Needs, wants and behaviour of “Drivers” and automated vehicles users today and into the future Contract No: 815001 D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Version 1.0
Work package WP1: “Driver”, traveller and stakeholder clustering a priori needs and wants and UC’s. Activity A1.1: User clusters, A1.2: User opinions, A1.4: research hypotheses, A1.5: transferability between modes, A1.6: taxonomy of knowledge and skills for AV operation, A1.7: Use Cases and priority scenarios Deliverable D1.1 Authors Karin Markvica (AIT), Paul Rosenkranz (AIT), Matina Loukea (CERTH), Evangelia Gaitanidou (CERTH), Evangelos Bekiaris (CERTH), Carlo Giro (IRU), Foteini Orfanou (NTUA), Eleni Vlahogianni (NTUA), George Yannis (NTUA), Filippo Fassina (DBL), Olivier Lenz (FIA) Status FINAL (F) Version 1.0 Dissemination Level Public (PU) Document date 31/05/2020 Delivery due date 30/04/2020 Actual delivery date 31/05/2020 Reviewers Michele Tozzi (UITP), Maria Panou (CERTH), Evangelia Gaitanidou (CERTH)
This project has received funding from the European Union’s Horizon 2020 Research and Innovation Programme under grant agreement no 815001.
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Version History Document history Version Date Modified by Comments 0.1 17.03.2020 AIT Initial draft with A1.1 content 0.2 09.04.2020 AIT, DBL, NTUA, FIA Input A1.5, A1.6 content 0.3 18.05.2020 CERTH Input A1.2, A1.3, A1.7 0.3.9 20.05.2020 IRU Input A1.1, A1.2 and A1.6 0.4 20.05.2020 AIT Final draft for peer review 0.5 29.05.2020 AIT, CERTH Final after review
Legal Disclaimer This document reflects only the views of the author(s). Neither the Innovation and Networks Executive Agency (INEA) nor the European Commission is in any way responsible for any use that may be made of the information it contains.
May 2020 2
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance Table of Contents
Table of Contents ...... 3
List of Figures...... 6
List of Tables ...... 7
Abbreviations List ...... 8
Executive Summary ...... 9
1. Introduction ...... 10 1.1. Purpose of the Document ...... 10 1.2. Intended audience ...... 10 1.3. Interrelations ...... 10
2. User clusters...... 11 2.1. Collection of stakeholders affected by automated vehicles and clustering of AV user’s ...... 11 2.1.1. Vehicle user ...... 14 2.1.2. Industry ...... 15 2.1.2.1. Automotive Industry ...... 15 2.1.2.2. Maritime Industry ...... 16 2.1.2.3. Aviation Industry ...... 16 2.1.2.4. Rail Industry ...... 17 2.1.3. (Remote) Operator ...... 17 2.1.4. Road User ...... 18 2.2. Updating the definition of VRUs ...... 18 2.2.1. Current definitions and underlying concepts ...... 18 2.2.2. Updated definition of VRUs ...... 19 2.3. Development of a common terminology for automated driving ...... 20 2.3.1. Common terminology for automated driving in Drive2theFuture ...... 20
3. Voice of customers’ survey ...... 22 3.1. Introduction ...... 22 3.2. Survey preliminary results ...... 23 3.2.1. Demographic information ...... 23 3.2.2. Analysis of user acceptance per mode ...... 28 3.2.2.1. Air transport ...... 28 3.2.2.2. Maritime transport...... 31 3.2.2.3. Rail transport ...... 34 3.2.2.4. Road transport ...... 37 3.3. Comparative remarks across transport modes ...... 44
4. Research hypotheses...... 46
5. Transferability from/to other modes ...... 51 5.1. Planning for the data collection ...... 52 5.1.1. Identification of the relevant topics ...... 52
May 2020 3
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
5.1.2. Development of the format for interviews and survey ...... 54 5.2. Interviews ...... 55 5.3. Survey ...... 57 5.4. Database of solutions ...... 57 5.5. Workshop with Consortium experts ...... 60 5.6. Results ...... 60 5.7. Next steps ...... 61
6 Taxonomy of Skills and Knowledge for AV operation ...... 63 6.1 Road Sector ...... 63 6.2 Rail Sector ...... 68 6.3 Maritime Sector ...... 72 6.4 Aviation Sector ...... 76
7 Use cases and priority scenarios ...... 79 7.1 Methodology ...... 79 7.2 Analysis of Use Cases...... 81 7.2.1 Training for road transport ...... 82 7.2.1.1 General description ...... 82 7.2.1.2 Analysis per Pilot site ...... 83 6.1.1 Training for rail transport ...... 95 6.1.1.1 General description ...... 95 6.1.1.2 Analysis per Pilot site ...... 96 6.1.2 Training for air transport ...... 98 6.1.2.1 General description ...... 98 6.1.2.2 Analysis per Pilot site ...... 99 6.1.3 Training for maritime transport ...... 100 6.1.3.1 General description ...... 100 6.1.3.2 Analysis per Pilot site ...... 101 6.1.4 Operators-based HMI& strategies for road transport ...... 102 6.1.4.1 General description ...... 102 6.1.4.2 Analysis per Pilot site ...... 103 6.1.5 AV Conspicuity HMI & strategies for interaction of automated road vehicles with non-equipped other road users ...... 107 6.1.5.1 General description ...... 107 6.1.5.2 Analysis per Pilot site ...... 108 6.1.6 In vehicle HMI & strategies for automated road vehicles ...... 113 6.1.6.1 General description ...... 113 6.1.6.2 Analysis per Pilot site ...... 114 6.1.7 Operators HMI & strategies for rail transport ...... 119 6.1.7.1 General description ...... 119 6.1.7.2 Analysis per Pilot site ...... 120 6.1.8 In vehicle HMI & strategies for rail vehicles ...... 122 6.1.8.1 General description ...... 122 6.1.8.2 Analysis per Pilot site ...... 123 6.1.9 Operators HMI & strategies for air transport ...... 124 6.1.9.1 General description ...... 124 6.1.9.2 Analysis per Pilot site ...... 125 6.1.10 Operators HMI & strategies for maritime transport ...... 127 May 2020 4
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
6.1.10.1 General description ...... 127 6.1.10.2 Analysis per Pilot site ...... 127 6.1.11 AV Conspicuity HMI & strategies for interaction of automated ships with other non-equipped vessels 129 6.1.11.1 General description ...... 129 6.1.11.2 Analysis per Pilot site ...... 129
7 Conclusions ...... 132
References ...... 133
ANNEX 1: Common terminology for automated driving in Drive2theFuture ...... 137
ANNEX 2: Drive2theFuture Voice of Customers Survey (full) ...... 147
ANNEX 3: Literature Review Template for Open Research Issues and Hypotheses ...... 174
May 2020 5
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance List of Figures Figure 1: Stakeholders in automated transport ...... 13 Figure 2: Effects of autonomous driving on the value chain (ATKearney, 2016: 13) ...... 15 Figure 3: New technology companies involved in the creation of ecosystems (ATKearney, 2016: 16) ...... 16 Figure 4: Age groups representation in the Drive2theFuture voice of customers’ survey ...... 24 Figure 5: Representation of European regions in the Drive2theFuture voice of customers’ survey ...... 24 Figure 6: Representation of countries in the Drive2theFuture voice of customers’ survey ...... 25 Figure 7: Respondents’ educational background ...... 26 Figure 8: Respondents’ annual gross income ...... 26 Figure 9: Modes and types of transport typically used by the respondents for their different types of travelling ...... 27 Figure 10: Respondents’ concerns regarding air automation ...... 29 Figure 11: Level of accidents as criterion for acceptance of automated air vehicles ...... 29 Figure 12: Cases where the use of automated flying vehicles (including drones) would be accepted by the respondents...... 30 Figure 13: Impact of automation in the employment of the air transport sector...... 30 Figure 14: Facilitation of PwD mobility by the automation of air transport ...... 31 Figure 15: General opinion of respondents regarding automated air vehicles ...... 31 Figure 16: Respondents’ concerns regarding maritime automation ...... 32 Figure 17: Cases where the use of automated vessels would be accepted by the respondents ...... 33 Figure 18: Level of accidents as criterion for acceptance of automated vessels ...... 33 Figure 19: Impact of automation in the employment of the maritime transport sector ...... 34 Figure 20: General opinion of respondents regarding automated vessels ...... 34 Figure 21: Respondents’ concerns regarding rail automation ...... 35 Figure 22: Impact of automation in the employment of the rail transport sector ...... 36 Figure 23: Cases where the use of automated vessels would be accepted by the respondents ...... 36 Figure 24: General opinion of respondents regarding automated and self-driving trains ...... 37 Figure 25: Conditions under automation in road vehicles would be preferred by drivers/ passengers ...... 38 Figure 26: Conditions under automation in road vehicles would be preferred by drivers/ passengers of non-automated road vehicle, in mixed flow with Automated Vehicles (AVs) ...... 38 Figure 27: Conditions under automation in road vehicles would be preferred by Vulnerable Road Users in mixed traffic with AVs and non AVs ...... 39 Figure 28: Respondents’ concerns regarding road automation ...... 40 Figure 29: Agreement rates of users regarding the presence of driver in public transport means ...... 41 Figure 30: Agreement rates of users regarding sped restrictions in automated road vehicles ...... 41 Figure 31: Level of accidents as criterion for acceptance of automated road vehicles ...... 42 Figure 32: Cost of automated vehicles as acceptance criterion of the survey respondents ...... 42 Figure 33: Need for reskilling of the road users to be able to use automated vehicles ...... 43 Figure 34: General opinion of respondents regarding automated and self-driving road vehicles ...... 43 Figure 35: General opinion of respondents regarding automated and self-driving vehicles of all modes ...... 44 Figure 36: Distribution of solutions across topics (survey) ...... 57 Figure 37: Screenshot of the Database of solutions...... 59 Figure 38: Autonomous levels in road transport (SAE, 2018) ...... 65 Figure 39: Autonomous levels in rail sector (IEC, 2009) ...... 69 Figure 40: Autonomous levels in maritime sector (IMO, 2018) ...... 73 Figure 41: Autonomous levels in maritime sector (Lloyd’s Register, 2016, 2017) ...... 73 Figure 42: Autonomous levels in aviation sector (Lloyd’s Register, 2016, 2017) ...... 76 Figure 43: Parameters combined for the development and description of the Drive2theFuture Use Cases ...... 79 Figure 44: Types of stakeholders that participated in the Drive2thFuture UCs workshop...... 80 Figure 45: Modes representation in the Drive2thFuture UCs workshop ...... 80 Figure 46: Prioritisation of UCs assessment criteria ...... 81
May 2020 6
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance List of Tables Table 1: Categories of vulnerable road users (Methorst, 2003: 5)...... 19 Table 2: VRU candidates and qualifications (Siulagi et al., 2016: 285) ...... 19 Table 3: Explanation of sources used for developing the common terminology for automated driving in Drive2theFuture ...... 20 Table 4: Dissemination channels for the Drive2theFuture Consumer Acceptance Survey ...... 23 Table 5: Literature sources for the analysis of Drive2theFuture research hypotheses and priorities ...... 46 Table 6: Drive2theFutrure research hypotheses and priorities ...... 49 Table 7: Selection of the most relevant goals, values, trends and action plans for the future transport ...... 52 Table 8: Final selection of relevant topics for Drive2theFuture ...... 53 Table 9: Format for the data collection ...... 54 Table 10: Interviews ...... 55 Table 11: List of factors and question to evaluate the transferability of solutions ...... 60 Table 12: Topics included in the EC Communication ‘On the road of automated mobility: an EU strategy for mobility of the future’ (ERTRAC CCAM Roadmap 2019) ...... 66 Table 13: Skills and Knowledge for AV operation in the road sector ...... 67 Table 14: Skills and Knowledge for AV operation in the rail sector ...... 71 Table 15: Skills and Knowledge for AV operation in the maritime sector ...... 75 Table 16: Skills and Knowledge for AV operation in the aviation sector ...... 77 Table 17: Drive2theFuture Use Cases ...... 79 Table 18: Ranking of Drive2theFuture Use Cases after the workshop ...... 81
May 2020 7
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance Abbreviations List
Abbreviation Definition AAWA Advanced Autonomous Waterborne Applications AR Augmented Reality ATC Automatic Train Control AV Automated Vehicles C-ITS Cooperative Intelligent Transport Systems DoA Description of Action ERTMS European Rail Traffic Management System ETP European Technology Platform FoT Field Operational Tests GoA Grades of Automation HMI Human Machine Interface HQ Headquarter IT Information Technology ITS Intelligent Transport Systems KPI Key Performance Indicator OEM Original Equipment Manufacturer PwD Persons with Disabilities SPaT/MAP Signal Phase and Time and Map Data TMC Traffic Message Channel UC Use case VMS Variable Message Signs VR Virtual Reality VRU Vulnerable Road User WoZ Wizard of Oz WTH Willingness to Have WTP Willingness to Pay
May 2020 8
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance Executive Summary
This Deliverable provides insights on the user clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance, as revealed by Drive2theFuture work as a main outcome of the first project year. The work performed will be further used as a baseline for the work packages to follow and can be divided into six main topics: user clustering, voice of customers’ survey, research hypotheses, transferability from/to other modes, taxonomy of skills and knowledge for AV operation, use cases and priority scenarios. All performed activities and results are described below per topic and chapter. • User clusters: Clustering of the different stakeholders in automated transport. This covers vehicle users, (remote) operators, other road users, legislation/licensing authorities and industry. First insights on their action frame and underlying motivation are provided as well as a redefinition of the term VRU. A comprehensive terminology of AV related terms, covering around 120 definitions was realised and will be used as reference in all project activities hereafter for common understanding purposes. This topic is covered in Chapter 2. • Voice of customers’ survey: To gain insights on the opinion of the public, a comprehensive online survey was performed in different countries. It included a generic part and a specific use case-based part for each transport mode (road, rail, maritime and aviation) and key scenario. The preliminary survey results are presented in Chapter 3. • Research hypotheses: Key research hypotheses per transport mode and AV function/level have been identified, based upon the initial list within the DoA. All results coming from this research are reported in Chapter 4. • Transferability from/to other modes: Innovative solutions in various transport modalities have been collected via interviews and an online survey. The collected solutions have been further described and categorised in a Database of Solutions (Figure 37). Common issues, approaches and lessons learnt of solutions across modes have been investigated applying a cross-fertilisation methodology and all results are described in Chapter 5. • Taxonomy of skills and knowledge for AV operation: The relevant skills and knowledge have been identified for each transport mode and automation level. The operation task was therefore decomposed into the cognitive domain and key related factors were extracted, revealing the anticipated skills of the operator as well as training needs. These outcomes are shown in Chapter 6. • Use cases and priority scenarios: Use cases and pilot scenarios have been formulated, with good balance among all modes of transport while covering key research hypotheses, also linked to the Drive2theFuture pilots. They comprise information –among others- on transport mode, AV level, key user clusters, relevant research issues and hypotheses, connected HMI to be developed and/or tested, related training requirements and operator skills/knowledge levels, operational, behavioural and legal risks, etc. The Use Cases are presented in detail in Chapter 7.
As autonomous transport is a very dynamic field, all taxonomies should be understood as snapshot of the current situation. The collection of terms will be continuously updated and expanded within the project.
May 2020 9
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance 1. Introduction 1.1. Purpose of the Document This Deliverable aims to identify and cluster the categories of “drivers”, travellers and stakeholders involved in or affected by autonomous vehicles, recognize their needs and wants and define relevant use cases, considering issues of transferability of solutions between different transport modes. The document therefore includes: • the results of user clustering (A1.1) • user opinions captured through the voice of customers surveys (A1.2) • the research hypotheses (A1.4) • issues of transferability between modes, including a comparative study between different types of vehicles and modes (A1.5) • the taxonomy of knowledge and skills for AV operation (A1.6) • the project’s identified Use cases and priority scenarios (A1.7)
The work performed therefore serves to clarify uncertainties and provides a common understanding that is essential for the work packages to follow. It is a reference work that can not only be used within the project but can also serve as a guide for interested parties. 1.2. Intended audience This report is a summary of the activities carried out under the umbrella of work package 1 regarding not only the autonomous systems in question, but also the involved individuals behind used technology. It is therefore addressed primarily to the funding agency. However, the information collected is also of interest to the general public, which means that the report can be used by researchers and citizens as a reference work. Care has therefore been taken to ensure that all technical terms are adequately described and that as little prior knowledge as possible is required to understand the facts described. 1.3. Interrelations As first work package, WP1 provides a scientific basis for the future project work and is therefore interwoven with other work packages and tasks. A strong connection between the project’s identified “Use cases and priority scenarios” and the project’s “Pilot tests” has to be pointed out.
May 2020 10
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance 2. User clusters
In current non-automated transport systems, the driver is responsible for all actions from driving, reacting and making decisions based on influences from the vehicle itself or the outside world. When it comes to different levels of vehicle automation, described as substituting challenges and responsibilities with semi- or fully-automatic processes, a new definition of the driver itself has to be made. In the case of transport automation innovations, the driver remains in the core, however the role and the possible impersonators of a “driver” – now put in quotation marks – differs significantly. Who can be the “driver” may vary according to automation level, i.e. in Level 5 automation the vehicle can be operated by someone that does not necessarily own a driver’s license. Also, regarding professional drivers (e.g. bus drivers), their role in higher levels of automation becomes different, involving tasks related more to the supervision of the system rather than the driving task as such. Another important issue is the interaction with other vehicles, especially non-automated ones, as well as other road and city infrastructure users (e.g. vulnerable road users - VRUs). Therefore, an updated definition of VRUs is developed in Drive2theFuture. To cover all these aspects, user clustering involves not only a categorization of the possible affected users’ (and stakeholders’) groups, but also considers the transformation of the roles of each group (e.g. with high levels of automation a VRU may be found in the role of the “driver”) as well as their particular needs and preferences in terms of automation. This chapter consists of the following items: • Identification of all traffic participants affected by automated vehicles and clustering of AV user’s • Updating the definition of VRUs • Development of a common terminology for automated driving
The systematically defined user clusters of AVs as well as traffic participants affected by automated vehicles will cover both professional (automated bus, taxi, truck, drone, ship, rail driver/pilot, rail signaler, TMC operator, etc.) and private (automated vehicle “driver”, passengers, other vehicle driver/passengers, VRUs, etc.) users and will help to guarantee they are all considered by the project activities dealing with relevant HMI (WP3), training and incentives (WP4), evaluation (WP5) or other project scopes. Additionally, the updated VRU definition as well as the overall unique terminology will be used as reference by all project activities for common understanding. This document should serve as a “living document” as it is quite likely that there will be new findings obtained within the project duration which should be included either in the user clusters or in the common terminology.
2.1. Collection of stakeholders affected by automated vehicles and clustering of AV user’s AVs will affect traffic participants of all modes - as well as all travellers - both drivers and passengers of AVs as well as the ones travelling in non-automated vehicles (surrounding traffic) in mixed flows, along with all other traffic participants. Different scenarios elaborated for Austria (MOB 2040, SozA) show a broad range of stakeholders involved and give a glimpse on new work profiles due to increased automation such as algorithm insurers, truck guides and drone pilots among the most prominent (Dörfelt/Scherf, 2017). Many predict that there will be a need for drivers in the future, since 100% automation will only be possible in the very distant future. Especially for the first and last mile, it will be necessary for a long time to control May 2020 11
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance the vehicle either locally or by remote control. It is also assumed that there will be remote supervisors instead of drivers. As for the current state of persons involved in automated transport, different stakeholder groups are part of this change (Figure 1). It is shown, that public transport drivers are often used in a remote way. This doesn’t mean, that all drivers of this part of the diagram are remote, it means, that in this field of transportation a greater level of automation is already implemented and most of the public transport system could replace their drivers with remote drivers.
May 2020 12
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Figure 1: Stakeholders in automated transport
May 2020 13
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
2.1.1. Vehicle user The user group is very diverse, as it not only covers professional and private drivers, but also passengers/travellers and in some cases crew members that ensure well-being and safety of the passengers. Our main target within the project are the drivers and travellers. Therefore, we will concentrate on the challenges faced by these groups.
One of the most common visions for the future comprises the assumption that drivers of automated vehicles will soon realize that the patterns of action (e.g. for carrying out steering manoeuvres) are no longer needed and may be forgotten. Instead, new skills (e.g. system monitoring) and a new understanding of the system must be acquired. That leads to a new understanding of the role of the driver in AV and ranges from the active operator to the passive supervisor of the vehicle (Wolf, 2015). This on-call service requires different cognitive processes than the continuous driving of the vehicle would do, but also offers the opportunity to use the time for other tasks, in the case of professional drivers, compiling administrative documents while inside the vehicle (if motion sickness is not an issue). In aviation, this development is quite similar. Cockpit computer systems take over the tasks from the cockpit crew which results in a decrease of manual tasks and increased activities of programming and monitoring aircraft automation. However, this assumes that all situations are either solved by the automated vehicle/system or solved by some kind of ‘supervision-action’ like emergency shutdown. It is unclear how the ‘complex situations’, which are now solved by a human taking over the entire driving task (e.g. airplane-pilot taking over in case of system failures; or unforeseen traffic/road situations where the ‘operator’ takes driver the driving task), are solved in the future. This becomes even more problematic as the driver/pilot is usually trained to do ‘driving’ in the classical sense but it has to be definitively repeated under supervision at certain levels of automation. Vehicle drivers of conventional vehicles have an additional disadvantage in that they must get used to the behaviour of automated vehicles and cannot use eye contact to mitigate unclear situations. This might be easier to manage in more ‘controlled’ road environments, such as in Europe, than in rather chaotic street situations in countries such as India or Egypt. For passengers and travellers, the use of AV intervenes less in the routines than that it conveys a different feeling of safety and security. If trust is placed in the vehicle, relatively little changes occur. On the contrary older, mobility impaired and very young people (children, teenager) or people who do own a driving licence will benefit from the change as they become more mobile. AVs are also available to those that have a problem with vision, reaction speed and reaction safety (IGPmagazin, 2019). The benefit attached to AV might lead to higher technology usage and its acceptance by older persons. According to Fraedrich et al. (2016) this is not necessarily the case for mobility impaired people as their interest in AVs is even lower than among the average road user. Moreover, the emergence of autonomous driving will impact professional drivers and commercial road transport companies. There are multitude of aspects that need to be taken into consideration. These include the replacement of some tasks that the driver performs and the necessary investment in re-training drivers. Bieg et al. (2020), after having examined the behaviour of professional drivers while in SAE Level 2 and 3 vehicles, concluded that drivers struggled to adapt to respective automation requirements. On the other hand, however, an increase in the level of technology in the vehicle could attract drivers. Automation could positively impact the overall perception of the commercial road transport sector while also leading to better working conditions (IRU Tackling Driver Shortage in Europe Report, 2019). Professional drivers should be
May 2020 14
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance guided throughout the transition to higher levels of automation and their acceptance is key for a successful uptake of autonomous vehicles (Hartwich et al., 2018).
2.1.2. Industry 2.1.2.1. Automotive Industry Autonomous driving is said to have a high potential to disrupt the automotive industry (ATKearney, 2016). The existing value chain is about to be replaced by hub-and-spoke. Most important players are therefore Original Equipment Manufacturers (OEM), IT suppliers, online players, telecom companies, device manufacturers and tier-x suppliers which have an equal share in the process (Figure 2).
Figure 2: Effects of autonomous driving on the value chain (ATKearney, 2016: 13)
Today’s value of a vehicle concentrates on hardware (90% of total value) which is about to change. AVs are centred on software (40%) and entertainment content (20%) which makes the introduction of new players inevitable. Technology companies in app development and operating systems are therefore important figures in the process. New alliances are being formed, which create powerful ecosystems in the future.
May 2020 15
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Figure 3: New technology companies involved in the creation of ecosystems (ATKearney, 2016: 16)
Autonomous driving is not the only field these traditional automotive players are involved in – e.g. Rolls- Royce aims to build autonomous ships.
2.1.2.2. Maritime Industry The shipbuilding industry is rather traditional in terms of technology involvement in existing production processes. Due to the economic risks involved, a high level of automation has not been the main priority for a long time (Andritsos/Perez-Prat, 2000). By now, companies such as Rolls-Royce Marine and Nippon Jusen work on autonomous ships (Safety4Sea, 2019). Apart from autonomous ships, Virtual Reality, Cyber Security, e-learning and digital certification are topics of relevance for the maritime industry (Safety4Sea, 2019). As the shipbuilding involves only few series-production features, the implementation of cost-effective automation is challenging. For a long time, short-term productivity has been more easily achieved through process optimization (Andritsos/Perez-Prat, 2000). By now, it is estimated that the introduction of automation of container vessels could save tens of thousands of dollars each day (Martek Marine, 2018). Possible savings are also related to the port infrastructure for the operation of container cranes.
2.1.2.3. Aviation Industry In the aviation industry, manufacturers and airlines are in favour of automation. It would not only generate cost savings (e.g. through redesign of the front in a more aerodynamic manner) but also solve the current shortage of pilots available (Rice/Winter, 2019). Some companies are currently working on fully autonomous aircrafts of different sizes (e.g. Amazon). Hand in hand with the adjustments to the legal framework necessary for the use of unmanned aerial vehicles, there is a range of considerations on how to use them for passenger transport. Three studies should be highlighted
May 2020 16
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance in this context, which make implementation within the next 15 years appear likely, although some assessments should be treated with caution:
• Porsche Consulting (2018): The Future of Vertical Mobility. Sizing the market for passenger, inspection, and goods services until 2035. A Porsche Consulting Study. • Roland Berger (2018): Urban air mobility. The rise of a new mode of transport. • UBER Elevate (2016): Fast-Forwarding to a Future of On-Demand Urban Air Transportation.
In the logistics sector there are a number of useful applications, including, for example, the person unloading the goods, maintenance technicians assembling the delivered spare parts, or similar. In passenger transport, the usage of autonomous air taxis is aspired by companies and start-ups such as Velocopter (Germany), Lilium (Germany), Volare (Austria), Matternet (US) or the helicopter company Bell (Rice/Winter, 2019).
2.1.2.4. Rail Industry The rail industry faces recruitment problems as there are too few engine drivers. Shift and night work as well as low salaries make it an unattractive profession for young people (Frey, 2017). Cost intensive processes with potential for automation can be found in rail freight transport (e.g. shifting, transhipment, train formation). A number of processes that could be automated and therefore contribute to safe working conditions are sequencing (high optimization potential), control of the departure of the train, control of vehicle technical data (doors, valves) via sensor and RFID technology, detection of the operational readiness of brakes, brake control, location and condition monitoring of the wagons, transhipment process in the terminal etc.
2.1.3. (Remote) Operator Under the guise of “(remote) operators” many different types of stakeholders can be grouped. One can differentiate between public suppliers of transport services, private suppliers of transport services, national entities responsible for the provision/maintenance of infrastructure and different cooperation networks. (Remote) Operators are particularly interested in increasing efficiency with regard to personnel deployment. Examples can be found in various areas, such as waterborne and air transport. The Advanced Autonomous Waterborne Applications (AAWA) initiative aims to increase in loading capacity due to fewer crew members. Autonomous control should be enabled by a combination of cameras, infrared, radar, sonar, drones, optical distance and speed measurement. Speech recognition, weather routing, global navigation satellite systems, smart screens and virtual reality are further topics with potential, especially as the command centre is not on board but on shore (remote operators). Current investigations proclaim that 10 to 14 employees in the control centre are able to monitor more than one hundred ships in real time (Brunel, n.a.) given that powerful satellite connections ensure data transmitted in real time to the control centre. On the ship, the board computers navigate in autonomous mode and the (potential remote) captain only takes control in critical situations. For air traffic control it is envisaged that some regional airport does not have staff in the tower but special cameras and special security technology for remote monitoring. Germany, Sweden and Australia planning on taking this approach in the future (WELT, 2018).
May 2020 17
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance 2.1.4. Road User Automation at sea, in the air or on rail will by no means affect as many people (not working in the field) as automation on the road does. Pedestrians, cyclists and users of small vehicles (scooter, Segway etc.) as well as conventional (non-automated) vehicles, will be of major importance to the roll-out of new technologies facing challenges in interpreting the AV’s behaviour as conventional drivers do. This also applies to any tourists and newcomers in town not familiar with the environment, transport system and/or language. Also, Vulnerable Road Users (VRU) which are a target group for AVs are worth mentioning as they must be redefined in terms of automated mobility. Especially for people who are currently experiencing mobility limitations, the increasing automation of vehicles is seen as an opportunity to achieve more self-determination, for example by giving all age groups equal access to mobility options (Frison et al., 2017). Further advantages of automated mobility are: improved road safety through lower accident probability and accident severity (Li/Kockelman, 2016), more efficient and therefore more environmentally friendly traffic flows, and greater comfort through better usability of the travel time (Litman, 2015; Trommer et al., 2016). However, these advantages could also have their price: comfortable self-propelled vehicles would also be attractive for people who are currently still travelling on foot, by bicycle or in mass transport vehicles. This, together with the newly developed target groups who themselves cannot currently drive a car, would lead to a much higher demand for private motorized transport and thus at least partially compensate for the positive effects (Alonso Raposo et al., 2017). In any case, the hoped-for advantages for people without driving skills or permission will largely only be possible with full automation (Level 5), since until then the drivers themselves will still be responsible for at least parts of the route.
2.2. Updating the definition of VRUs The introduction of AVs will change the framework and the role of traffic participants significantly. An important issue is the interaction with other vehicles, especially non-automated ones, as well as other users of the transportation system (e.g. VRUs). Therefore, a new definition of VRU will be needed.
2.2.1. Current definitions and underlying concepts To achieve a VRU redefinition towards Automated Vehicles (AV) per transport mode and AV level it is essential to take stock of current definitions and underlying concepts. As for now, VRU can be defined (as)… … non-motorised road users, such as pedestrians and cyclists as well as motor-cyclists and persons with disabilities or reduced mobility and orientation (European Union, 2010). … road users who are most at risk for serious injury or fatality when they are involved in a motor-vehicle- related collision (US DOT FHWA, 2019). … with regard to the amount of protection in traffic (e.g. pedestrians and cyclists) or by the amount of task capability (e.g. the young and the elderly) (SWOV, 2012). … a term applied to those most at risk in traffic. Thus, vulnerable road users are mainly those unprotected by an outside shield (OECD, 1998). … road user who is present in a crash involving vehicles which do not have a protective shell (Avenoso, 2005).
May 2020 18
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
VRU are distinguished from other road users by the amount of external protection, the task capability and the resilience. Some also take the speed difference as indicator (Sucha, 2014). Most definitions include pedestrians and cyclists, children and adolescents as well as motorised two wheelers as they have less protection and/or have physical disadvantages compared to the average road user. Other definitions also include seniors, mobility impaired persons, scooter riders, skateboarders and Segway riders. Depending on the broadness of the definition, VRUs currently includes the categories given in Table 1.
Table 1: Categories of vulnerable road users (Methorst, 2003: 5) Functional groups of vulnerable road users 1. Pedestrians 8. Pre school children 15. Handicap – lost function 2. Pedestrians Plus 9. Elementary school (4-8 yrs. old) 16. Limited stamina 3. Bicyclist 10. Elementary school (9-11 yrs.) 17. Limited perception 4. Slow moped drivers 11. Special schools 18. Mentally handicapped 5. Moped drivers 12. Secondary school (12-15 yrs.) 19. Motor handicapped 6. Motorcycle drivers 13. 16-17 years old 20. Foreigners 7. Special vehicle drivers 14. 18-25 years old 21. Addicted / homeless
The usage of the VRU categories mentioned is not without critique. Main topics in this regard are (Methorst, 2003): • overlapping groups (e.g. elderly and mobility impaired); • age as an insufficient indicator in terms of the necessary financial and manpower resources and the cost effectiveness of measures; • age as an insufficient indicator for illness, physical and mental health condition; • heterogeneity within groups such as among elderly people and mobility impaired; • traffic roles as starting point instead of user-centred conception.
There are indications that automated vehicles might bring a new type of “vulnerability” that includes those not connected via devices and those not able to use the technology properly (CS MARE, 2019). Table 2 shows candidates and qualifications in terms of road automation.
Table 2: VRU candidates and qualifications (Siulagi et al., 2016: 285) VRU candidates Qualification Pedestrians Physical vulnerability and intersecting or shared right-of-way with vehicles Bicyclists Physical vulnerability and intersecting or shared right-of-way with vehicles Motorcyclists, scooters, Physical vulnerability and intersecting or shared right-of-way with vehicles; electric bikes however they present additional challenges such as white-lining behaviour Skateboarders, Segway Physical vulnerability and intersecting or shared right-of-way with vehicles riders Seniors Physical vulnerability and prone to cause more collisions Disabled Depends on the particular disability (e.g., slower as pedestrians, deafness, or blindness)
2.2.2. Updated definition of VRUs Vulnerable Road Users (VRU) are defined in the ITS Directive as "non-motorized road users, such as pedestrians and cyclists as well as motor-cyclists and persons with disabilities or reduced mobility and orientation", due to their reduced passive safety protection or physical vulnerability. May 2020 19
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Taking into account the current definitions and underlying concepts as mentioned in chapter 2.2.1 as well as the change in the role of traffic participants by the introduction of AVs, the following updated definition of VRUs was developed in Drive2theFuture:
Automation related Vulnerable Road Users are defined as those that have a higher risk of accident involvement or vulnerability in case of it, due to their lack of connectivity (i.e. non-connected users having low conspicuity vis a vis the automated road traffic environment), or lack of design compatibility (i.e. out- of-position passengers of automated vehicles).
2.3. Development of a common terminology for automated driving A necessary starting point for defining and clustering road users is a basic terminology for driving automation systems. This terminology helps to clarify the role of the (human) driver engaging in driving automation systems, and it provides clarity and stability of communication on the topic of driving automation. In this context, several activities for creating terminology guidelines have been started (see i.e. [SAE-J3016], [SAE-J3063], [DINSAE-91381]). In addition, documents on the development strategy in automotive industry, railway industry, ship industry and aviation industry contain definitions of emerging and future functions and services. Furthermore, research project reports constitute a source for new technical terms and functions. Therefore, selected guidelines and strategy documents have been analysed. A database has been created containing the term, its abbreviation, a definition of the term and its source document. The basic version of the database contains about 120 terms. Whereas one part of the terms applies to all automation levels, the other part of the terms is specific for one or more automation levels. Hence, applicability for one or more automation levels can be used for grouping the terms. In case the terminology database is going to be extended with new terms, the following steps are recommended: • The source of the term and its definition should be documented. • It should be noted in case the term is a synonym of an existing term. • It should be checked whether the term is unambiguous and precise. A list of ambiguous and imprecise terms in the context of driving automation and an explanation which problems may occur when using these terms can be found in [SAE-J3016].
2.3.1. Common terminology for automated driving in Drive2theFuture To ensure a common terminology, all relevant terms have been collected and categorised. The term collection comprises more than 120 terms and can be found in ANNEX 1. Table 3 indicates which sources were used for developing the common terminology for automated driving in Drive2theFuture.
Table 3: Explanation of sources used for developing the common terminology for automated driving in Drive2theFuture Source ID Title Author Distributor Date SAE-J3016 Taxonomy and Definitions for Society for Automotive SAE June 2018 Terms Related to Driving Engineering International J3016:JUN2018 (SAE International) May 2020 20
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Source ID Title Author Distributor Date Automation Systems for On- Road Motor Vehicles SAE-J3063 Active Safety Systems Terms & Society for Automotive SAE November Definitions Engineering International J3063:NOV2015 2015 (SAE International) DINSAE- Terms and Definitions Related Society for Automotive DIN SAE SPEC June 2019 91381 to Testing of Automated Engineering International 91381:JUN2019 Vehicle Technologies (SAE International) WiPed- Wikipedia, The Free October 201910 Encyclopedia 2019 ERTRAC- Automated Driving Roadmap ERTRAC Working Group ERTRAC May 2017 2017 (V7.0) “Connectivity and Automated Driving” ProjectDef definition developed in the project
May 2020 21
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance 3. Voice of customers’ survey 3.1. Introduction Regarding the integration of automation into the European transport system, a crucial issue is the acceptance of all transport users and this is one of the main focus areas of Drive2theFuture project. Even though technology is gradually progressing, a main point to be determined is whether humans are ready to abandon their old habits (e.g. the driving task and/or even the car ownership – in combination with car sharing/pooling applications) and also board on a vehicle with no driver. For the investigation of this tendency, several surveys have been conducted, in order to pulsate the public’s opinion. For example, the EC 2020 Eurobarometer survey (European Commission, 2020) showed that 58% of the sample have heard, seen or read something about automated vehicles, with proportion differences at national levels. The majority of respondents stated that they would not feel comfortable in a fully automated vehicle without the supervision of a human operator but 70% of the respondents would feel comfortable traveling with the supervision of a human operator in it. However, acceptance of automation in the driving task seems to be evolving with time as, according to the 2017 (Deloitte, 2017) and 2018 (Deloitte, 2018) Deloitte global automotive consumer studies, people throughout the world are becoming convinced that travelling with autonomous vehicles is safe. The acceptance of all relevant transport user groups’ is also considered throughout the Drive2theFuture project and is going to be performed in a multi-parametric way; by means of a voice of customer survey that has been developed and circulated within WP1, the sentiment analysis performed within A2.5 to the involvement of users in the pilot testing in WP5; with particular emphasis to People with Reduced Mobility (PRMs) and other vulnerable user groups (redefining the relevant VRU concept described in Section 2.2). The purpose of the Voice of Customers’ survey that has been developed within Activity 1.2 is to explore the opinions of all transport users about acceptance of Autonomous Vehicles for different transport modes, respecting a good balance of gender, geographical coverage, transport modes and user clusters.
The survey has been designed and developed with the contribution of all Drive2theFuture partners, especially ensuring the necessary representation of all transport modes, in order to include the proper description of automation levels of each mode, while covering the different needs of each mode by the questions. The survey has been structured in a modular way, with four main areas corresponding to each transport domain (aviation, maritime, rail, road). Each area includes a simple description of what automation means for each mode, with illustrative examples and a short list of multiple choice questions, concerning several aspects of automated vehicles and their prerequisites for the acceptance, including their general concerns, their opinion on related employment issues, impact of mobility of persons with disabilities, etc.. An introductory part also exists, facilitating the collection of the demographic information of the respondents, as well as of data concerning their travelling patterns and preferences. All the information collected is being treated by the researchers in an anonymous and confidential manner, in accordance to the ethical requirements defined within the project, while all demographic information will be used only to contextualize the statistical analysis of the aggregated results, and they will not be published or used in any form, other than the statistical analysis of the survey’s results.
The survey aims to depict and represent, to a large extend, the transport users' opinions about the automated vehicles of all different levels and scopes (e.g. for public transport vehicles, private cars, drones for urban deliveries, air transport vehicles for passenger inter-city flights, cargo ships, rescues boats, passenger ships, passenger and freight trains, etc.) and also provide a comprehensive estimation of their willingness to use them under different circumstances. In order for this to be achieved, and attract responses of as many user groups possible, the survey developed has been translated in 18 different European
May 2020 22
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance languages, so as to facilitate the respondents and give them the opportunity to provide their responses to their mother language or their language of preference. In ANNEX 2 the full survey questionnaire is included. The survey has been integrated in the project’s website (http://www.drive2thefuture.eu /2020/01/28/welcome-to-our-survey/), while for its dissemination all partners have joined forces, using many different channels, as the ones indicated in the list below:
Table 4: Dissemination channels for the Drive2theFuture Consumer Acceptance Survey ➢ Project’s level ✓ Drive2theFuture newsletter - http://www.drive2thefuture.eu/proyectos/newsletter-april-2020/
✓ Drive2theFuture website and Social Media (Twitter, LinkedIn) ✓ Presentation on the 1st Drive2theFuture Workshop that took place at 6th of March 2020 in Brussels. ➢ Partners’ level ✓ Newsletters of organisations ✓ Members of Associations (e.g EURNEX, FIA, IAM RoadSmart, WEGEMT, IRF, HUMANSIT , IRU, PZM, UITP) ✓ Social media (Facebook, Twitter, Instagram, LinkedIn) ✓ Partners’ websites ✓ FIA Forum for Mobility and Society ✓ Other EU projects consortia (e.g. ARCADE, AUTOPILOT, C-MobILE, L3Pilot, etc.) ✓ Transport Research Associations (e.g. ECTRI, FERSI, etc.) ✓ Press Releases (e.g. https://www.gocar.gr/news/feed/29971,Aytonoma_oxhmata_Pes_thn_apoyh_soy.html) ✓ Personal contacts
Currently, an impressive number of nearly 10.000 questionnaires have been completed, leaving us however behind the aimed numbers, due also to the COVID-19 crisis outbreak that has caused the cancellation or postponing of many conferences and dissemination events that we were planning to attend, reaching greater audiences. To this extend, the dissemination of the survey is still on-going and its final results will be provided at Month 18, through an updated version of D1.1 that will be generated exclusively for this purpose. 3.2. Survey preliminary results 3.2.1. Demographic information From the respondents of the survey, 72% are male and 26.5% female, while the rest 1.5% has selected not to declare gender or chose the option “other”. Regarding the different age groups, there is an almost equal representation of the 25-35, 36-45 and 46-60 groups, however significant shares of participants exist also from the remaining age clusters, providing to the survey results to the desirable opportunity to depict the opinions of different age categories and cultural backgrounds.
May 2020 23
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
30%
25%
20%
15%
10%
5%
0% 18-24 25-35 36-45 46-60 >60
Figure 4: Age groups representation in the Drive2theFuture voice of customers’ survey
As for the representation of the European countries from which replies have been collected so far, the majority of the answers comes from Central European countries (41.46%), while countries from Eastern and Southern Europe are almost equally represented (23.47% and 21.27% respectively). In addition, 12.79% of the replies come from Northern Europe, while a small share of 1% originates from other countries outside Europe (e.g India, USA). Figure 5 below depicts the representation of the aforementioned regions, while Figure 6 presents the distribution of questionnaires per country.
1,01%
12,79%
41,46% 21,27%
23,47%
North Europe Sourth Europe East Europe Central Europe Other
Figure 5: Representation of European regions in the Drive2theFuture voice of customers’ survey
May 2020 24
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Other United Kingdom Sweden Spain Slovenia Slovakia Romania Portugal Poland Norway Netherlands Malta Luxembourg Lithuania Latvia Italy Ireland Hungary Greece Germany France Finland Estonia Denmark Czechia Cyprus Croatia Bulgaria Belgium Austria
0% 5% 10% 15% 20% 25%
Figure 6: Representation of countries in the Drive2theFuture voice of customers’ survey
Through the whole duration of the survey dissemination (which was launched on 24 January 2020), the monitoring of the replies per country is being updated weekly, in order for more targeted actions to be organised regarding the representation of countries; thus ensuring the geographical balance required. Such actions will continue during the remaining time of the survey being online. Regarding the education background of the respondents, 49.8% are of Master and/or PhD level, 20.12% have a bachelor degree, while also 26.31% are primary/elementary/high school graduates or have received technical training. Thus, the sample is somehow biased towards highly educated people, but still has a fair representation of all types of users.
May 2020 25
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Other
Ph.D.
M.sc.
Bachelor degree
Trade/technical training
Primary/Elementary/High Shool Degree
0% 5% 10% 15% 20% 25% 30% 35%
Figure 7: Respondents’ educational background
Moreover, the majority of the respondents (58.41%) are full-time employed and the 44.9% are operating a vehicle (even if rarely) as an aspect of their work (including car, motorcycle, minibus, tractor, etc.). Also, the various income clusters (Figure 8) reveal a good socioeconomic spread of the sample.
prefer not to say
more than €100,000
€81.000-€100.000
€61.000-€80.000
€41.000-€60.000
€26.000-€40.000
€10.000-€25.000
Below €10,000
0% 5% 10% 15% 20% 25% 30%
Figure 8: Respondents’ annual gross income
As for the modes and types of transport that the respondents typically use for the different kind of trips, according to Figure 9 below, the passengers car is presented as the most frequently used vehicle type (with an average of 65.24% of users stating that they use it for most of their trips), followed by the means of public transport (average: 33.75%) and bicycle or walking (average 26.29%). As expected, airplanes present their biggest rates for business travel and vacation, while ships mostly for vacation.
May 2020 26
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
I do not take such trips
Sharing services, i.e. car sharing
Ship
Airplane
Bicycle or walking
Motorcycle or scooter
Taxi
Public transport
Passenger car
0% 10% 20% 30% 40% 50% 60% 70% 80%
Vacation Errands Leisure/ Social Business Travel Commuting
Figure 9: Modes and types of transport typically used by the respondents for their different types of travelling
May 2020 27
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Regarding the respondents relation to technology, most of them (60.3%) stated that they usually are in prone of trying a new technological product, while almost all of them (95%) have heard about automated and autonomous vehicles, but again the vast majority of them (80%) has never had any relevant experience. Still, 20% stating some previous experience constitutes a non-representative high rate and is attributed to several respondents stemming for the project’s pilot sites, where relevant pilot services pre-existed. 3.2.2. Analysis of user acceptance per mode Moving on to the analysis of the users’ acceptance in relation to automated vehicles, it needs to be mentioned that at the beginning of this section in the survey, a note has been included informing the respondents about what we mean by automation, providing g the following definition:
Automation is the use of control systems and information technologies reducing the need for human intervention. Several trends have been evolving across sectors around digitalisation, increased interconnectivity levels in production processes and advanced automation levels. Implementation of these technologies has already begun in many transport chain areas and will keep impacting upon all transport modes in the future.
As mentioned before, in the beginning of each part of the survey, a brief description of the automation levels (for each mode) has been included, presented in a simplified way with the use of illustrative examples, created specifically for this survey, while a link also gives the opportunity to the respondents to also check the official definition of the levels of automation. 3.2.2.1. Air transport From the total number of the respondents, 72,6% selected to answer the questions related to air transport automation. After the description of the air transport automation levels, level 3 (where the plane itself can react automatically to each situation and the pilot can intervene at all time) has been ranked as the most preferable by the users (47.63%), followed by level 2 (the reaction of the pilot is guided and assisted by the computer on board of the plane. The pilot has still all control), rated with 36.73%. Moreover, regarding the different levels of automation and operation of drones, level 2 has been stated as the more preferred, with a rate of 41% (where the drone is programmed to follow a certain route, it does not stay all the time in the visual line of sight of the pilot. The pilot can follow the drone on a monitor and, if needed can intervene to correct movements), followed by level 3 (31.8%), (where swarms of drones autonomously make decisions based on shared information. The use of drones for real-time data collection is becoming common practice in the areas of (indicatively) precision agriculture and civil defence, such as firefighting). Regarding the main concerns expressed by the respondents for the use of automated air vehicles, most of them (57.6%) pointed out cybersecurity (fear that it will be hijacked and used for a terrorist attack) as being very critical or critical, followed by safety fear dealing with the possibility of a technological failure that would lead to an accident (35.38%) and the employment related risks, mainly concerning loss of jobs (25.8%). Although the majority (69,5%) of the respondents did not provide any additional concerns (depicted by the yellow peak of the respective diagram), some of the those that have been mentioned include environmental sustainability, policy issues, training of operators and costs).
May 2020 28
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Not answered
5-Very critical
4
3
2
1- Not critical at all
0% 10% 20% 30% 40% 50% 60% 70%
Other Employment risk Cybersecurity Safety
Figure 10: Respondents’ concerns regarding air automation
Moreover, and taking under consideration the level of accidents as a criterion for the acceptance of automated air vehicles, the majority of the users (62.72%) stated that they would accept automated air vehicles if they had much fewer or near to zero accidents, in comparison to the conventional vehicles.
40%
30%
20%
10%
0% Have the same Have somehow Have much fewer Have close to zero Not answered level of accidents fewer accidents accidents (i.e. accidents as non automated than redustion of 50% ones today non'automated or more) than ones today non automated ones today Figure 11: Level of accidents as criterion for acceptance of automated air vehicles
The survey participants were also asked to declare in which cases they would agree with automated flying vehicles to be used. According to the replies received, the vast majority (83.7%) stated that they would accept drones to carry medicine, food, etc. during emergencies, while the following option (with an acceptance rate
May 2020 29
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance of 47%) would be the use of drones for deliveries outside the cities. On the other hand, the option with the lowest acceptance rate (18.5%) is the case of drones used for urban deliveries without a capacity limitation. Figure 12 below gives an overview of all relevant cases.
70%
60%
50%
40%
30%
20%
10%
0% Drones to Drones for Drones for Drones for Urban air Urban air Air transport Air transport carry urban urban deliveries vehicles for vehicles for vehicles for vehicles for medicine, deliveries deliveries outside passenger cargo cargo inter- passenger food, etc. with limited without a cities. transport transport city flights. inter-city during carrying capacity within cities. within cities. flights. emergencies capacity (i.e. limitation. up to 5 kg) 1- Not at all 2 3 4 5-Very much Not answered
Figure 12: Cases where the use of automated flying vehicles (including drones) would be accepted by the respondents.
Moving on to another important issue that deals with the implications to the employment of the air transport due to automation, most of the participants (36%) believe that automation will cause job losses in the air transport sector, however, a slightly lower (yet significant) rate (29%) states that automation will bring new jobs in the air transport sector. 23.3% believes that no significant changes will be caused to the employment of the sector due to automation.
40%
30%
20%
10%
0% Yes, automation Yes, automation No significant Other (please Not answered will cause job will bring new jobs changes will be specify) losses in the air in the air transport caused in the air transport sector sector transport sector
Figure 13: Impact of automation in the employment of the air transport sector.
As for the facilitation of persons with disabilities (PwD) by the automation of air vehicles, the opinions are divided. More specifically, 42.5% of the respondents stated that automated air vehicles are expected to facilitate the mobility of persons with disabilities, while another 45.5% stated the exact opposite. 9.5% of the
May 2020 30
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance respondents said that the facilitation will result under some specific conditions, while 2.5% did not answered this question.
50%
40%
30%
20%
10%
0% Yes No Under conditions Not Answered (please specify)
Figure 14: Facilitation of PwD mobility by the automation of air transport
In general, the overall opinion of the respondents regarding the automated air vehicles, ranges from neutral to positive, as nearly 47% of them have a good or a very good opinion for the automated air vehicles, while the opinion of 33.4% of the respondents is neutral (Figure 15).
40%
30%
20%
10%
0% 1- Not good 2 3 4 5-Very good Not at all answered
Figure 15: General opinion of respondents regarding automated air vehicles
3.2.2.2. Maritime transport From the total number of the respondents, 81.45% selected to answer the questions related to maritime transport automation. After the description of the vessels automation levels, the respondents rated (by 44.4%) level 2 as the most preferable (where the on-board systems are connected through a satellite uplink to a command centre, which will help the captain and his crew to make the right decisions), followed by level 3 (where the control of the ship is being realised by a remote control), which was rated preferable by 30.3%. The level that was ranked May 2020 31
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance last in the preferences of the users is level 4, in which the ship is fully automated to the point that, in many cases, there is no bridge at all. Regarding the concern of the respondents for the acceptance and use of automated vessels, again (as in the case of air transport) cybersecurity comes first, followed by safety issues, however their difference in maritime transport is very small (nearly 55% of the respondents characterised cybersecurity as critical or very critical issue for their acceptance, while the same did the 51.4% of them for safety). The risk of job losses is considered as of great concern for 24.8% of the participants. Moreover a 6% of the respondents stated as critical other issues, mainly related to environmental sustainability and the impact of extreme weather conditions.
75,59% Not answered
5-Very critical
4
3
2
1- Not critical at all
0% 10% 20% 30% 40% 50% 60% 70% 80%
Other Employment risk Cybersecurity Safety
Figure 16: Respondents’ concerns regarding maritime automation
Regarding the opinion of the respondents about the proper use of automated vessels, the 64.7% of them agreed that they should be used as rescue boats and for other special purposes, operating in harsh environments, 59% agreed to be used for transferring of cargoes in inland waterways and to other destinations near the coast line, while smaller acceptance rates were noted for the passenger services, as depicted in Figure 17.
May 2020 32
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
40% 35% 30% 25% 20% 15% 10% 5% 0% Rescue boats and Small barges Cargo ships of Passenger Passenger Passenger ships other special transferring any type services for short services for of all types purpose vessels cargoes in inland sea trips (i.e. for medium sea trips operating in waterways and less than 1 hour) (i.e. for less than harsh to other 24 hour) environments destinations near the coast line
1- Not at all 2 3 4 5-Very much Not answered
Figure 17: Cases where the use of automated vessels would be accepted by the respondents
In relation to the interaction of the automated vessels with other vessels (including non-automated ones) and passengers, the vast majority of the respondents stated that the automated vessels should be somehow indicated. 65.2% of them suggested to be marked with Variable Message Signs (VMS) and/ or at the ticketing counters, 19.3% of them to be announced at the ship’s arrival/ departure, while 3.7% have chosen the “other” option. Regarding the level of accidents that would act as facilitator for the acceptance of automated vessels, the results are similar to the ones for air transport, with the majority of the users selecting much less or zero accidents, while an almost similar share of respondents would also accept them if leading to somehow less accidents than non-automated ones have today (Figure 18).
40%
30%
20%
10%
0% have the same have somehow have much less have close to zero Not answered level of accidents less accidents accidents (i.e. accidents. as non-automated than non- reduction of 50% ones today. automated ones or more) than today. non-automated ones today.
Figure 18: Level of accidents as criterion for acceptance of automated vessels
Regarding employment risks for the maritime sector due to the introduction of the automated vessels, the majority of the respondents (43.3%) assume that automation will cause job losses in the maritime transport
May 2020 33
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance sector however, the second preferred answer (24.5%) was that no significant changes are expected in the maritime transport sector due to automation, while a slightly lower rate (22.2%) supported that automation will bring new jobs in the sector.
50%
40%
30%
20%
10%
0% Yes, automation Yes, automation, No significant Other (please Not answered will cause job will bring new changes will be specify) losses in the jobs in the caused in the maritime maritime maritime transport sector transport sector transport sector
Figure 19: Impact of automation in the employment of the maritime transport sector
Finally, the general opinion of the respondents about automated vessels is similar to the one presented for air transport, with 41.3% among them expressing a good or very good opinion, and 33.4% having a neutral opinion.
40%
30%
20%
10%
0% 1- Not good 2 3 4 5-Very good Not at all answered
Figure 20: General opinion of respondents regarding automated vessels
3.2.2.3. Rail transport From the total number of the respondents, 68.15% selected to answer the questions related to rail transport automation.
May 2020 34
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Concerning the levels of automation for the rail transport (including also light rail and metro systems) and the preference of the respondents, level 2 (where the locomotive is controlled remotely but the operator is still at the driver’s seat, overseeing the tracks and deciding whether or not to intervene. While at the station, he/she operates the doors and watches over the passengers’ safe disembarkation.) has been chosen by most of them (37%) as the most preferable for the rail vehicles operation, followed however by level 4 (where the train’s control is fully automated and its operation is being monitored remotely), with a preference rate of 27.8% of the users. The level that has ranked last in the preferences of the users is level 1, in which the train operator is in full control. As for the concerns of the users regarding rail transport automation, a similar pattern is noticed, as for the 2 previous modes. Notably, in this case the 2 biggest concerns, safety and cybersecurity, have also the same users’ rates in reporting them as crucial (47.6% and 46.7% respectively). Similar rates (as for the previous two modes) also appear for the rail transport regarding the employment risks, with a percentage of 26.8% of the users considering it as a critical or very critical issue towards their acceptance of automated rail vehicles. Other concerns that have been mentioned by a minor percentage of the respondents (6%), are mainly about environmental sustainability, but also the need of on-board staff to deal with incidents (such as drunks, hooligans, harassment and other emergencies), as well as overcrowding.
Not answered
5-Very critical
4
3
2
1- Not critical at all
0% 10% 20% 30% 40% 50% 60% 70% 80%
Other Employment risk Cybersecurity Safety
Figure 21: Respondents’ concerns regarding rail automation
In terms of employment risks of the rail sector due to automation, the rates are almost identical to the ones for the maritime sector, with the biggest share of the respondents expressing the opinion that automation will cause job losses in the rail sector.
May 2020 35
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
50%
40%
30%
20%
10%
0% Yes, automation Yes, automation, No significant Other (please Not answered will cause job losses will bring new jobs changes will be specify) in the rail transport in the rail transport caused in the rail sector sector transport sector
Figure 22: Impact of automation in the employment of the rail transport sector
The respondents have been also asked to rate their acceptance level for some specific cases regarding the operation of automated trains. As presented in Figure 23 below, the majority among them (64.3%) expressed their preference for a driverless passenger train, operating under the co-supervision of a human attendant in train, while the least preferred option is the driverless passenger train, controlled fully by an automated system. However, it needs to be stressed out that there are slight differences in the acceptance rates between the provided options.
40% 35% 30% 25% 20% 15% 10% 5% 0% A driverless A driverless A driverless A driverless A driverless An automated An automated passenger train, passenger train, passenger train, freight train, freight train, signalling signalling controlled fully under the co- under the co- controlled fully under the co- system for system for by an supervision of a supervision of a by an supervision of a trains that are trains that are automated human remote human automated human remote operating operating with system controller attendant in system controller. without human remote human train. supervision supervision
1- Not at all 2 3 4 5-Very much Not answered
Figure 23: Cases where the use of automated vessels would be accepted by the respondents
Also, in this case, the majority of the respondents (30.8%) stated that they would accept automated rail vehicles, as soon as they have close to zero accidents, in comparison to the conventional ones. Moreover, May 2020 36
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance the vast majority of them (78%) also stated that an automated train should be marked, either with Variable Message Signs (VMS) and/ or at the ticketing counters (60%) or through announcement at the train’s arrival/ departure (18%). Finally, it needs to be noted that a bigger (in comparison to the previous two modes) rate of the rail transport users (57.8%), expressed a positive general opinion regarding automated and self-driving rail vehicles.
30%
20%
10%
0% 1- Not good at 2 3 4 5-Very good Not answered all
Figure 24: General opinion of respondents regarding automated and self-driving trains
3.2.2.4. Road transport Among the total number of the respondents, 83% selected to answer the questions related to road transport automation. The road transport automation level that has been selected by the majority of the respondents (22.7%) as the most preferred one, is level 3 (where there is constant shared responsibility between the driver and the car and the driver can ask the car to fully drive by itself under specific scenaria (i.e. on the motorway) and the car informs the driver to take back control whenever necessary). However, there is a very slight difference between the preference rate of level 3 and the one for level 4, where the car drives by itself and the driver is still there to react only in case of an unforeseen scenario (19.7%) and level 2, where the car is responsible only for a limited number of manoeuvres or driving scenaria, such as self-parking, or drives at low speed in case of traffic-jam (16.4%). Since the road transport mode differs from the other modes in the way that a user can be involved in driving the vehicle, the Drive2theFuture voice of customers’ survey has examined the preferences and acceptance of the respondents, regarding their various different roles. In this context, the respondents were asked to answer under which conditions they would prefer automation support, in different occasions. More specifically, when using the road vehicles as drivers and/or passengers, most users (52.3%) stated that they would use automation in a different traffic environment (i.e. left or right driving), followed by unknown environments (i.e. foreign country, unknown city, rural area or countryside), preferred by the 48.9% of the users (see Figure 25 ). The majority of the respondents did not provide any additional aspects to this question (as depicted by the green peak in the respective diagram).
May 2020 37
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
80%
70%
60%
50%
40%
30%
20%
10%
0% 1- Not at all 2 3 4 5-Very much Not answered
Adverse weather (i.e. heavy rain, wind, heavy snow, etc.) In unknown environment (i.e. foreign country, unknown city, rural area or countryside) Different traffic environment (i.e. left or right driving) Unknown vehicle type (i.e. with or without automated gearbox, electric vehicle, etc.) Other (please specify):
Figure 25: Conditions under automation in road vehicles would be preferred by drivers/ passengers
The ranking of the options remains almost the same, again with slight range differences, when the question refers to the condition of the respondent being a driver/passenger of non-automated road vehicle, in mixed flow with Automated Vehicles (AVs) (Figure 26), or as a Vulnerable Road User (VRU) (i.e. pedestrian, cyclist, etc.) in mixed traffic with AVs and non AVs (Figure 27). The majority of the respondents did not provide any additional aspects to this question (as depicted by the green peak in the respective diagrams).
80% 70% 60% 50% 40% 30% 20% 10% 0% 1- Not at all 2 3 4 5-Very much Not answered
Adverse weather (i.e. heavy rain, wind, heavy snow, etc.) In unknown environment (i.e. foreign country, unknown city, rural area or countryside) Different traffic environment (i.e. left or right driving) Unknown vehicle type (i.e. with or without automated gearbox, electric vehicle, etc.) Other (please specify):
Figure 26: Conditions under automation in road vehicles would be preferred by drivers/ passengers of non- automated road vehicle, in mixed flow with Automated Vehicles (AVs)
May 2020 38
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
80% 70% 60% 50% 40% 30% 20% 10% 0% 1- Not at all 2 3 4 5-Very much Not answered
Adverse weather (i.e. heavy rain, wind, heavy snow, etc.) In unknown environment (i.e. foreign country, unknown city, rural area or countryside) Different traffic environment (i.e. left or right driving) Unknown vehicle type (i.e. with or without automated gearbox, electric vehicle, etc.) Other (please specify):
Figure 27: Conditions under automation in road vehicles would be preferred by Vulnerable Road Users in mixed traffic with AVs and non AVs
Regarding the option of marking automated road vehicles in traffic, the majority of the users (62.2%) agreed with this, choosing them to be marked either by a sign (43.9%) or a flashing siren (12.7%) on top of them or even by a flashing siren and sound on top of them (5.6%). 22.2% of the respondents stated that no marking is necessary. When examining the concerns of the users regarding automated road transport vehicles, some more types of concerns have been added, due again to the fact of the possibility of different types of uses of road vehicles (as passenger, driver, owner, interaction in the traffic environment as road user, VRU, etc.). As presented in Figure 28 below, the 61% of the respondents find cybersecurity to be either critical or very critical for them regarding their use of automated road vehicles, keeping the same pattern as in all other modes. Safety is the next most important concern of the users (with 60.2% stating it’s a critical or very critical issue), while concerns on data privacy, cost of AVs and security issues follow. The risk in employment is again ranked last, with the 31.8% of the respondents considering it to be (very) critical. Environmental sustainability, loss of driving pleasure and infrastructure costs are some more concerns, mentioned by fewer respondents.
May 2020 39
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Other (please specify)
Cost of AVs
Employment risk
Data Privacy
Cybersecurity
Safety
Security
0% 10% 20% 30% 40% 50% 60% 70% 80%
Not answered 5-Very critical 4 3 2 1- Not critical at all
Figure 28: Respondents’ concerns regarding road automation
Moving on to the expected impact of road transport automation in the sector’s labour force, again the picture resembles the ones of the other modes, with a rate of 42.4% expressing the opinion that automation will cause job losses in the road transport sector, 24.8% of the users believing that, on the contrary, automation will bring new jobs, while a percentage of 23.6% states that no significant changes will be caused in the employment status of the road transport sector. In the next sections of the road transport questionnaire, the respondents were provided with some statements, referring to both public transport and private cars, and were requested to express their level of (dis)agreement. Regarding the operation of public transport, (i.e. automated busses), most users (53.5%) agreed that a remote supervisor is required, with the possibility to talk to, hear passengers and react to their requests, while the 46.9% said that a driver should always be there, as a back-up to the system.
May 2020 40
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
30%
25%
20%
15%
10%
5%
0% No driver or A remote A remote A remote A remote A driver should human supervision supervisor is supervisor is supervisor is supervisor is always be there, as will be required in required but does required, with the required, with the required, also a back-up to the future automated not need to be possibility to talk possibility to also presented as a system. busses seen or felt by the to, hear be presented as an real-time video- passengers. passengers and avatar in a vehicle audio connection react to their screen. in a vehicle screen. requests.
1- I don’t agree at all 2 3 4 5- Fully agree Not answered
Figure 29: Agreement rates of users regarding the presence of driver in public transport means
On the other hand, the majority of the responders stated that they do not think that the fully automated public transport vehicles or private cars should have speed restrictions (42.1% and 42.4% respectively), while the 52.7% agreed that fully automated road vehicles should have the same speed limits as non-automated vehicles (Figure 30).
40%
35% Fully automated public transport vehicles (bus, tram, etc.) should 30% have speed restrictions (i.e. up to 20km/h). 25% Private cars should have speed 20% restrictions (i.e. up to 30km/h) only in towns.
15% Τίτλος Τίτλος άξονα Private cars should have speed 10% restrictions (i.e. up to 70km/h) in motorways also. 5%
0% Fully automated road vehicles 1- I don’t 2 3 4 5- Fully Not should have the same speed agree at agree answered limits as non-automated vehicles. all Τίτλος άξονα
Figure 30: Agreement rates of users regarding sped restrictions in automated road vehicles
When examining other acceptance criteria for automated road vehicles, it emerges that, likely to the other modes, most respondents (29%) stated that they would accept them if they had much less accidents (i.e. reduction of 50% or more) than non-automated ones today or close to zero accidents (33.4%).
May 2020 41
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
40% 35% 30% 25% 20% 15% 10% 5% 0% have the same level have somehow less have much less have close to zero Not answered of accidents as non- accidents than non- accidents (i.e. accidents. automated ones automated ones reduction of 50% or today. today. more) than non- automated ones today.
Figure 31: Level of accidents as criterion for acceptance of automated road vehicles Regarding the cost parameter though, the biggest percent of the respondents (48%) would accept automated vehicles, even if they had the same cost as the non-automated ones of today.
50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% have higher price have the same cost have somehow less have much less cost Not answered than the non- as the non- cost than the non- than the non- automated ones of automated ones of automated ones of automated ones of today. today today today
Figure 32: Cost of automated vehicles as acceptance criterion of the survey respondents
The survey participants were also asked about their opinion regarding future training needs of road users, in order to be able to use automated vehicles. The vast majority among them replied that road users would need to be somehow (40.2%) or significantly reskilled (35.5%), on issues concerning the understanding of the automation system, potential malfunctions and corrective actions, decision making and competence to intervene when necessary.
May 2020 42
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
50%
40%
30%
20%
10%
0% Yes, somehow Yes, significantly No Not answered reskilled (for which reskilled (for which skills related to skills related to automated road automated road vehicles, would you vehicles, would you like to receive training like to receive training for?) for?)
Figure 33: Need for reskilling of the road users to be able to use automated vehicles
Moreover, the majority of the respondents (62.3%) assume that automated road vehicles would facilitate the mobility of persons with disabilities, while 7.3% agreed that this would be the case under some specific conditions; provided that, for example, these vehicles will have suitable access, ingress/egress and the remote support that PwD require, or upon an assessment of the users, to ensure they are capable of "taking back control" when the vehicle asks it. In general, the 48.3% of the respondents, has a good or a very good opinion about the automated and self- driving road vehicles, while the 23.3% of them expressed a neutral opinion.
30%
20%
10%
0% 1- Not good at 2 3 4 5-Very good Not answered all
Figure 34: General opinion of respondents regarding automated and self-driving road vehicles
May 2020 43
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance 3.3. Comparative remarks across transport modes
From the above analysis of the replies on the different transport modes questionnaires and the expressed acceptance of their respective automated vehicles by the respondents, we can assume that there are no remarkable differences between modes. Reviewing the overall opinion of the respondents about the automated vehicles of the different modes, we may note slight differences among them, with the respondents being mostly neutral in the case of air and maritime transport, while expressing a more positive opinion for the rail and road transport.
40%
30%
20%
10%
0% 1- Not good at 2 3 4 5-Very good Not answered all
Air transport Maritime transport Rail tranport Road transport
Figure 35: General opinion of respondents regarding automated and self-driving vehicles of all modes
However, it needs to be noted here that the preference of the majority of the respondents regarding the automation levels of all modes, focused on the middle levels, where cooperation between the vehicle and the driver still exists. More specifically, in the air transport, level 3 is the one that has gathered then biggest share among the respondents, where the plane itself can react automatically to each situation but the pilot can intervene at all time, while the same also applies for drones, where level 2 has been selected by the respondents’ majority, in which the operator can follow the drone on a monitor and, if needed can intervene to correct movements. As for the maritime and rail transport, level 2 is the most preferred, where for the vessels the on-board systems are connected through a satellite uplink to a command centre, helping the captain and his crew to make the right decisions and for the trains the operator is still at the driver’s seat, overseeing the tracks and deciding whether or not to intervene. While at the station, he/she operates the doors and watches over the passengers’ safe disembarkation. Finally, for the road transport, the level where there is constant shared responsibility between the driver and the vehicle (level 3), emerged again as the most desirable among the users. However, in the latter case the rating of higher automation (level 4) is quite close. Moving on to another point, the results show that also users’ concerns and fears are almost identical for all modes, focusing mainly on cybersecurity and safety, while rating much lower the risks to employment. This fact is also inter-connected to the fact that in all modes, the majority of the respondents stated that they would accept automated vehicles, as soon as they have close to zero accidents (average rate of 31.3%) or at least much less accidents (i.e. reduction of 50% or more), in comparison to the conventional ones (average rate of 29%).
May 2020 44
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Another concern that has been also emphasised throughout the modes is the one of environmental sustainability. In relation to the interaction of the automated vehicles with other users (either road users and/or non- automated vehicles), again the vast majority of the respondents states that they would prefer automated vehicles to be somehow marked (e.g. either by the use of Variable Message Signs (VMS) or through announcement at the vehicle’s arrival/ departure), in order to be clearly distinguished. Finally, regarding the impact on employment and employability, the majority of the respondents (in all modes) believe that automation will cause job losses to the transport sector. However, at the same time there is a quite big rate of respondents also expressing the opinion that automation will also bring new jobs, concluding to a rather balanced situation. More specifically, in road transport the second biggest percent of respondents (24.8%) believe that automation will bring new jobs, while a percentage of 23.6% states that no significant changes will be caused in the employment status of the sector, while the opposite pattern appear in the other three modes, where the second biggest percentage believes that there will be no major changes, while a slightly lower rate of users states that automation will bring new jobs in the sector.
May 2020 45
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance 4. Research hypotheses
One of the main purposes of the Drive2theFuture project is to explore the factors affecting users’ acceptance of automated vehicles, envisaging a concise approach towards enhancing user acceptance for the upcoming deployment of automated vehicles. The holistic approach of the Drive2theFuture approach is based upon the fact that: • it considers all clusters of users (“drivers”, passengers, operators, stakeholders), their needs, preferences and specificities; • it addresses all modes of transport, along with transferability issues between and across them (including MaaS strategies); • it involves multiple tools to achieve its goals (behavioural models, simulators, optimized HMI per automation level, training schemes, active involvement of users, demonstrations, ethical and regulatory aspects, policy and incentives).
Consequently, the project puts the user at the centre of automation deployment, while employing technological and regulatory means to facilitate this deployment; thus aiming to maximise acceptance, satisfaction, willingness to use along with safety, security, compliance and sustainability. In this process (which includes all transport modes and levels of automation but is differentiated for each of them), the study of user opinion and behaviour towards existing solutions will guide the iterative development of novel HMI concepts and training schemes, while their acceptance will be (objectively and subjectively) measured in a series of 12 Pilots across Europe. In this framework, Drive2theFuture indicatively considers specific research priorities (per mode), which are being further specified and addressed within the UCs of the project and the realisation of each pilots (see Section 7 of this Deliverable). Based on a preliminary survey, open research issues and hypotheses per transport mode and AV function/level have been initially recognised, through a thorough literature review of 23 sources that includes input from several different types of sources, such as roadmaps of ETPs and partnerships on all modes (ERTRAC, WATERBORNE, SESAR, S2R , etc.), policy papers by stakeholders associations (e.g. ITF/OECD report, STRIA roadmap), European Commission’s Communications, research projects outputs and minutes from research working groups, as well as scientific papers. The table below lists the sources that have been analysed and taken under consideration.
Table 5: Literature sources for the analysis of Drive2theFuture research hypotheses and priorities A/A Title of source
1 European Commission: Special Eurobarometer 496 – Report on expectation and concerns of connected & automated driving (2019) - https://data.europa.eu/euodp/en/data/dataset/S2231_92_1_496_ENG 2 ERTRAC’s Automated Driving Roadmap (2019) - https://www.ertrac.org/uploads/documentsearch/id57/ERTRAC-CAD-Roadmap-2019.pdf 3 STRIA Roadmap on Connected and Automated transport (2019) - https://ec.europa.eu/research/transport/pdf/stria/stria- roadmap_on_connected_and_automated_transport2019-TRIMIS_website.pdf 4 Autonomous Vehicles Survey Report of Perkins Coie LLP and the Association for Unmanned Vehicle Systems International (AUVSI) (2019) - https://www.perkinscoie.com/images/content/2/1/v3/216738/2019-Autonomous-Vehicles- Survey-Report-v.3.pdf 5 Forster, Y., Hergeth, S., Naujoks, F., Krems, J., & Keinath, A. (2019). Tell Them How They Did: Feedback on Operator Performance Helps Calibrate Perceived Ease of Use in Automated Driving. Multimodal Technologies and Interaction, 3(2), 29.
May 2020 46
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
A/A Title of source
6 European Commission - On the road to automated mobility: An EU strategy for mobility of the future (2018) - https://ec.europa.eu/transport/sites/transport/files/3rd-mobility- pack/com20180283_en.pdf 7 ERRAC - Rail 2050 Vision Rail - The Backbone of Europe’s Mobility (2018) - https://errac.org/wp-content/uploads/2019/03/122017_ERRAC-RAIL-2050.pdf 8 European ATM Master Plan: Roadmap for the safe integration of drones into all classes of airspace (2018) - https://www.sesarju.eu/node/2993 9 Shift2Rail Multi-Annual Action Plan (2018) - https://shift2rail.org/publications/multi-annual- action-plan/ 10 Pakusch, C., Stevens, G., Bossauer, P., & Weber, T. (2018). The Users' Perspective on Autonomous Driving-A Comparative Analysis of Partworth Utilities. 11 Germany Federal Ministry of Transport & Digital infrastructure: Task Force on Ethical Aspects of Connected and Automated Driving (Ethics Task Force) (2018) - https://www.bmvi.de/SharedDocs/EN/publications/report-ethics-task-force-automated- driving.pdf?__blob=publicationFile 12 ITF/OECD, Managing the transition to driverless road freight transport (2017) - https://www.itf-oecd.org/sites/default/files/docs/managing-transition-driverless-road- freight-transport.pdf 13 DIMECC – One Sea: Roadmap towards commercial autonomous shipping in 2025 (2017) - https://www.oneseaecosystem.net/wp- content/uploads/sites/2/2017/08/onesea_roadmaps-august-2017_paivi-haikkola_rev.pdf 14 Draft road map and action plan to facilitate automated driving on TEN road network (2017) - https://www.its-platform.eu/highlights/draft-road-map-and-action-plan-facilitate- automated-driving-ten-road-network-report 15 EU – US – Japan ITS Cooperation: Trilateral Impact Assessment Sub-Group for ART (20017) - https://connectedautomateddriving.eu/mediaroom/impact-assessment-framework- automatisation/ 16 WATERBORNE Vision 2030 & Innovation Opportunities (2016) - https://www.waterborne.eu/media/35622/folder-august-2016.pdf 17 EC-funded research contribution to roadmapping of urban transport - SETRIS Deliverable Report (2016) - https://www.waterborne.eu/media/35594/setris-d14_vf-11-03-2016-final- v2.pdf 18 Rolls Royce Marine, Autonomous ships - The next step (2016) - https://www.rolls- royce.com/~/media/Files/R/Rolls-Royce/documents/customers/marine/ship-intel/rr-ship- intel-aawa-8pg.pdf 19 GEAR 2030 roadmap on highly automated vehicles (2016) - https://circabc.europa.eu/sd/a/40c4104e-5f7d-4cc4-a418-679e72c5fc7e/Road-20160615- WG-pres-EU%20activities%20on%20automated%20vehicles.pdf 20 European Commission, Autonomous Systems, Special Eurobarometer 427 / Wave EB82.4 – TNS Opinion & Social (2015) - https://ec.europa.eu/commfrontoffice/publicopinion/archives/ebs/ebs_427_en.pdf 21 Shift2Rail Multi-Annual Action Plan (2015) - https://www.shift2rail.org/wp- content/uploads/2013/07/MAAP-final_final.pdf 22 Gasser, T. M., Schmidt, E. A., Bengler, K., Chiellino, U., Diederichs, F., Eckstein, L., ... & Hoyer, R. (2015). Report on the Need for Research. Round Table on Automated Driving, Federal Ministry of Transport and Digital Infrastructure, BMVI2015. 23 EY - Deploying autonomous vehicles - Commercial considerations and urban mobility scenarios (2014) - https://www.ey.com/Publication/vwLUAssets/EY-Deploying-autonomous-vehicles- 30May14/%24File/EY-Deploying-autonomous-vehicles-30May14.pdf
May 2020 47
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
A template has been developed from CERTH/HIT that has been used for the analysis of these sources (ANNEX 3). Moreover, feedback has been provided through the other WP1 Activities, while a relevant discussion has taken place during a dedicated session on the project’s Use Cases that took place at the 6th of March 2020 in Brussels. In Table 6 below the final research hypotheses and priorities of the Drive2theFuture project are presented, which have been slightly updated since their preliminary version (included in the project’s DoA).
May 2020 48
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Table 6: Drive2theFutrure research hypotheses and priorities A. Road B. Rail C. Maritime D. Air • HRO1: Acceptance after hands-on experience of all levels of • HRA1: Train-centric • HMA1: Acceptance of • HAV1: Simulated behaviour automation in urban, rural, highway and specific applications, concepts for automatic passengers, pilots and training in non-standard such as tunnels, constructions and bridges, and environmental operation. operators. situations (cyber-attack, mass conditions (i.e. co-pilot for adverse weather, unknown • HRA2: Development and • HMA2: Impact on events in urban settings). environments, unknown type of vehicle, etc.). examination of HMI for operators through • HAV2: Impact of adaptive HMI • HRO2: Acceptance considering age, gender, IT literacy, GoA3/4 operation spectrum of automation on drone flight planning and socioeconomic factors and understanding of automation for all (signaller/train operator levels and quantitative execution. cohorts by Kansei/Citarasa methodologies. perspective). prognosis of behavioural • HAV3: Public acceptance of • HRO3: Public acceptance of the possibility of accidents • HRA3: Impact on training adaptations. drones’ violation of privacy. occurrence with automated road vehicles, even if they are and education, ensuring • HMA3: Deskilling issues • HAV4: Drone purpose of use fewer than the number of the conventional ones. safety culture in automated and decreased system correlation to its appearance. • HRO4: Acceptance of other vehicles’ drivers, passengers and operations supervision. understanding. • HAV5: Risk of drone accidents. VRUs. • HRA4: Passenger and • HMA4: Perceived • HAV6: Drone’s noisiness • HRO5: Conspicuity of automated vehicles and the mode they freight information situation awareness vs. acceptance. operate at (automated or not). systems for the future actual system status. • HAV7: Vigilance and • HRO6: Vigilance and complacency issues in Level 3 and Level 4. automated railway system. • HMA5: Vigilance and complacency issues for the • HRO7: Driver-Readiness in transitions between manual and • HRA5: Full automated complacency issues in drone operator and the automated driving. railway ecosystem and transition from operator supervising controllers. • HRO8: Manned and unmanned traffic safe operation and in connected business to systems monitor. • HAV8: Impact on training and harmony, especially in rural environments. models’ acceptance. • HMA6: Cost efficiency of education, ensuring safety • HRO9: Transfer of expertise from rail, water, air sectors. • HRA6: Vigilance and automated vs non- culture in automated • HRO10: Behaviour adaptation (“mimicking”, “flocking”) of non- complacency issues in automated operation in operations supervision. equipped vehicles. transition from operator to a wide range of missions. • HAV9: Liability and operational • HRO11: Reliability, availability, and flexibility of an systems monitor. • HMA7: Issues of issues. autonomous car, due to the inexperience of the users with the • HRA7: Issues of European international technology, and lack of confidence in this new technology. harmonisation. harmonisation.
May 2020 49
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
A. Road B. Rail C. Maritime D. Air • HRO12: Impact of mixed and automated flows to traffic flow • HAV10: Cost efficiency of (micro/macro) simulation, incl. big data analytics for scaling. drone-based logistics • HRO13: Negative Impacts of unintended shifts in mode choice operations. (i.e. especially attractive for car owners, neglecting the fact that • HAV11: Issues of international they could also unintendedly impact the mobility behaviour of harmonisation. other, currently non-car, users). • HRO14: Training and dissemination with multi-platform tools for VR/AR simulation, WoZ and simulator scenarios for public acceptance and expectations. • HRO15: Liability and operational issues per automation level and user cluster. • HRO16: Cost efficiency of automated vs non-automated vehicles. • HRO17: Dependence of AV acceptance on HMI. • HRO18 Issues of international harmonisation.
More specifically, 7 new research hypotheses have been identified for the road sector, 1 for the rail and maritime sectors and 2 for the air sector, resulting to an overall number of 43 research hypotheses and priorities that have also been connected to the project’s Use Cases and are being further analysed and mapped for each Pilot Site in Section 7 of this Deliverable, indicating the focus of each pilot/ pilot site, as well as its expected outcomes.
May 2020 50
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
5. Transferability from/to other modes
According to European Commission strategic agenda (European Union, 2011), the transport of the future will be automated and integrated. In particular, for reaching the strategic goal of an integrated transport, great value would come from the exchange and transferability of solutions across the various transport modalities, to share best practices and lessons learnt. In the framework of Drive2theFuture project, as “solution” is defined any innovation that has the purpose to solve a problem in a specific domain, being it a past innovation already recognized, a present innovation or an innovation still under development (e.g. prototype, beta versions, etc.). To foster the transferability of solutions proposed within Drive2theFuture across transport modes, a methodology has been developed to guide the analysis of common issues, approaches and lesson learnt of solutions across transportation sectors. The methodology is based on a “Database of Solutions” and consists of: collection and selection of a list of relevant solutions, their description and categorisation in the database and guidelines to assess the potential of transferability across modalities. The expected benefits of this method are the following: • to build a coherent strategy for cross-fertilisation initiatives, avoiding “silver-bullet” approaches, by working on different solutions and different levels • to build an open database of solutions that can be expanded over time and where experts of all transport domains can draw and contribute according to their necessities; • to foster the cross-fertilisation of solutions by providing guidelines using two main criteria: benefits/issues expected and applicability.
Such methodology was initially developed by DBL for EU FP7 funded project EXCROSS (2011-2014), where its aim was to foster the transfer of resilience resources across transport domains, with a specific focus on safety solutions. In Drive2theFuture, the methodology has been extensively adapted to meet the aims of the project. In particular, during the process of adaptation, two factors had been kept in mind: 1) the different focus; 2) the different timeframe and resources. Concerning the first factor, the focus of the solutions has been extended compared to the original methodology. While EXCROSS project had a deep focus on safety, explored through seven safety-related topics (i.e. certification, dangerous goods, fatigue, training, incident investigation, safety enhancement, safety facilitation), the new methodology has to deal with a wider variety of topics related to the acceptance of the future automated and integrated transport. More details about the topics selected will be described in the next section (see Table 8). For what concerns the time and resources, the original methodology was developed during the entire duration of the EXCROSS project and implied the involvement of a large number of domain experts, safety experts, end users and stakeholders. Therefore, in Drive2theFuture, the methodology application had to be downsized in order to reach the efficiency and effectiveness needed to fit the time allocation and resources of the activity. The activity followed five (5) different steps: • Step 1: Planning for the data collection. During the first step, the methodology was adapted to the Drive2theFuture framework in terms of definition of objectives and focus and the data collection plan was established and scheduled.
May 2020 51
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
• Step 2: Interviews. In the second step, a series of interviews were carried out, involving experts of the various domains. In this step, DBL interviewed representatives from the aviation and road sectors, whilst EURNEX and WEGEMT carried out the interviews with rail and maritime experts, respectively. • Step 3: Survey. Parallelly to step 2, a survey based on the interview structure was prepared and launched with the aim of collecting more information to be integrated with the interviews. In this step, DBL was supported by CERTH, EURNEX e WEGEMT and the rest of the consortium for the dissemination of the survey among their network targeting professionals as well as non- professionals. • Step 4: Workshop. Once all the solutions were collected, analysed and integrated in the “database of solutions”, a workshop was held during the project 2nd Plenary Meeting in Karlsruhe (28th and 29th of November 2019) with the objective of collecting inputs from consortium experts to refine and finalize the methodology. The inputs were analysed and the methodology completed. • Step 5: External discussion and dissemination. The fifth step regards the external discussion about the transferability of solutions and the dissemination of the activity. Step 5 will develop through the entire project, with periodically updates, until month 36.
5.1. Planning for the data collection During the first step of the activity, DBL laid down the basis for the data collection of solutions that constitute the inputs for the methodology, following two sub-steps: 1. Identification of the relevant topics 2. Development of the format for the collection of solutions
5.1.1. Identification of the relevant topics As previously introduced, compared to the original methodology in EXCROSS and the description of the activity in the proposal phase, the focus of the methodology has been extended from the seven safety- related topics to a broader range that tries to cover all relevant topics concerning future automated and integrated transport. For doing so, a literature review was done with the aim of understanding which are the most relevant trends identified by the European Union for the transport of the future (European Union, 2011; Mobility4EU, 2016, 2018a, 2018b). As showed in Table 7, a selection of the most relevant values (Mobility4EU, 2018a), trends (Mobility4EU, 2016) and action areas (Mobility4EU, 2018b) was made. Furthermore, macro-categories of goals from the European Commission’s strategy on transport (European Union, 2011) were identified.
Table 7: Selection of the most relevant goals, values, trends and action plans for the future transport Mobility4EU White Paper for Vision for Transport Societal trends Action Plan Transport 2030 Environmental Sustainable Low-zero emission Sustainability protection mobility Novel business models Seamless Cross-modal/cross- Multimodal/seamless and innovation in border transport and transport for people and transport integration of novel freight mobility services in public transport / / Automation and Standardisation across- connected driving modes (e.g.
May 2020 52
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Mobility4EU White Paper for Vision for Transport Societal trends Action Plan Transport 2030 certifications, intermodal tickets, etc) / / Automation and Standardisation across- connected driving modes (e.g. certifications, intermodal tickets, etc) Inclusive society, Inclusive and user- Inclusion: putting the Passenger rights personalisation, centric user in the centre accessibility Safety and Security in / Safety and Security and safety transport (cyber)security in transport Urbanisation and Smart / Mobility planning / cities
Drawing from the literature review, a first selection of relevant topics was made, including: 1) sustainability (green in Table 7), 2) multimodal/seamless transport (light blue), 3) inclusion and user-centeredness (orange), 4) security and safety (yellow) and 5) smart mobility for smart cities (violet). This first selection of topics was then discussed through a loop of brainstorming sessions and feedback from the consortium experts, leading to the dismissal of some topics and the addition of new ones. The final selection of the topics is shown in Table 8. Those topics were used as the base for the data collection of solutions.
Table 8: Final selection of relevant topics for Drive2theFuture Topic Description Future automated and integrated transport holds the expectation to be much safer than present transport by reducing first line human error that accounts for 90% of road accidents. Nevertheless, in other domains as rail, maritime and aviation, some Safety level of automation has already been introduced, with a positive impact on safety. This topic focuses on analysing those solutions that deal with the improvement of safety and reduction of risk for users, passengers and stakeholders in any domain. As stated in the European Commission's White Paper on Transport, "transport security is high on the EU's agenda". Concerning all modes of transport, security covers the prevention, mitigation and system resilience against a range of intentionally malicious activities. This topic focuses on analysing solutions related to Security prevention and mitigation of the effects of terrorist attacks, vandalism and sabotage, piracy and any crime committed in the premises of transport operators or impacting on passengers. Of growing importance is the topic of cyber-attacks and data privacy in automated and connected transports. In this context, the Human Factors topic intends to highlight the physical, cognitive and emotional interactions among users/stakeholders and the environment (their vehicle, other vehicles, the infrastructure). These include all kinds of factors involved in the user's performance while "using" the vehicle (e.g. handovers of control, Human Factors workload, fatigue, etc.), the physical fit between users and means, and also the overall positive or negative experience they have while interacting with the environment. Therefore, the solutions can address aspects related to physical ergonomics, Human-Machine Interaction (HMI) and User Experience (UX).
May 2020 53
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Topic Description Within this topic, inclusion refers to accessible and usable transport for all users. The mainstreaming of universal design principles to vehicles and infrastructures facilitates the physical accessibility of transport for people with reduced mobility (disabled or older people, but also with pushchairs or heavy luggage). Furthermore, Inclusion and User- transport equity and spatial accessibility of transport must be supported in order to centeredness enable access to essential services, employment, shopping opportunities and healthcare to everyone. Hence the geographical spread of public transport stations and other mobility services as well as the affordability of transport offers are considered. Sustainability is considered one of the most important topics in present and future mobility. The aim of EU Commission is to reduce at least 60% of greenhouse gasses by 2050 for the transport sector. This area focuses on the topic of sustainability from the "users" (drivers, passengers, etc.) and relevant stakeholders' (VRUs, etc.) point Sustainability of view. This topic considers all kinds of solutions for reducing individual or collective footprint of mobility on the environment and their impact on users/stakeholders. Potential solutions could deal with, but are not limited to, changing travel habits (e.g. cycling, car sharing, public transport, rail, etc.), traffic reduction, acceptance and use of more environmentally friendly vehicles. Trust is a major predictive factor of technology acceptance. People will not accept, and therefore use, future transport technologies if they don't trust them. In this topic will be considered all those solutions that aim to facilitate and increase Trust users/stakeholders' trust (or perception of trust) towards present or future transport technologies. These solutions may affect users' trust directly (e.g. through physical or HMI's features) or indirectly (e.g. through brand reputation or communication). Transport is a rapidly developing and changing sector which faces problems to develop, attract and retain appropriate staff. Future jobs will therefore require new and advanced skills in engineering as well as in back-office operations, but at the same time, the growing interdisciplinary elements of transport activities will also Training require transport professionals with developed skills in safety, security, logistics, IT, behavioural sciences, marketing and economics. In this topic, will be therefore considered solutions that deal with improvement or adaptation of workers/users training in any transport domain. These can be related to training techniques and approaches, tools, certification and accreditation systems and others. This topic refers to the necessary innovation strategies including the appropriate Innovation policy development, governance, financing instrument, dissemination, etc. in order to ensure a rapid deployment of results developed in the research process.
5.1.2. Development of the format for interviews and survey For collecting information about innovative solutions in the various transport modalities, two data collection techniques were used: interviews and an online survey. Both the interviews and the survey were built using the same format in order to investigate the main aspects related to the solutions. The aspects investigated are described in Table 9.
Table 9: Format for the data collection Topic Description Interviewees and respondents were asked to provide personal information such as Personal their role and experience, their background, their organisation/institution and their information topic of expertise. Other sensitive information like names, gender and age were May 2020 54
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Topic Description considered not relevant for the activity’s purposes and they were not collected. Personal information is useful to contextualize the further information collected through interviews and survey. Every expert selected a solution based on their reference domain and field of Description of the expertise (based on the topics presented in table 12). Once the solution had been solution selected, it was described in terms of its objectives, functioning, users’ needs addressed, target, readiness and costs/availability. The rationale behind the solution. Is it Approach technology/human/organizational/procedural (or other-) based? Explanation of why such approach was used to solve a certain problem. Focus regarding potential issues that the solution might have (or will) encounter during its development and/or implementation. Issues described could be expected Issues (well-known and anticipated) or unexpected. Moreover, it was asked if and how (some of) the issues have been solved. What has been learnt from the development and implementation of the solution. Lessons learnt This is the takeaway message for future solutions and for potential transfers to other modalities.
Both the interviews and the survey were developed using these aspects as main structure, with minor differences due to the different data collection techniques and tools. Interviews were conducted online, following a semi-structured script. The duration of each interview was approximatively of 30 minutes and, previous agreement with the interviewees, they were recorded to support later analyses. The survey was developed using the online platform Surveymonkey and consisted in a mix of closed (multiple choices) and open questions, to allow the experts to freely describe their solutions. The two data collection techniques were conducted in parallel during month 4 and 5.
5.2. Interviews Interview were conducted with the aim of obtaining in depth description of solutions from selected experts in the various domains and fields, both inside and outside the consortium. Experts were selected and contacted through various means, in details:
• External experts for aviation, rail and maritime were selected and contacted directly by DBL, EURNEX and WEGEMT respectively, through their contacts’ list. A list of road experts was provided by CERTH and a selection of experts was then contacted by DBL; • Internal experts were selected through an open request for participation in the activity to all project’s partners.
In total, 17 interviews were conducted, covering 20 solutions in all four transport domains and some multi domain. All interviews with respective domain, topic and solution are listed in Table 10.
Table 10: Interviews Internal Expert Domain / Topic Solution (company/institution) external Aviation Airbus External Innovation Airbus BizLab Start-up Accellerator May 2020 55
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Internal Expert Domain / Topic Solution (company/institution) external Delft University of Aviation External Safety Certification process in aviation Technology (TU Delft) Deutsches Zentrum für Single European Sky ATM Research Aviation Luft- und Raumfahrt External Innovation program (SESAR) (DLR) Model-based technique for risk Aviation Airbus External Safety assessment Regulatory framework for safety Human Road IAM RoadSmart Internal systems’ HMIs across vehicles and Factors manufacturers CITS (Communication between Intelligent Transport Systems) use case: Road Wiener Linien Internal Safety I2V crossroads traffic monitoring for autonomous buses Road HIT/CERTH Internal Safety Warning system for VRUs based on IoT Austrian Institute of Road Internal Safety Risk Assessment based on dynamic data Technology GmbH (AIT) Embedded computer safety system for Road Easymile External Safety EZ10 Maritim American Bureau of External Safety Alarm Management System in ships e Shipping Maritim Bureau Veritas External Safety Periodically Unmanned Bridge e Maritim Maran Dry Real time vessel and machinery External Safety e Management performance monitoring systems Consultant Railway European Railway Traffic Management Rail External Safety Signalling* System (ERMTS) Innovation Shift2Rail Creation of one-stop shop for safety Safety certificates of vehicles for multiple Rail EURNEX Internal European state Publication of Open Data by rail Innovation infrastructure managers and rail operators The in-accessibility index: advantages Inclusion Rail CambiaMO External and potential for improving strategic and UCD planning and investment ** Inclusion Ludwig-Maximilians- Participatory design for mobility ** and UCD Rail Universität München External Inclusion Integration of transport services in a (LMU) and UCD single smartphone App ** Smart Public Transport Sustainabilit Rail External Bike adaption to public transport ** Lab y * The interviewee requested to keep the project mentioned confidential ** The solutions were than categorised as “multimodal”.
May 2020 56
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance 5.3. Survey The online survey was developed and disseminated with the aim of reaching a larger number of experts and to collect a higher number of solutions to be integrated with the ones collected through interviews. For doing so, DBL developed the survey following the same format of the interviews, using a combination of closed and open questions. Once prepared, the survey was first disseminated internally within the consortium and then it was asked to every partner to share it with their direct contacts and through their social media channels (LinkedIn and Twitter). The survey was also shared through Drive2theFuture’s website and social channels, as well as in those of other related EU-funded projects such as Mobility4EU and TRA VISIONS. Finally, a thread in the ETM Forum (https://www.etmforum.eu/ ) was also opened to promote the survey. The survey obtained 41 valid responses from a total of 90, resulting in a selection of 18 solutions. The respondents were composed by 32% of internal experts and 68% of external experts. Solutions’ distribution across topics can be seen and Figure 36.
Figure 36: Distribution of solutions across topics (survey) 5.4. Database of solutions Once the data collection was finished and all the solution were selected, they were inserted in the “database of solutions” (Figure 37). The database of solutions is a database where all solutions have been stored and described using different categories. The categories are, in order:
• Transport modality: the transport modality from which the solutions were taken. The transport modalities are road, rail, maritime and aviation plus a multimodality category that comprehend those solutions that already consider more than one transport mode. • Topics: the topic that the solution covers (see Table 12). • Sub-topics: inside the main topic, a further specification of the type of solution. Every topic has its own sub-topics (e.g. Safety can be further split in safety enhancement, automation, standards, certification, safety assessment and safety culture). The sub-topics identification was deduced once all the solutions had been collected. • Data source: where the information about the solution were collected (i.e. interview or survey). • Name of the solution
May 2020 57
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
• Description: description of the solution in terms of needs addressed, objectives and functioning. All contextual information is also described in this section. • Target: the categories of people that are impacted by the solution, directly or indirectly. • Type of vehicle: to which types of vehicles is applied. • Type of solution: for example, if the solution is (mainly) a technology, methodology, physical artefact, etc. • Issues: description of the issues, expected and/or unexpected, that have been encountered during all the development stages and/or during and after the implementation. For those solutions that are still under development, there are listed potential issues that are expected. • Issues addressed: explanation of how (some of) the issues have been addressed or are intended to address. • Lessons learnt: what have been learnt from the development and implementation of the solution. This is the takeaway message for future solutions and for potential transfers to other modalities. • References: when available, bibliographic references were added.
The database of solutions constitutes an open resource that can be used as a database for searching, studying and comparing a variety of innovative solutions from all transport modalities, covering different topics. All the categories of the database have a twofold purpose: in addition to facilitating the description and the comparison among solutions, they serve also as filter to search the solutions desired according to the preferred characteristics. At the current stage, the database lists 37 solutions. The database will remain open to be further expanded during the entire duration of the project (and beyond).
May 2020 58
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Type of Type of Mode Topics Sub-topics Data source Solution Description Target Issues Issues addressed Lessons learnt References vehicle solution
Are there any issues What did we learn from the To which type Technology, related to the implementation of such Please provide a short description of the Who is the beneficiary of vehicle is the methodology, Have (any of) these issues Name of the solution implementation of this solution? solution. of the solution? solution physical artifact, been addressed? How? solution? If so, please What is the takeaway for referred? etc describe them. future solutions? EZ10 is a fully automated (SAE level 5) bus Passengers Buses/Trucks Technology 1) In autonomous driving, 1) One solution could be To be compliant with safety EZ10: Embedded Safety for public transport developed by manufacturers have to using LIDAR sensors that standards required by https://easymile.com/so Easymile. ensure that any function have been certified authorities, it takes huge lutions-easymile/ez10- Road Safety Automation Interview System for a fully For safety, Easymile is developing an reaches appropriate according to industry amount of workload and it is autonomous-shuttle- automated bus (EZ10) embedded computer system, that will be safety levels. However, it standards, but not road very challenging and time easymile/ certified according to ISO 26262: Road is complicated to reach standards. This is does not consuming, However, from vehiclesThe solution – Functional is a warning safety, system with based ASIL D on Vulnerable Road Users Cars + Technology thoseUnexpected level because technical some represent1) To address the this best issue, option, in ourThe perspectivemain lesson the is that, whole Obstacle detection: : IoT (Internet of Things) that warns VRUs Buses/Trucks + Issues: future developments, it especially in the future Safety Warning system for (Vulnerable Road Users) in case of critical VRUs 1) Delay in the should change the context of autonomous Road Safety Interview Enhancement VRUs based on IoT events (e.g. when a VRU is about to cross communication. The communication protocols driving, more attention the road and there is a vehicle system uses a web server (i.e. 5G). With new should be dedicated to VRUs approaching). for the communications. protocols, the web server and their interactions with TheWiener VRUs Linien (pedestrians is developing and cyclists) an automated are Drivers/Riders Buses/Trucks Technology This1) Technological increases the time for will1) Technological be no longer needed AVs.Autonomous driving, and in Crossroads traffic crossroad traffic monitoring system to limitations: The system is limitations: They think particular this I2V crossroad Safety Road Safety Interview monitoring for improve buses safety by reducing traffic not working satisfactorily. that they can improve monitoring solutions requires Enhancement accidents. Cameras are able to cameras’ performances by huge amount of computing autonomous buses Such system employs a set of cameras to detect pedestrian very refining the algorithm. power to have a good quality monitor crossroads. The cameras are able well, they detect cyclists They are also thinking picture recognition. It is a feature that alerts a driver to an Drivers/Riders Cars + Technology 1) Increase the price of imminent crash and helps them use the Buses/Trucks vehicles Safety Auto Emergency maximum braking capacity of the car. AEB 2) Measure poorly Road Safety Survey Enhancement Braking (AEB) will independently brake if the situation implemented in becomes critical and no human response is developing countries made. The solution is a risk assessment of roads, Drivers/Riders Cars + Technology Two main technical 1) The managed to correct The major lessons learnt is based on vehicles’ dynamic data. Vehicles’ Buses/Trucks + issues: lots of data that it is possible to deliver a data such as acceleration, rolling angles, PTW 1) Data of positions were safety rating using collected Safety Roads Risk Assessment Road Safety Interview strength force, traction control, etc. are not consistent. dynamic data. Traditional Enhancement based on Dynamic Data measured by vehicles and, through a 2) They had difficulties in safety assessment depends machine learning algorithm, it is possible the straight forward on the assessors and can vary to identify unsafe roads and critical road classification of across different evaluators. sections. Roads and sections are rated driving/riding styles for Using dynamic data, results Adaptive Cruise Control for adjusting Drivers/Riders Cars + Technology Safety vehicle's speed and keeping a safe distance Buses/Trucks Road Safety Survey Adaptive Cruise Control Enhancement from the preceding vehicle.
Regulatory framework This can be intended as a meta-solution, Drivers/Riders Cars + Regulations 1) Managing expectations: 1) IAM is trying to address When there is lack of http://www.pacts.org.uk that is, the framework needed before a Buses/Trucks people do not fully this issue through their consistency in the wide offer /2019/09/pacts-launches- for safety systems’ Road Human Factors Regulations Interview solution can happen. understand what are the trainings for advanced of ADAS or automated new-report-what-does- HMIs across vehicles vehicle capabilities and drivers. features, consumers are the my-car-do/ Currently, in present cars, manufacturers they build incorrect ones who are losing out. and manufacturers and vehicles have different names and expectations about what Autoliv’s night vision system uses an Drivers/Riders Cars + Technology If the driver trust the The lesson learnt was to https://www.youtube.co Cognitive Autoliv night vision infrared camera mounted in the front grille Buses/Trucks system, there may be move the night vision m/watch?v=Kb9jBRo9U4 Road Human Factors Survey of the vehicle that senses temperature problems with false information from an 7 Ergonomics spotlight differences as sensitive as a tenth of a negatives. additional display inside the degree to create a highly detailed thermal vehicle (that required The solution addresses the need to reduce Drivers/Riders Buses/Trucks Technology The problem is to define a It has not been evaluated https://www.adasandme the high cognitive load among bus drivers, system that is as good as on real road but in .com/about-adasme/use- and to reduce their stress and fatigue the driver to do the simulators. cases/ problems. But also to make the docking docking so he or she do more smooth and environmental good. not try to override it. Driving the bus in complex urban environment is demanding and for a bus driver there is need to have a good view on all VRUs and at the same time stop and start in a smooth way, stil keeping the time table. Cognitive Road Human Factors Survey Safe docking at bus stop Ergonomics Passenger pick up/drop off automation for buses. Example: 1) As Peter, a bus driver, approaches a specified bus stop area, the system asks for driving control, that Peter agrees. 2) The system takes over and approaches the bus stop in a safe and comfortable manner. During the approach, Peter can leave his seat and interact with the passengers. 3) Once passengers have ended Making worldwide surveys in order to Passengers Cars Methodology We are about to launch We expect to have ride- Inclusion and User- User-centered Worldwide survey for have a wide experience how people use surveys in Australia, sharing aspetcs which could Road Survey centrdness design ride-sharing mobility and wilingness to share ride with France and the US, so no differ from the countries but public transport and to find decisive factors results up to now. very much looking for a for ride-sharing. For example: travel common key Figure 37: Screenshot of the Database of solutions
May 2020 59
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
5.5. Workshop with Consortium experts During the 2nd project Plenary Meeting held in Karlsruhe in November 2019, DBL organized a workshop involving all the meeting’s participants with the aim of collecting inputs from consortium experts to refine and finalize the methodology. The workshop had the duration of 2 hours and consisted in a group work activity during which the participants had to brainstorm and debate about relevant factors related to two evaluation categories of transferability:
• Applicability • Potential benefits and issues.
The participants were divided in 5 groups composed by 7-8 people and worked independently with the help of a facilitator; every group had a specific use case taken from the database of solutions. Since the vast majority of the experts were coming from the road domain, there were selected 5 solutions from other domains and it was asked the participants to try to transfer them in the road context. The brainstorming activity was supported by a poster and coloured post-its where every participant could write down what were in his/her opinion on the most important aspects to consider when trying to apply the solution to the road domain. Once all the inputs were collected, the group debated and tried to summarize the different contributions. The same activity was carried out also focussing on the potential benefit and issues. Towards the end of the workshop, all facilitators presented in plenary the results of their groups. The outputs of all the groups, as well as notes of the various presentations, were taken by DBL for further analyses.
5.6. Results From the analysis of the workshop outputs, four factors emerged as relevant to be considered when evaluating the transferability of solutions to other domains:
• Users • Technology • Regulations • Legislation
For all these factors, a series of questions was elaborated (Table 11) with the aim of guiding the user of the database in considering all the relevant potential issues and advantages when applying the solution to other domains.
Table 11: List of factors and question to evaluate the transferability of solutions Factors Questions Premise Considering the domain(s) where the solutions would be transferred: • Does the solution take into consideration special needs of users? • Is the solution inclusive in terms of physical, digital and geographical accessibility for users? Users • Is the solution inclusive in terms of costs for the users? • Are there specific skills needed to use the solution? • Would users need training to use it? May 2020 60
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Factors Questions • Does the solution imply new forms of interactions/communication between different users (teamwork) or different levels of stakeholders (e.g. front-end operators/organisational management)? • Would the solution change the nature of the work operators currently do with an impact on reducing/increasing their stress and workload? • Can the solution positively impact the level of safety of the overall system? • Would the solution need to introduce new tools/HMIs? • Is the solution’s HMI easy to use for all users? • Are there any barriers that might prevent the users’ acceptance of the solution? • Can the technology on which the solution is based be applied? • Can the solution’s technological features be adapted to the new domains’ technical and physical characteristics? Technology • Are the solution’s technological characteristics based on specific domain-related infrastructures? • Does the solution’s technological characteristics comply with the technical standards? • Does the solution comply with the regulations? Is the adaptation required worth? • Does the solution comply with the regulations in the countries where it would be Regulations applied? Is the adaptation required worth? • Does the solution need or entail a certification to be used? • Does the solution fit with the legislation? • Does the solution fit with the legislation in the countries where it would be applied? Legislation • Does the solution entail liability issues? • Does the solution entail liability issues in the countries where it would be applied? • Does the solution entail privacy issues? • Does the solution entail liability issues in the countries where it would be applied?
According to the type of solution, not all the questions will be relevant and applicable. The list of questions has to be considered as a checklist and does not require the evaluator to answer, but it suggests which aspects to study when evaluating the transferability of solutions and their transferability timeframe. Finally, the database of solutions, together with the checklist, offer an easy and agile tool to evaluate and foster the transferability of solutions across transport modes. The tool is conceived as an open resource to be used by experts from all transport modes and that can be further expanded and refined as more solutions are collected and evaluated.
5.7. Next steps During the entire length of the project, the activity will continue through the expansion of the database. On this regard, a thread will be opened in the ETM Forum where the database will be presented and shared with all forum’s participants. The thread will be periodically updated by DBL and will have a threefold purpose of:
• Collecting feedback from external experts in various transport domains
May 2020 61
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
• Integrating new solutions in the database • Disseminating the project updates and events.
Finally, the database may provide useful input to the other WPs. In particular, WP5 and the pilots’ leaders could take advantage of the database for reasoning about potential cross-fertilization of solutions related to their pilots.
May 2020 62
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
6 Taxonomy of Skills and Knowledge for AV operation
It is obvious that the advent of autonomous vehicles will bring changes to many fields and sectors and most of them have already been widely addresses by researchers and car manufacturers. But the above mentioned aspects are not the only ones affected by the future vehicles. The impact on the role of the ‘driver’ is anticipated to be significant in all transportation modes (cars, trucks, vessels, trains, airplanes) as the driver will have the opportunity to abstain from the driving process and be focused on secondary tasks. On the other hand, the driver should be aware of all the principles governing the automated function and their limitations and be trained adequately in order to be able to take over the control of the vehicle whenever necessary in the most appropriate and safe way. The driver is not the only user type affected by the advent of the automation. The labour sector will also be affected by the new type of vehicles as jobs will be alleviated and new will be created. People involved in the autonomous vehicle operation should be reskilled or upskilled in order to be able to deal with the new technology and the various systems. One of the outcomes of the Drive2TheFuture project is identifying and investigating the skills and knowledge required for an efficient and proper operation of any autonomous vehicle. Skills are defined as the expertise and the ability to efficiently perform a simple or more complex task and can be categorized in technical skills, social skills (like team working or communication), soft skills (like IT skills) or labour skills. Knowledge is the information, awareness or familiarity someone should have in order to get expertise and be able to perform a task. Both skills and knowledge can be obtained through training programs or courses, seminars or education as it will be presented in Activity 4.1 and Deliverable D4.1 of Drive2theFuture project. Project deliverables, reports, studies, articles, scientific papers, websites were reviewed for identifying the requirements for all workers and drivers. Both professional and private operators will be considered and all transportation sectors (road, rail, maritime, aviation) will be considered and vehicle types will be analysed in order to reveal the new needs arisen from the advent of automation. Furthermore, all automation levels will be taken into consideration as each one has different requirements while the autonomous vehicle operation will be decomposed into the cognitive domain. Finally, the skills and knowledge will be evaluated for their importance and prioritized accordingly. 6.1 Road Sector According to SAE (Society of Automotive Engineers), there are 6 levels of automation in the road transportation sector: no automation, driver assistance, partial automation, conditional automation, high and full automation. Each level requires different systems and sensors inside the vehicle and therefore additional driver skills and knowledge as automation level increases and gets more complicated. A study conducted by the Michigan Department of Transport and the Centre for Automotive Research (2016) identified the skills required to handle properly the automated driving systems, the impact of the automated functions in driving skills as well as how the driver should react when the systems reach their limitations. In case of autonomous vehicles of levels 2 and 3 the driver should continuously cooperate and collaborate with the vehicle on the driving task. The driver should be able to monitor the systems not only in terms of supervising their status but also their performance and their appropriate operation. The intensity of monitoring or “vigilant attention” as it is called, is higher in levels 2 and 3 and decreases in level 4 and 5 and requires driver concentration maintenance and supervision skills (Alonso Raposo et al., 2018). Since the automated operation is based on the various systems and sensors the vehicle is equipped with, the driver should be familiar with all the electronic devices, technologies and functions and be aware of how they work, the principles governing them, their capabilities and limitations so that he could understand their decisions and actions, recognize errors, received warnings and act accordingly (Manser et al., 2019, McDonald et al., 2015, McDonald et al., 2016). Knowledge of the location of these systems and sensors should also be gained so that they are not blocked resulting in inaccurate or no information. Awareness of the decisions taken by the various systems based on the surroundings and the information received is also important while
May 2020 63
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance information about the differences between the levels of automation is considered a requirement for a safe AV operation (Manser et al., 2019). Additionally, it is necessary for the driver of levels of automation 3 and 4 to be able to take over the vehicle control when necessary either due to system limitations or failure. The control take over could be either from automated function to manual or between different levels of automation (Spulber et al., 2016, Manser et al., 2019, Marinik et al., 2014). This transition of control should be performed in a safe way and within a specific time period (short notice) and therefore it is necessary that the driver keeps high levels of situational awareness and concentration even if he is engaged in secondary tasks (Hutchins, 2018, Google, Ford). For the most effective and efficient driver reaction in cases of systems failures or limitations of automated functions, the driver should be adequately trained in various emergency situations when it is necessary to take over the vehicle control. The autonomous operation of the vehicles of all automation levels is exclusively based on the various sensors, systems and the algorithms developed making them work properly as well as the efficient communication with the other vehicles and the infrastructure. ICT skills are necessary for vehicle manufacturing (European commission, 2017) and people with programming skills and experts in machine learning and artificial intelligence are needed for the algorithms and software development (European commission, 2019) so that the autonomous vehicles can understand the surroundings, detect any (physical) object around (vehicle, motorcycle, bicycle, pedestrian, lane crossing, lane marking, etc) and (re)act in a safe and efficient way. Since big amount of data (big data) from autonomous vehicles will be recorded continuously from its various sensors, backend software engineers are necessary for data storage services, design of APIs for proper communication between the sensors, between the vehicle and the infrastructure or between the sensors and the server/platform/cloud. Additionally, the cohesion and compatibility of the programs required for the operation of such a vehicle should be ensured. Design of the proper communication tools should be combined with the existence of communication models and wireless networks enabling the information and data transmission and exchange between the vehicle, the infrastructure as well as the traffic management center (TMC). The road sector is considered to be very complex as the vehicle coexists with other autonomous vehicles, conventional ones, motorcycles, cyclists and pedestrians and therefore the safe and appropriate interaction and communication between all road users is of major importance. The sensors installed in the vehicle serving the communication, detection and reaction should be able to scan the surroundings of the vehicle and transmit the correct information and messages to the driver or the “driver” so that he understands the information and reacts appropriately. In automation levels 3 and 4 the vehicle communicates with the infrastructure and its various units (lane marking, traffic lights, etc) and as a result both the vehicle and infrastructure units should exchange information in order to keep the vehicle into track or adapt its speed according to the traffic light signalization. For this reason infrastructure characteristics should be appropriately designed for supporting communication and information exchange and the road surface should be of good quality to support it (Fiedler et al., 2019). The interaction and communication skills of the autonomous vehicle with the surroundings is the basis for its safe operation and coexistence with the other users of the road network. People working in the TMC should have the knowledge of recognizing the data received and the skills to pro- cess and analyse it.
May 2020 64
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
•The full- time performance by the human driver of all aspects of the No dynamic driving task, even when enhanced by warning or intervention SAE Automation Level 0 systems
•The driving mode-specific execution by a driver assistance system of either steering or acceleration/deceleration using information about the Driver SAE driving environment and with the expectation that the human driver Assistance Level 1 perform all remaining aspects of the dynamic driving task
•The driving mode-specific execution by one or more driver assistance system of both steering and acceleration/deceleration using information Partial SAE about the driving environment and with expectation that the human Automation Level 2 driver perform all remaining aspects of the dynamic drivng task
•The driving mode - specific performance by an automated driving system Conditional SAE of all aspects of the dynamic driving task with the expectation that the Automation Level 3 human driver will respond appropriately to a request to intervene
•The driving mode - specific performance by an automated driving system High SAE of all aspects of the dynamic driving task even if a human driver does not Automation Level 4 respond appropriately to a request to intervene
•The full-time performance by an automated driving system of all aspects of Full SAE the dynamic driving task under all roadway and environmant conditions that Automation Level 5 can be managed by a human driver
Figure 38: Autonomous levels in road transport (SAE, 2018) As far as the public transport is concerned, the driver should be trained for operating an autonomous vehicle and should have the knowledge and skills of an autonomous car driver as it has been described previously. Additionally, the driver should be skilled for monitoring the operation of the vehicle remotely ensuring safety for the passengers and the other interacting road users. On the other hand, the passengers of an autonomous public transport bus should be familiar with its functions and operation principles so that they can recognize its actions and know who they should contact in case of emergency or how they should evacuate the bus if needed (Lundquist, 2018). Systems for stopping the vehicle when arriving at a bus stop, opening the doors and allow passengers (dis)embarkation and closing the doors belong to the previous categories with the engineering/programming skills. Due to the absence of driver, there will be need for high quality in-vehicle means of communication (Mirnig et al., 2019). Finally, a very promising concept is the Autonomous Mobility on Demand (AMoD) services where modelling and programming skills are also required (Zhang et al.,2015, Pavone, 2015). Currently, in Europe, pilots are taking place with SAE Level 3 vehicles in first and last-mile scenarios. These vehicles drive at usually low speeds and have dedicated infrastructure supporting them. For example, in Brussels, the public transport operator STIB started testing its shuttles around the Woluwe Park during the summer of 2019 (STIB, 2019). At EU level, the AVENUE project (AVENUE, 2020) is currently running by having shuttles tested across pilot sites in Europe without a driver’s seat but with a driver on-board. The shuttles operate under SAE Level 3 and drivers, despite not having a driver’s seat, are required to intervene in case the vehicle demands it. The hazard of cyber-attacks is of vital importance and therefore it is necessary to eliminate or alleviate any levels of hacking vulnerability. Experts with continuous education on these issues are required for ensuring cybersecurity and encryption protection, as well as monitoring traffic for detection of any suspicious activity. Along with the software engineers, robotics and electrical engineers will be involved in the design of hardware, electrical and communication systems as the autonomous vehicle can be considered as a robot
May 2020 65
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance that has to perform many complex tasks simultaneously. Finally, the automotive engineering skills already required for the design and construction of conventional vehicle will continue to be necessary but will be upgraded since the future cars need additional features and characteristics. Logistics operators should also be retrained and informed of the novelties that are related to their daily work. Different use cases are identified for freight transport operations. The use cases include hub-to-hub, confined areas and open roads. Hub-to-hub and confined areas could have the logistics operator performing remote- controlled operations or monitoring on board the vehicle. Open roads foresees primarily truck platooning where a platoon of trucks is remotely controlled. This is particularly relevant when considering driverless vehicles that are remote-controlled and used for remotely controlling the process and managing the delivered cargo. According to the ERTRAC Long Distance Freight Roadmap (2019), transport between hubs provide an opportunity to improve the efficiency of transport operations. Short distances such as from factories to ports or terminals can be considered when operating an automated or remote-controlled vehicle. In the case of a remote-controlled vehicle for both indoor and outdoor logistics operations, a controller is needed to control the autonomous vehicles in terms of dispatching, scheduling and routing (Vis, 2006) and monitor the process (Flaemig, 2016), while intervention is also possible via an operation center using wireless transmission. Logistics operators will also be needed in terminals in harbours (Skillful project, D1.1). Law specialists are required for formulating and establishing suitable regulations and legal framework for the operation of autonomous vehicles and managing issues related to the future vehicles as well as solve liability issues in case of failure or incident occurrence and develop ethical rules for the robots (Skillful project, D1.1).
Law specialists are required for formulating and establishing suitable regulations and legal framework for the operation of autonomous vehicles and managing issues related to the future vehicles as well as solve liability issues in case of failure or incident occurrence and develop ethical rules for the robots (Skillful project, D1.1). Regulations have already been established in national, European Level and international level, still limited though. Since 2016, numerous EU Member States started developing and testing autonomous vehicles, such as France, UK or Germany (ERTRAC CAD Roadmap, 2019). At EU level, the Declaration of Amsterdam (April 2016) represents the first of Europe working towards a harmonised approach to introduce connected and automated driving (ERTRAC CAD Roadmap, 2019). Key action points involve mainly the need to address legal and practical barriers to the testing and deployment of connected and automated vehicles. In May, 2018, the European Commission (EC) published legislative initiatives under the 3rd Mobility Package (EC, 2018) such as the Communication ‘On the road to automated mobility: an EU strategy for mobility of the future’ (EU, 2018) whose main topics are summarized in Table 12. Additionally, in 2019 the revised General Safety Regulation was revised (EU, 2019) sets forth that all types of road vehicles (cars, vans, trucks and busses) must be retrofitted with event data recorders for establishing who was driving (the automated vehicle or the driver) during an accident and to determine liability. Other EU regulations related to automated driving are Regulation (EC) 561/2006 on driving and resting times (EC, 2006), Regulation (EU) 165/2014 on tachographs in road transport (EU, 2014) and Directive 2003/88/EC on working time (EU, 2003), Directive 2009/103/EC on motor insurance (EC, 2009) and Directive 85/374/EEC on product liability (EEC, 1985).
Table 12: Topics included in the EC Communication ‘On the road of automated mobility: an EU strategy for mobility of the future’ (ERTRAC CCAM Roadmap 2019) Topics Description Technical/ 1. Systems providing the vehicle with real-time information on the state of the vehicle and Technological the surrounding area driver readiness monitoring systems 2. Accident data recorder for automated vehicles 3. Harmonised format for the exchange of data for the purposes of multi-brand vehicle platooning 4. Update of rules on road infrastructure safety management Legal 1. New vehicle type-approval rules 2. Proposal for new safety measures for driver assistance systems and autonomous driving 3. Proposal for new safety requirements for roads to support AVs May 2020 66
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Topics Description 4. Proposal for mandatory black box in automated vehicles 5. Upcoming guidelines on the product liability framework 6. Develop a balanced and fair framework for the sharing of vehicle data 7. Adopt rules to ensure secured communication, data protection and interoperability 8. Set recommendations on the use of spectrum for 5G large scale testing Societal 1. Assessment of the medium and long term socio-economic and environmental impacts 2. Support to the reskilling of the workforce 3. EU forum on ethics to address issues related driverless mobility 4. Set up ethical guidelines on the development of artificial intelligence
Finally, at international level, there are existing Conventions such as the Geneva Convention on Road Traffic (1949) (Geneva, 1949) which aims to promote the development and safety of international road traffic by establishing certain uniform rules. Its Article 8 is currently being debated upon at UNECE level and countries such as Germany have proceeded to amend the Vienna Convention on Road Traffic by allowing the transfer of driving tasks to the vehicle, provided that the United Nations vehicle regulations or can be overridden or switched off by the driver (LOC, 2016). In October 2018, the UNECE adopted a resolution on the deployment of highly and fully automated vehicles in road traffic. This resolution sets recommendations to ensure a safe interaction between automated vehicles, other vehicles and more generally all road users (ECE, 2018).
Table 13 summarizes the main outcomes of the review of the skills and knowledge necessary for AV operation in the road sector.
Regulation on cooperative, connected and automated driving is currently limited.
Table 13: Skills and Knowledge for AV operation in the road sector Skills Description Social Skills Communication, Team working, organization, problem-solving Programming Artificial Intelligence, Algorithms, software development, backend/frontend skills, and Computer machine learning, higher-order skills in big data analytics Skills Cybersecurity and encryption protection, security systems for protecting external communication for AVs, data protection Engineering/ Sensors and systems development, hardware development, Robotics, electrical Technical Skills engineering, automotive engineering, digital road map database access, firmware, Smart Traffic Light controller system, smart signs, advisory road marking, etc Testing and Simulation Skills Driver Skills and Cooperation and collaboration with the vehicle, Efficiently monitoring and supervising Knowledge the system, Concentration maintenance, Familiarity with all electronic devices and sensors on and inside the vehicle, limitations and capabilities, Understanding of the information and warnings from the systems based on the surroundings, Knowledge of differences among different levels of automation, Situational awareness and transition of control skills, Capability of recognizing errors and malfunctions and act properly Remote Skills and knowledge for efficient remote monitoring of the PT and freight and logistics operation transport operations in confined areas Communication V2I and V2V communication model, Wireless communication, ad hoc network, DSRC skills Multi-Channel Test Tool Traffic Collection and processing skills from the data transmitted from the infrastructure and management the vehicles center May 2020 67
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Skills Description Legal Legal framework and standards for the autonomous vehicle operation knowledge Social legislation and its adaptation to autonomous vehicle operations (driving and rest time rules) Liability issues in case of incident occurrence Data generated by V2X infrastructures to be compliant with national or international law
A safe and good driver is an experienced driver. In the first levels of automation where all or most functions and decisions are taken by the human driver, experience is of vital importance. Experience has two important functions. First of all, people learn more insight and develop their skills in different situation that cannot be learned or implemented in a training. Secondly, experience has also the function of making reactions (physical and cognitive) more naturally, obvious and automatic. This gives the possibility to react faster and more accurate in many situations. Different studies on driver education and training (OECD-ECMT, 2006) indicate that the first six months or during the first 5000 km of driving, the number of mistakes and accidents decline most pronounced, but it will take two to four years of driving independently before the ‘normal’ high rate of accidents will disappear. The study of Vlakveld (2004, 2005) indicates that this important role of experience plays a role at every age, although the accident risk at the time of licensing will differ between ages. In the case of levels 3 and 4 where the vehicle overtake several driving tasks and reacts in most cases autonomously, the driver has less opportunity to get full experience of the driving task and will probably need more time to learn it. This will probably have the effect that, in case of the need to overtake the control, the reaction will be slower and less accurate by missing a cognitive “arsenal”. In this case, continuous training is necessary so that the driver will get the experience of fast intervening always in strong combination with appropriate reaction. All training needs will be included in the report of A4.1. 6.2 Rail Sector The automation concept in the rail sector is much different than in the road sector and therefore the level of automation are defined in a different way and under different principles in operation. The automatic train protection is already in-stalled in GoA1 for ensuring the automatic activation of the brakes in case of speeding or other risky situations while the automated train protection is introduced from GoA2 for the control of acceleration and braking of the train. Since the first two GoAs require the presence of the driver, he or she should be aware of these systems, the consequences and risks from their operations as well as their capabilities and under which circumstances they are activated and which are the decisions they take. The manual operation of the train in GoA1 is substituted by semi- automated train operation in GoA2 where the driver should be familiar with the various displays and more monitoring and supervising tasks. Continuous monitoring, track super-vision, communication, operational knowledge, dependability, selective attention and other social, perceptual and cognitive skills are considered important for the train driver of GoA1 and GoA2 (Brandenburger and Naumann, 2016, 2019, Brandenburger et al., 2016).
May 2020 68
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
•Ensure safe speed: Automatic Train Protection (ATP) wih Driver •Driving (Control acceleration and braking): Driver Non Automated •Supervise track (prevent collision): Driver Train Operation GoA 1 •Supervision passenger transfer: Driver (NTO) •Operation in event of disruption: Driver
•Ensure safe speed: Automatic Train Protection system (ATP) •Driving (Control acceleration and braking): Automatic Train Operation Semi - (ATO) Automated Train •Supervise track (prevent collision): Driver Operation GoA 2 •Supervision passenger transfer: Driver •Operation in event of disruption: Driver •Ensure safe speed: Automatic Train Protection system (ATP) •Driving (Control acceleration and braking): Automatic Driverless Train •Supervise track (prevent collision): Automatic Operation (DTO) GoA 3 •Supervision passenger transfer: Train Attendant •Operation in event of disruption: Train Attendant
•Ensure safe speed: Automatic train protection system (ATP) •(Control acceleration and braking): Automatic Unattended Train •Supervise track (prevent collision): Automatic Operation (UTO) GoA 4 •Supervision passenger transfer: Automatic •Operation in event of disruption: Automatic (Train Attendant
Figure 39: Autonomous levels in rail sector (IEC, 2009) Additionally, high levels of situational awareness should be kept as the driver may need to intervene (e.g. manual speed control) when automation fails or its limitations are exceeded and therefore his reaction and performance should be appropriate to ensure safety (Brandenburger and Naumann, 2019, Wickens et al., 2010). Therefore, skills of critical diagnosis and maintenance are considered significant (Rail IRC Skills Forecast 2018). Due to the fact that in GoA1 and GoA2 the driver is on board, network controllers should ensure safety by providing the right information to the driver real time (Rail IRC Skills Forecast 2018). Concerning GoA3, the driver is an operator controlling remotely the train in strong cooperation with the train attendant who is still in the train supervising passengers exchange and detecting emergencies (Fiedler et al., 2019). Communication and team working skills are necessary for both of them for ensuring the proper information exchange and understanding. Due to the large amount of data transmitted to the control center and their complexity, remote operators are required to have high skills in big data analytics and problem- solving for maintaining high operation and reliability levels (Rail IRC Skills Forecast 2018, Department of Infrastructure and Regional Development, 2016). Manual intervention is still necessary in cases of emergency when the train operator takes over the train control remotely (Brandenburger and Naumann, 2018). Remote operators and drivers should also have safety management skills for incident recovery including fault identification (Case for Change – Autonomous Rail Vehicle & Human Factors, 2018) and fault rectification (Rail IRC Skills Forecast, 2017) and thus on route driving skills should be maintained in case of emergency (Rail IRC Skills Forecast, 2017). The people involved in the development of these technologies should be skilled for their proper design, installation and operation for ensuring safety and immediate response and activation when needed. Technical and engineering skills (Shift2Rail, 2019) are necessary for GoA3 and GoA4 for enabling driverless and unattended train operation respectively in terms of track and passenger transfer supervision as well as the train operation in event of disruption, obstacle, people and animal detection for collision avoidance, existence of other trains on the route (Bienfait et al., 2012) or lineside signalling observation (Rangra et al, 2018). The systems include signalling, on board and trackside modifications (GoA1 and GoA2) and additionally communication and monitoring systems, and measures to supervise track for GoA3 and GoA4.
May 2020 69
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Furthermore, wireless signalling, sensors and communication technologies needed to be designed and developed for enabling the data capture and transmission between the train, track and signals (Fiedler et al., 2019, Rail IRC Skills Forecast 2018 and Rail IRC Skills Forecast 2017) According to Ghantous-Mouawad et al. (2006) the transition to GoA3 and GoA4 requires communication skills for enabling the communication with the operational center, skills in timetable management and passenger exchange while slit second decision making, cooperation, teamwork and productivity are among the social skills considered important for all transportation sectors (CEDEFOP, Case for Change – Autonomous Rail Vehicle & Human Factors, 2018, Rail Sector Overview, 2018). Since rail safety work is considered to be dangerous enough, skills and knowledge in human factors for ensuring health and safety of the workers and the passengers as well as for increasing human performance and reduce failures and errors are also important (Case for Change – Autonomous Rail Vehicle & Human Factors, 2018). Similarly to the road sector, artificial intelligence and software and hardware skills are also required (Shift2Rail, 2019) for efficient visual perception and remote control (European Commission, 2019) in case of driverless trains while according to Rangra et al. (2018), skills on software and hardware assessment are necessary for ensuring safety. Finally, augmented and virtual reality and simulation skills are a prerequisite for developing, testing as well as constructing a rail infrastructure and control or maintenance operations (Case for Change – Autonomous Rail Vehicle & Human Factors, 2018, Rail Sector Overview, 2018, Rail IRC Skills Forecast, 2018, Rail Education and Training, 2016, A national rail industry plan for the benefit of Australia, 2017, Skillful project, D1.1). Regulations and guidelines should be also established in the rail sector for enabling the operation of trains of different levels of automation. Specific regulations should be stablished in case that both semi and fully automated trains are using the same track (Ghantous-Mouawad et al. (2006). Liability issues for driverless trains are necessary to be defined in terms of if the cooperation or any other entity is responsible in case of incident or failure. The role of signaller will also be affected by the advent of automation as they are required to have deep knowledge of all the new signalling systems, continuous monitoring the decisions the systems take and intervene when necessary (Brandenburger et al, 2018). Especially in line crossings or in case of incident occurrence and delays they should be highly skilled to arrange the train schedule efficiently so that no conflict will occur. The number of trains plays a significant role as the more trains are under the signaller’s control, the more skilled the signaller should be to monitor and arrange every journey and rearrange the schedule and manually intervene in the system. Additionally, according to Hayden-Smith (2013) the signaller has under his operation many lane crossings simultaneously but each time he can intervene only in one of them and thus he should set priorities and emergencies. As automation in signalling increases, the signaller abstains from manual work and should continuously monitor the system for ensuring safety performance (Hayden-Smith, 2013) and be able to intervene in case of unexpected events (such as incident occurrence) or in cases outside the capabilities of the automated system (Sharples et al., 2010). Therefore, it is necessary that the signaller should be capable of maintaining high levels of skills when assuming manual operation of a complex system. Finally, train drivers, signallers, controllers and workers should gain adequate knowledge, get familiar with new technologies and have specific skills for their implementation and maintenance. Such technologies are the European Train Control System (ETCS) and the European Rail Traffic Management System (ERTMS) which standardizes and automates various services and tasks (Fiedler et al., 2019, AAIMESC PROJECT) as well as the Satellite Based Augmentation Systems (SBAS) in Australia for navigation purposes (Rail Sector Overview, 2018, Rail IRC Skills Forecast 2017). Finally, workers should be further upskilled and qualified for maintaining a safe and efficient pre-journey, in journey and post journey operation of an autonomous train (Case for Change – Autonomous Rail Vehicle & Human Factors, 2018, Rail IRC Skills Forecast 2017). The main skills and knowledge required in the rail sector concerning autonomous vehicles are summarized in Table 15.
Concerning the urban environment, differences exist among the underground and elevated rail systems. Similarly, to a high-speed train, an automated underground metro line has a fully segregated infrastructure
May 2020 70
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance and there is no interaction with other modes or vulnerable road users (SYSTRA, 2018). The levels of automation are defined similarly to the rail trains but in case of metro lines, fully automated vehicles already exist worldwide. When manually, the driver is responsible for detecting any obstacles or other trains in front of the vehicle, observe the signalling and act accordingly. As level of automation increases, systems for collision avoidance, obstacle detection or spacing with the preceding vehicle are conducted by systems. On the other hand, trams are fully integrated in the mixed traffic and have strong interaction with passenger cars, trucks, cyclists and pedestrians while track crossing is also possible. An automated or autonomous tram is necessary to be equipped with various sensors (lidar, radar and cameras) for horizontal and vertical object scanning, distance estimation and environment/surroundings sensing and object or signal recognition (SIEMENS, 2019). Processes such as signal or station approaching and detection of vehicles or pedestrian track crossing (SIEMENS, 2019) are of vital importance and therefore appropriate sensors and algorithms development, artificial intelligence and deep learning techniques are necessary for adjusting speed and ensuring safety. On board intelligence systems are also required for noticing early enough any danger, avoid emergency braking (SYSTRA, 2018) and handle complex situations that may occur (SIEMENS, 2019). In case of non-autonomous trams, the driver should keep safe spacing from the preceding tram vehicle, control the vehicle, monitor the surrounding traffic, supervise the assistance systems and maintain higher levels of awareness than a metro driver especially in the cases that he has to intervene (SYSTRA, 2018). Communication and legal skills are also necessary for enabling the information exchange and handle regulations and liability issues.
Table 14: Skills and Knowledge for AV operation in the rail sector Skills Description Social Skills Communication, Team working, organization, skills in timetable management, problem- solving, slit-second decision making, Knowledge in human factors for passengers and workers safety Programming Artificial Intelligence, Algorithms, software development, backend/frontend skills, machine and Computer learning, higher-order skills in big data analysis Skills Cybersecurity and encryption protection, security systems for protecting external communication for AVs, data protection Engineering/ Sensors and systems development, hardware development, Robotics, electrical Technical Skills engineering, automotive engineering, systems for driverless and unattended train operation, automatic train protection and automatic train operation, train operation in event of disruption, obstacle, people and animal detection for collision avoidance, existence of other trains on the route or lineside signaling observation, diagnostics Signaling technologies, Testing and Simulation Skills Technical Knowledge in new signalling and position technologies, Knowledge of the European Train Knowledge Control System (ETCS) and wireless delivery of mission-critical rail communications, digital interlocking system Driver/Crew Same as for road sector (Table 1) Skills and Maintenance of on route driving skills, knowledge of new on board systems Knowledge Monitoring of the passenger exchange, detection and accomplishment of emergency conditions, supervision of the train’s state. Communication V2I communication model skills Wireless communication, ad hoc network, Wireless interface/connection and components, data transmission systems Legal skills Legal framework and standards for the autonomous vehicle operation Liability issues in case of incident occurrence Data generated by V2X infrastructures to be compliant with national or international law Skills for Rail vehicle setup and deconstruction skills and knowledge for a safe and efficient pre- workers in train journey, in journey and post journey autonomous train operation, Skilled rail network driving, front controllers May 2020 71
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Skills Description line and network control Safety Preparing for emergencies related to both safety and environmental protection, fatigue management management, safe management of door closures skills Remote Control Off site and remote fault support skills, skills and knowledge for Incident recovery Skills procedures for autonomous trains and rail vehicles, including fault identification and rectification, remote operations, processing of large amount of data Signaller Knowledge of all new signalling technologies and systems, ready to intervene efficiently any time Track Knowledge of replacing the components Maintenance
6.3 Maritime Sector In the maritime sector, there is not an international and standardizes definition of the levels of automation and therefore different taxonomies have been defined such as from the International Maritime Organization or the Lloyd’s Register Group Limited. Systems for autonomous vessels have already been developed and tested, creating the necessary circumstances for encouraging automation also in the maritime sector. Similarly to the road sector, the professional driver of an autonomous vessel is required to have technical and engineering skills in order to be able to deal with any malfunction or failure of the hull structure, the machinery and other systems that may occur (Skillful project, D3.1, Skillful project, D3.6). In case of remote vessel control, the people working in the shore control center should be able to understand and interpret the pertinent data transmitted from the various systems and sensors installed on the vessel to the shore- based facility in case of a machinery/equipment/hull dam-age event (Fiedler et al., 2019) as well as being able to navigate it (World Maritime University, 2019). Operation monitoring, emergency situations handling, autonomous ship surveillance and additional safety related tasks are performed from the shore control center by skilled and trained personnel. Data should be monitored and controlled via maritime broadband radio, and satellite communication (World Maritime University, 2019). Human intervention may be required at any time and under various conditions and as a result seafarers on the bridge or in the shore based center should be properly skilled and maintain high levels of situational awareness (World Maritime University, 2019, Man et al.,2014, Porathe et al., 2014). Due to the fact that the vessels may switching from one autonomy level to the other (i.e. with crew on board when navigating through narrow channels and without crew on-board when navigating in the open seas etc.) the people involved should be able to distinguish the different principles governing each mode-autonomy level (Skillful project, D3.1, Skillful project, D3.6). Additionally, the personnel in charge should be able to control the satellite communication capacity and the bandwidth to ensure that the vessels is safe when operated remotely. Coastal crew can get on the vessel and assist in navigation inside the port or in the mooring process (Fiedler et al., 2019).
May 2020 72
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
•Ship with automated processes and decision support: Seafarers are on board to operate and control shipboard systems and functions. Some operations may be automated and at times Degree be unsupervised but with seafarers on board ready to take control. 1 •Remotely controlled ship with seafarers on board: The ship is controlled and operated from another location. Seafarers are available on board to take control and to operate the Degree shipboard systems and functions 2
•Remotely controlled ship without seafarers on board: The ship is controlled and operated from another location. There are no seafarers on board. Degree 3
•The operating system of the ship is able to make decisions and determine actions by itself. Degree 4
Figure 40: Autonomous levels in maritime sector (IMO, 2018)
•No autonomous function. All action and decision-making performed Manual AL 0 manually, i.e human controls all actions
•All actions taken by a human operator, but decision support tool can On-board present options or otherwise influence the actions chosen. Data is provided AL 1 by systems on board. decision support
•All actions taken by a human operator, but decision support tool can On and off-board present options or otherwise influence the actions chosen. Data may be AL 2 provided by systems on or off board. decision support
•Decisions and actions are performed with human supervision. Data may be “Active” human AL 3 provided by systems on or off board. in the loop
•Decisions and actions are performed autonomously with human Human on loop supervision. High impact decisions are implemented in way to give human operator/ AL 4 operators the opportunity to intercede and over-ride supervisory
•Rarely supervised operation where decisions are entirely made and Fully AL 5 actioned by the system autonomous
•Unsupervised operation where decisions are entirely made and Fully AL 6 actioned by the system during the mission autonomous
Figure 41: Autonomous levels in maritime sector (Lloyd’s Register, 2016, 2017)
May 2020 73
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
According to World Maritime University (2019a, 2019b), the shore-based personnel (i.e. located at the remote control center) should have, apart from maritime knowledge, digital and software engineering skills as well as data fluency and programming skills for interpreting large amount of data. Since there are different types of marine vessels, interoperability skills are necessary for the AV operator in order to efficiently handle different operational profiles. The AV operator would need to deal with various parties (i.e. shipyards, classification societies, pilots, port authorities and more), in a way that differs significantly with the ship type involved (i.e. operation of passenger vessels is very much different to that of tanker vessel etc.) and there- fore communication skills are considered among the most important competencies for the ship operation (International Chamber of Shipping, 2018). One of the limitations and major concerns related to the advent of autonomous vehicles is the new legislation that has to be established. Currently, legislation issues do not favour the unrestricted operation of autonomous vessels. For example, fundamental principles for the operation of autonomous vessels (i.e. unmanned bridge) are in conflict with the current pertinent legislation to prevent collisions at sea (IMO, 1972, 2018a, 2018b). Protection and indemnity clubs (P&I Clubs), charterers and other third parties in the maritime sector, currently do not take any liability for accidents that would involve autonomous vessels. It is apparent, that AV operators need to acquire knowledge about the legal framework associated with the autonomous vessels (i.e. remotely operated or/and unmanned shipping), which is about to change considerably in order to account for the new era-reality of the autonomous vessels (International Chamber of Shipping, 2018). It must be ensured that applicable codes, guidelines and standards prescribed by the IMO, Flag States, Classification Societies and other policy makers are taken into account. Programming, engineering and technical skills are also required for developing all the systems, sensors and technologies the autonomous vessel should be equipped with so that it can navigate itself on a specific route, detect obstacle on its route and avoid collisions as well as and perform the necessary actions (e.g. manoeuvring) (Bureau Veritas, 2017). Fully autonomous ships will be also equipped with automatic mooring and unmooring systems or with detachable bridge that will be removed when the vessel is ready for autonomous operation (Kongsberg). For auto mooring the required infrastructure and communication infrastructure should also be developed by skilled people (Fiedler et al., 2019). Besides the mooring system, V2V and V2I connectivity and communication should also be designed and be enabled by developing sensors and platforms for information exchange as well as systems such as maritime broadband radio, WIFI or Global System for Mobile Communications (GSM) (Kongsberg). Maritime operations and ship maintenance can be either performed manually or remotely using robots (project 'Robotic Vessels as-a-Service' (RoboVaaS)) creating safer conditions for the coastal workers. Data transmission networks, sensors, augmented and virtual reality can enable surface and under water communication, remote services and maintenance work. Drone technology can also assist crew in surveying and inspecting any part of the ship (Fraunhofer CML, 2019, World Maritime University, 2019b). As it has already been mentioned in the case of the road sector, cybersecurity is extremely important for the safe AV operation. Continuous development of cybersecurity skills is essential to support safe and secure shipping, which is operationally resilient to cyber risks (i.e. cybercrime, terrorism etc.) (Komianos, 2018, Vartdal et al., 2018). Safety management skills are also required for an autonomous vessel operation “related to both safety and environment protection” (MSC, 2017). Since a vessel with any degree of autonomy affect the operational organization, -responsibilities and -decision process, the safety management system will be affected. The pertinent skills would need to be acquired on a lifelong learning basis, including preparing for emergencies related to both safety and environ-mental protection. It is foreseen that the operators’ safety management system will be subject to audits based on elements from the ISM scheme. Communication and team working skills are essential for the successful accomplishment of task associated with activities of on board and shore-based personnel, all working as a team to operate autonomous ships (i.e. for navigation and other operation activities etc.). For vessels with personnel on board, local/manual restoration by on-board crew may be relied upon if adequate competence, instructions or assistance by a remote control center is available, all under the umbrella of good communication and teamwork (Carey, 2019).
May 2020 74
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Finally, port workers such as quay cranes operators, gate entrance controllers and dockers are also going to be upskilled and retrained. A new remote-control driver of an automated quay crane should have general knowledge on electronics and mechanics along with control panel handling skills while for an experienced one only the handling skills are necessary (World Maritime University, 2019b). On the other hand, since each port or terminal may have different operational processes a docker is required to have deep knowledge of the different process. In case of fully automated terminals skills such as efficient planning, scheduling and equipment dis-patching, monitoring and remote control of ship handling are necessary for a docker (World Maritime University, 2019b). Table 16 summarizes the outcomes of the review of the main skills and knowledge required in the maritime sector in relation to automation.
Table 15: Skills and Knowledge for AV operation in the maritime sector Skills Description Social Skills Communication, Team working, organization, skills in timetable management, problem- solving, slit-second decision making, on-board and shore-based personnel Knowledge in human factors for passengers and workers safety Programming Artificial Intelligence, Algorithms, software development, backend/frontend skills, and Computer machine learning, higher-order skills in big data analysis, augmented and virtual reality Skills skills and knowledge Cybersecurity and encryption protection, security systems for protecting external communication for AVs, data protection Engineering/ Sensors and systems development, hardware development, Robotics, electrical Technical Skills engineering, automotive engineering, obstacle detection, surroundings mapping, mooring and unmooring systems, HD Maps of the relevant port transport infrastructure, naval engineer Testing and Simulation Skills Airborne or underwater drones can perform potentially hazardous inspection and maintenance tasks, either by remote control or autonomously (in cooperation with programming and computer skills). Driver/Crew Same as for road sector (Table 1) Skills and Knowledge of new on board systems, Interoperability Skills, Docking skills, Coast water Knowledge crews inner-port navigation the mooring skills Monitoring of the passenger exchange, detection and accomplishment of emergency conditions, supervision of the vessel’s state. Communication Satellite communication capacity and the bandwidth, advanced data transmission skills technology systems, communication network, V2V and V2I communication Legal skills Legal framework and standards for the autonomous vehicle operation, liability issues in case of incident occurrence, data generated by V2X infrastructures to be compliant with national or international law Safety Preparing for emergencies related to both safety and environmental protection management skills Remote Control Understand and interpret the pertinent data transmitted from the vessel to the shore- Skills based facility in case of a machinery/equipment/hull damage event and any other case concerning safety Distinguish the different principles governing each type -Interoperability skills Mooring and unmooring operation skills Complex engines and machinery aboard monitoring Data analytic experts and system controllers
May 2020 75
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance 6.4 Aviation Sector Monitoring the automation systems is a skill required also in the aviation sector when the state of autonomy is level 3 or 4. The pilot should be capable of detecting any system malfunction or suspicious performance as well as be ready to react properly in case of failure in order to avoid air crashes. In contrast to the car drivers, who can recall their driving skills even if automated systems are used, the basic flying skills of a pilot are not retained (wsp.com) due to the complexity of the pilot tasks. Additionally, the pilot of airplanes of these autonomous levels is required to have a deep knowledge of the systems and appropriately distinguish the various kind of information received from them and take the right decisions accordingly.
•Information Acquisition: A0-Manually, A1-Supported by artefact •Information Analysis: B0-Manually, B1-Supported by artefact •Decision and Action Selection: C0-Manually, C1-Supported by artefact Done by Humans •Action Implementation: D0-Manually, D1-Supported by artefact
•Information Acquisition: A2-With user filtering and highlighting criteria, A3-With user control of filtering and highlighting criteria, A4-With user awareness of filtering and highlighting criteria, A5 - With filtering and highlighting criteria not visible to the user •Information Analysis: B2-On user request, B3-On user request with alerting mechanism, B4-With user setting of elerting parameters, B5-With alerting parameters not visible to the user Supported by •Decision and Action Selection: C2-With user choice and acceptance among proposals, Automation C3-With user acceptance of one proposal •Action Implementation: D2-With user activationand control on actions, D3-with user activation and control of actions, D4-With user activation , monitoring and interruption of action sequence
•Decision and Action Selection: C4-With user informed, C5-With user informed on requet, C6-With user not informed •Action Implementation: D5-With user monitoring, modification and interruption capabilities, D6-With user monitoring and interruption capabilities, D7-With limited Done by user monitoring, modification and interruption capabilities, D8 -With no user Automation monitoring, modification and interruption capabilities
Figure 42: Autonomous levels in aviation sector (Lloyd’s Register, 2016, 2017) In the case of autonomous airplanes the pilot should have all the necessary and required skills and knowledge in order to efficiently and safely supervising remotely the airplane. Due to the fact that remote control is more difficult and demanding than the on board control and supervision, he should be completely aware of every system installed in the plane as well as its level of automation, its capabilities and limitations (Pavlas et al., 2009). It is necessary that he is capable of promptly detecting any suspicious activities of the systems and any abnormal behaviour of the plane and be prepared for handling any situation. Monitoring tasks are among the basic skills a UAS operator should have including instrument monitoring, navigation, route and long term monitoring so that safety is ensured (Pavlas et al., 2009). Additionally, in the future, one pilot may have to supervise more than one un-manned airplanes simultaneously and it is obvious that he should be highly skilled and well trained. Different types of aircrafts have different types of systems (Pavlas et al., 2009), they follow different routes and they are flying on different airways and therefore the remote controller May 2020 76
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance should ensure their safe operation and journey. They should ensure that the plane is on a predefined route at cruising altitudes. Additionally, a pre-flight check of the systems of the plane is necessary from the remote pilot controller. Social skills are also considered important for an UAS operator such as decision making, risk assessment skills or team leadership and communication skills (Pavlas et al., 2009, Tvaryanas et al., 2007, Wilson, 2002, Sharma and Chakravarti, 2005). Similarly to the other transportation sec-tors, high levels of situation awareness should be preserved also in the case of an UAS operation (Pavlas et al., 2009).
In the future, the use of unmanned aircrafts will be expanded also in the urban environment serving personal and goods transportation resulting in alleviating traffic congestion phenomena. According to Colombo (2019), the safe operation and performance of these aircrafts are based on three functions: the landing and taking off process will be executed without a runway, the aircrafts should be able to detect, see and avoid obstacles, like buildings and vehicles and last but not least the efficient management of emergency situations such as weather conditions. Similarly to all the other transport modes, engineering skills are required for sensors (lidar, radar, cameras) and systems development, simulation modelling, software and hardware development and testing so that the autonomous aircraft will optimum perform the tasks of perception, decision/planning and execution. People involved in these processes should ensure that information is received from all systems and sensors and that this information is accurate and correct. The most complete and precise information will enable a safe flight.
Apart from the technical part, the operation of UAS needs the establishment of standards, regulations and operational rules. The Federal Aviation Administration (FAA, 2016) formulated rules for small unmanned aircraft systems in 2016 including (a) operational limitations, (b) Remote Pilot in Command Certification and Responsibilities, (c) Aircraft Requirements and (d) model aircraft. Towards the full automation of bigger airplanes, it is necessary that people involved in the technical and legal sector should be cooperated in order to formulate the appropriate legal and operation framework as well as rules for the safe operation of UAS.
Drones are nowadays used for short range surveillance purposes and their operators should be able to identify obstacles and modify the drone route accordingly. In the future, drones will also be used for transporting people and goods within urban environments and therefore they will cover longer distances. One of the biggest challenges in their safe operation is the appropriate infrastructure where drones will land and take off. Additionally, since drones will fly in low altitudes it is necessary that they are equipped with systems for obstacle detection and avoidance (buildings, etc) while regulations and legislation should be established for their safe operation and efficient management of the airspace (Duval et al., 2019, EASA, 2015, McNeal, 2014). The guidelines should also include the characteristics and specific requirements of the drones. Finally, unmanned traffic management including sensors, communication systems and servers will be introduced for coordinating and monitoring the large number of drones in an urban or rural environment as well as for receiving data and providing real time information (Duval et al., 2019). All services consisting the UTM should communicate based on common rules (AIRBUS, 2019). Table 16 is a summary of the main skills and knowledge for autonomous aircrafts and drones operation.
Table 16: Skills and Knowledge for AV operation in the aviation sector Skills Description Interpersonal teamwork skills, conflict management skills, stress management skills, organization, Skills leadership, skills in timetable management, slit-second decision making Personal Identify and manage risks effectively, ability to cope with complex and stressful situations, resilience and problem solving, workload management skills critical thinking Programming Artificial Intelligence, Algorithms, software development, backend/frontend skills, and Computer machine learning, higher-order skills in big data analytics, augmented and virtual reality Skills skills and knowledge
May 2020 77
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Skills Description Cybersecurity and encryption protection, security systems for protecting external communication for AVs, data protection Engineering/ Sensors and systems development, hardware development, Robotics (able of performing Technical Skills maintenance work that cannot be handled by humans), electrical engineering, aeronautics, automotive engineering, safe navigation systems development ,Testing and Simulation Skills, Airborne or underwater drones can perform potentially hazardous inspection and maintenance tasks, either by remote control or autonomously (in cooperation with programming and computer skills). Driver/Crew Same as for road sector (Table 1) Skills and Knowledge of new on board systems, Interoperability Skills, Monitoring of the passenger Knowledge exchange, detection and accomplishment of emergency conditions. Communication Satellite communication capacity and the bandwidth, advanced data transmission skills technology systems, communication network, effective communication skills, emergency communication skills Legal skills Legal framework and standards for the autonomous vehicle operation, liability issues in case of incident occurrence, data generated by V2X infrastructures to be compliant with national or international law Safety Emergency Plan preparation, risk assessment, emergency management, application of management procedures, effectively monitors aircraft using automation skills Remote Control More difficult and demanding than the on board control and supervision Skills Detection of suspicious activities or abnormal behaviour of the plane Simultaneously monitoring and supervision of more than one unmanned airplanes Knowledge of characteristics of different types of aircraft, the routes they follow Pre-flight Check Urban Engineering/Technical/Programming Skills (Landing and take-off without a runway, Environment obstacle detection and avoidance) Operation
May 2020 78
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance 7 Use cases and priority scenarios 7.1 Methodology The upper goal of this Chapter is to present the Drive2theFuture Use Cases that will constitute the reference for all future implementation and demonstration activities that will take place within the project. The Drive2theFuture Use Cases have been defined within the context of WP1: “Driver”, traveller and stakeholder clustering a priori needs and wants and UC’s”, upon a defined user-centered methodological approach and are reflecting all the relevant users’ needs as well as the International Advisory Board (IAB) and Consortium experts’ views and filtering. The Drive2theFuture Use Cases are related to 2 main outcomes of the project, which are also going to be tested during the project’s pilots, namely the (1) HMI & strategies for the different clusters of users and levels of automation (WP3) and (2) the training programmes and schemes that will be suggested by the project for the “drivers”, users and stakeholders (WP4). For the initial selection of the Use Cases, 3 main parameters have been considered and combined, in relation to the project implementation. More specifically, the different modes and types of vehicles that are included in the project’s pilots, the respective users of these vehicles, as well as the Drive2theFuture tools that will be tested (Figure 43).
Figure 43: Parameters combined for the development and description of the Drive2theFuture Use Cases From the work that followed, involving the matching of these parameters, 13 different Use Cases came up, covering all modes and all types of users, as presented in Table 17 below:
Table 17: Drive2theFuture Use Cases Air transport Maritime transport Rail transport Road transport Training for air Training for maritime Training for rail Training for road transport (for drone transport (including transport (including transport (including operators). training of operators, training of operators, training of drivers, regarding also the signallers and passengers, VRUs, interaction of other passengers). operators, etc.). vessels with automated ships). In vehicle HMI & In vehicle HMI & strategies for rail strategies for automated vehicles (for rail drivers road vehicles (for and passengers). drivers/riders and passengers).
May 2020 79
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Air transport Maritime transport Rail transport Road transport AV Conspicuity HMI & AV Conspicuity HMI& AV Conspicuity HMI& strategies for strategies for strategies for interaction interaction of interaction of of automated road automated ships with automated rail vehicles with non- other non-equipped vehicles with other equipped other road vessels. non-equipped trains. users (including HMI& strategies for interaction with non-equipped vehicles and VRUs). Operators HMI & Operators HMI & Operators HMI & Operators HMI & strategies for air strategies for maritime strategies for rail strategies for road transport (for drone transport. transport. transport. operators).
For the finalisation of the Use Cases but also in order for them to be prioritised, a dedicated session was organised, within the context of the 1st Drive2theFuture Workshop in Brussels on March 6th 2020, with the participation of more than 40 people (physically and online). As presented in Figure 44 and Figure 45 below, the majority of the participants were researchers, however, almost all types of stakeholders and modes were represented.
0% Users
32% PT Operator Local authorities 53% Manufacturers (OEMs) 8% 1% Mobility Service Providers 1% Researcher 5% EC
Figure 44: Types of stakeholders that participated in the Drive2thFuture UCs workshop
4% 3%
32% 10% Air Maritime Rail 51% Road Multimodal
Figure 45: Modes representation in the Drive2thFuture UCs workshop
During this dedicated session, upon the presentation of the preliminary Drive2theFuture Use Cases, along with the different criteria for their prioritisation, stemming from the corresponding assessment areas, the participants were asked to rank, first the criteria (in terms of their importance), and then each Use Case according to each criterion, in both cases using an online tool through their mobile phones. This exercise has
May 2020 80
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance concluded to the prioritisation of the criteria (Figure 46) and the prioritisation of the Use Cases as such (as presented in Table 18).
Figure 46: Prioritisation of UCs assessment criteria
Table 18: Ranking of Drive2theFuture Use Cases after the workshop Rank Use Case #1 Training for road transport #2 Training for rail transport #3 Training for air transport #4 Training for maritime transport #5 Operators-based HMI& strategies for road transport #6 Operators HMI& strategies for rail transport #7 Operators HMI& strategies for air transport #8 AV Conspicuity HMI& strategies for interaction of automated road vehicles with non- equipped other road users #9 Operators HMI& strategies for maritime transport #10 AV Conspicuity HMI& strategies for interaction of automated rail vehicles with other non-equipped trains #11 In vehicle HMI & strategies for rail vehicles #12 In vehicle HMI & strategies for automated road vehicles #13 AV Conspicuity HMI& strategies for interaction of automated ships with other non- equipped vessels
During the analysis of the Use Cases and their description per Pilot Site, it has been decided that the UC regarding “AV Conspicuity HMI& strategies for interaction of automated rail vehicles with other non- equipped trains”, could not be efficiently supported by the rail pilots of the project, so it was decided not to be included in final project’s UCs, considering also the fact that it has been ranked low in the rail transport related UCs. So, there are 12 final Drive2theFuture Use Cases, which are in detail analysed below in Section 0 below, also in Pilot Site(s) level.Analysis of Use Cases After the finalisation and the prioritisation of the Use Cases, the mapping to each of the project’s Pilot Site followed, together with the full analysis and description of each Use Case, in order to facilitate the organisation and the implementation of the project’s pilots and the evaluation of its outcomes.
May 2020 81
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance 7.1.1 Training for road transport 7.1.1.1 General description Title: Training for road transport
Relevant WP/Activity WP4, WP5/ A1.7 Short description (output - This Use Case focuses on the training of users involved in the objectives) Drive2theFuture piloting of road transport automated vehicles. In particular, it can be divided in 4 sub- Use Cases, dealing with the training of (1) AV “drivers” and operators, (2) AV passengers, (3) other road users (including VRUs) and (4) infrastructure/ fleet operators/ planners/ supervisors. The purpose of this Use Case is the provision of training, which will be tested in the respective pilot sites, aiming to prepare all users (including VRUs) for the interaction with autonomous vehicles and the identification of critical situations and scenarios, as well as the interaction between automated fleet with non-automated drivers/riders (mixed flows), along with the behavioural adaptation of users when using an automated vehicle. This will also lead to the evidence of gained acceptance of road vehicle users and the assessment of awareness and perception of automated vehicles in different contexts. Target user/stakeholder • Cyclists clusters • Pedestrians • Drivers/riders of automated vehicles • Drivers/riders of non-automated vehicles • Powered two Wheelers (PtW) • Bus passengers • Bus operators (of L4 bus) • Bus drivers • Traffic Management Center (TMC) operators Mode(s) Road transport Type of vehicles • AV shuttle buses • Automated cars • PtW simulator (level 0) and mixed vehicle flows (level 0-5) Target SAE Levels • Levels 2-3-4 Key Research Hypothesis • Adaptation of AV’s behaviour according to observed, abstracted behaviours, so that the new AV behaviour is accepted by users in simulation and reality (HRO1 & HRO4). • Increase of users’ acceptance and confidence towards an automated vehicle (HRO1 & HRO4). • Increase of acceptance and capacity of users in mixed flows, along with the behavioural adaptation (including mimicking and conspicuity issues of automated cars and VRUs, also when facing critical manoeuvres) (HRO5 & HRO10) • Increase of acceptance of automation and behaviour/reaction /skills of drivers with regards to HMI (HRO17). • Increase of acceptance considering also age, gender, IT literacy, socioeconomic factors and understanding of automation for all cohorts (HRO1).
May 2020 82
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Title: Training for road transport
• Increase of acceptance and operation capacity of the Traffic Management Centre (TMC) operators towards autonomous vehicles and mixed flow (HRO1). • Improvement of interaction between the AVs and other road users, which will contribute to increased acceptance (HRO4). • Improvement of vigilance and complacency issues (HRO6). • Improvement of driver-readiness in transitions between manual and automated driving (HRO7). Relevant Pilot Sites • Oslo, Norway (RO-1) • Karlsruhe, Germany (RO-2) • Versailles, France (RO-3A) • Versailles, France (RO-3B) • Warsaw, Poland (RO-4) • Vienna, Austria (RO-5) • Brussels/Gent, Belgium (RO-6) • Rome, Italy (RO-7) • Linköping, Sweden (RO-8)
7.1.1.2 Analysis per Pilot site Oslo, Norway (RO-1) – Responsible Partner: TOI
Use Case title: Training for road transport
Instantiated research Improvement of interactions between the AV-shuttle and other road hypotheses (see separate file users, due to HMI/training combination, which will contribute to with research priorities from increased acceptance (HRO4 & HRO17). DoA and/or add new ones or revise) Indicative step-wise scenario(s) • Initial testing (Phase II) of WP4 training by video recordings will be (please descried the steps/ conducted at locations where safety relevant situations already phases that your pilot will follow identified in Phase I are likely to happen, and all interactions with in relation to this specific Use cyclists (and where relevant also other road users). Case) • In Phase III, a final testing and evaluation of the training will be conducted. The location will depend on where the automated shuttle bus will be, at that time, operating. Roadside interviews using the same questionnaire as in Phase I or video recordings using the same set-up as in Phase I/II will be conducted before and after the implementation of the HMI/training combination. Which of these two methods will be selected is dependent on and needs to be tailored to the situation at hand in 2021 (route, relevant locations). Participating actors (primary - • VRUs (Cyclists) secondary) • Other road users when relevant Type, model and number of AV shuttle bus (Navya Arma) vehicles SAE Level(s) (per vehicle) 3-4 – steward on board Environmental conditions (e.g. Public road, route length is 1.4 kilometre (Ormøya), mixed traffic, weather conditions) urban environment
May 2020 83
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Training for road transport
Traffic context (urban – mixed Urban, mixed traffic (cyclists, pedestrians, cars, buses) flow, rural, …) Traffic density (low – medium - Low-medium high) Speed range (per vehicle) 8-18 km/h Specific digital infrastructure to Fixed time-table (part of Ruterbillett app) be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None. Fully autonomous based on on-board sensors. requirements (e.g. 5G, G5, IoT) Training requirements and user The HMI/training combination is aimed at informing, instructing other skills road users interacting with the shuttle to overtake at safe locations and how to overtake the shuttle. Key Risks (technical, • Implementation of the HMI/training combination into existing operational, legal, behavioural) display in the shuttle might not be feasible. In that case a tailor made solution needs to be developed. • COVID-19 logistic impact and delays. Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. AV Conspicuity HMI& strategies for interaction of automated road related to optimal use of road vehicles with non-equipped other road users HMI, common part of training of other modes, etc.) Future extensions/ The HMI/training combination can be used in similar AV-shuttles transferability driving elsewhere; if needed adapted to local route Relevant KPIs • KPI.9: Number of involved vulnerable road users (pedestrians, cyclists, elderly, children) in accidents. Target: In spite of their potential enhanced vulnerability in automated traffic flows, keep current numbers at least at today’s levels (no VRUs accidents enhancement). • KPI.21: User's perception of security levels while riding the vehicle (without operators on board). Target: Overall mean perceived level of security increased after hands-on experience. Starting point & target HMI/training combination aimed at safe overtaking has not been innovation/added value implemented and evaluated before
Karlsruhe, Germany (RO-2) – Responsible Partner: FZI
Use Case title: Training for road transport users (VRUs for the interaction with automated road vehicles) Instantiated research -Acceptance of other vehicles’ drivers, passengers and VRUs (HRO4). hypotheses (see separate file -Training and dissemination with multi-platform tools for VR/AR with research priorities from simulation, WoZ and simulator scenarios for public acceptance and DoA and/or add new ones or expectations (HRO14). revise) Indicative step-wise scenario(s) • Person immerses into simulation as VRU (pedestrian) using VR (please descried the steps/ headset. phases that your pilot will follow
May 2020 84
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Training for road transport users (VRUs for the interaction with automated road vehicles) in relation to this specific Use • Person is demonstrated the behaviour of automated vehicles in Case) interactive scenarios. • Training is supposed to increase pedestrian’s acceptance and confidence towards automated vehicles. Participating actors (primary - Automated vehicle(s) and VRU (pedestrian). secondary) Type, model and number of • 1x Modified Audi Q5 vehicles • 1x Modified Smart electric drive SAE Level(s) (per vehicle) SAE Level 3 (automated vehicles) Environmental conditions (e.g. Majority of experiments done at good weather conditions (as they are weather conditions) done in simulation using AR/VR environment, weather conditions are highly variable and can be changed) Traffic context (urban – mixed Urban flow, rural, …) Traffic density (low – medium - Low high) Speed range (per vehicle) 0 – 50 km/h (all vehicles) Specific digital infrastructure to Automated vehicles drive on high definition digital maps (lanelets) be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None required requirements (e.g. 5G, G5, IoT) Training requirements and user None specific skills Key Risks (technical, Training is done in simulation, therefore no risks. operational, legal, behavioural) Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. Operators-based HMI& strategies for road transport (resulting related to optimal use of road interaction strategy, i.e. AV behaviour model is evaluated in training) HMI, common part of training of other modes, etc.) Future extensions/ Optionally create and publish dataset of virtually created, highly transferability interactive scenarios consisting of vehicle and VRU trajectories Relevant KPIs • KPI.1: User acceptance rating on UAS scale. Target: Overall mean user acceptance above 6 after the Pilots and overall mean user acceptance above 7 for VRUs (as redefined in the scope of the project); • KPI.3: User acceptance after hands-on experience of AVs (Conflicts between automated vehicles and other traffic participants). Target: User acceptance after hands-on experience to increase by 50%. (Level of conflicts to be reduced by 50% on average) • KPI.10: User opinion/rating of AVs. Target: Users’ view of automated functions after Pilots to be much closer to actual performance than before. Starting point & target Increase pedestrian’s acceptance and confidence towards an innovation/added value automated vehicle.
May 2020 85
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Versailles, France (RO-3A) – Responsible Partner: VEDECOM
Use Case title: Training for road transport
Instantiated research • Acceptance after hands-on experience of all levels of automation in hypotheses (see separate file urban, rural, highway and specific applications, such as tunnels, with research priorities from constructions and bridges, and environmental conditions (HRO1). DoA and/or add new ones or • Increase of vigilance and complacency in Level 3 and Level 4 (HRO6). revise) • Driver-Readiness in transitions between manual and automated driving (HRO7). • Training and dissemination with multi-platform tools for VR/AR simulation, WoZ and simulator scenarios for public acceptance and expectations (HRO14). Indicative step-wise scenario(s) • Participants are assigned to one experimental group (with (please descried the steps/ training/without training). phases that your pilot will follow • According to the group, they receive the training protocol. in relation to this specific Use • Then, all participants will be asked to drive on public road and to Case) activate the automated driving mode when available. Participating actors (primary - • Participants: experienced, non-professional drivers secondary) • Experimenters: human factor researchers & expert driver Type, model and number of Main vehicle: Renault Zoe (Wizard of Oz). For security reasons, a vehicles following vehicle could be involved. SAE Level(s) (per vehicle) During manual driving: SAE level 1 During automated driving: SAE level 3 (simulated by the wizard) Environmental conditions (e.g. Depending on the local weather (except in case of low-visibility during weather conditions) heavy rain/fog) Traffic context (urban – mixed Urban & peri-urban public roads flow, rural, …) Traffic density (low – medium - Low – medium. For safety reasons rush hours will be excluded high)
Speed range (per vehicle) 0 to 50 km/h
Specific digital infrastructure to None required due to the WoZ nature. be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None required due to the WoZ nature. requirements (e.g. 5G, G5, IoT) Training requirements and user Non-professional drivers with a valid driver’s license for a minimum of skills 3 years and regular driving experience (at least 2-3 days/week). Some of the participants will receive a training before driving, while the others will not. Key Risks (technical, • Any kind of road hazard & a possible confusion due to the simulated operational, legal, behavioural) automated driving conditions. To minimize these risks, the expert driver will be an experienced driver in charge of the security on board of the vehicle. • COVID-19 logistic impact and delays. Liability and operational issues VEDECOM already obtained the authorization from the French (if any) Government for the WoZ to go on public roads. Connected Use Cases (e.g. In vehicle HMI & strategies for automated road vehicles. related to optimal use of road May 2020 86
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Training for road transport
HMI, common part of training of other modes, etc.) Future extensions/ Based on the developed training protocol for non-professional drivers, transferability the protocol may be adapted to train professional drivers Relevant KPIs KPI.1: User acceptance rating on UAS scale. Target: Overall mean user acceptance above 6 after the Pilots and overall mean user acceptance above 7 for VRUs (as redefined in the scope of the project). Starting point & target This experiment builds up on the outcomes of a previous experiment innovation/added value carried out as part of the AutoConduct project by VED, which tested an earlier version of the training protocol.
Versailles, France (RO-3B) – Responsible Partner: IFSTTAR
Use Case title: Training for road transport
Instantiated research • Impact of training to drivers’ behaviour (HRO1). hypotheses (see separate file • Objective assessment of AV passengers/ “driver” acceptance level with research priorities from (HRO1 & HRO4). DoA and/or add new ones or revise) Indicative step-wise scenario(s) i) Drivers will be invited to travel on-board an electric vehicle (please descried the steps/ performing automated driving on a test track (No traffic or other users phases that your pilot will follow are managed in this pilot). ii) Drivers’ attention will be focused to other in relation to this specific Use activities than driving (ex. reading, question-answer or other attention Case) load). iii) The AV will carry out automated driving. iv) Face to a potential obstacle collision (i.e. static obstacle on the test track), the AV will initiate an automated obstacle avoidance manoeuvre execution. v) Any effect or reaction on drivers is expected to be observable through in- car video recordings during such a situation. Participating actors (primary - Drivers (experiment participants), researcher (potential) as a secondary) passenger. Type, model and number of Experimental electric vehicle platform (Renault ZOE), 1 vehicle vehicles integrating prototyped automated driving functions (x-by-wire). Non- approved vehicle. SAE Level(s) (per vehicle) Emulated SAE 4 level Environmental conditions (e.g. Non-adverse weather is required. Day-light conditions. weather conditions) Traffic context (urban – mixed None, due to test track context flow, rural, …) Traffic density (low – medium - No traffic high) Speed range (per vehicle) lower than 70 Km/h Specific digital infrastructure to None required be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure Test track and instrumented vehicle. Precise localization (base station). requirements (e.g. 5G, G5, IoT) Clear GPS visibility.
May 2020 87
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Training for road transport
Training requirements and user Subject to Ethical approval and directives : non-professional drivers or skills professional driver Key Risks (technical, • Ethical approval, technical dysfunction. operational, legal, behavioural) • COVID19 logistic impact and delays. Liability and operational issues • Ethical and technical issues. (if any) • GDPR issues for in-car video recordings. Connected Use Cases (e.g. Training for road transport related to optimal use of road HMI, common part of training of other modes, etc.) Future extensions/ Can be later extended to real road conditions through a FoT initiative. transferability Relevant KPIs KPI.3: User acceptance after hands-on experience of AVs (Conflicts between automated vehicles and other traffic participants). Target: User acceptance after hands-on experience to increase by 50%. (Level of conflicts to be reduced by 50% on average). Starting point & target Meaningful data for studying and characterizing drivers’ acceptance innovation/added value and reactions face objective assessment to automated functions in a secured and controlled full-scale environment. In-car video recordings can then be used for demonstrating the relevance of AV functions.
Warsaw, Poland (RO-4) – Responsible Partner: PZM
Use Case title: Training for road transport
Instantiated research • Increase acceptance and/or better operation of AVs by the trained hypotheses (see separate file drivers (HRO1). with research priorities from • Dependence of acceptance on age, gender, IT literacy, DoA and/or add new ones or socioeconomic factors and understanding of automation (HRO2). revise) Indicative step-wise scenario(s) Questionnaire to check the current acceptance level, test in vehicles (please descried the steps/ with wearables, focus group discussion, after few months: training, test phases that your pilot will follow in vehicles, focus group discussion, questionnaire to check the final in relation to this specific Use acceptance level. Case) Participating actors (primary - • Primary: 20 drivers. secondary) • Secondary: 40 passengers, public authorities, interested institutes, universities, media. Type, model and number of 1 x Volvo L2, 1 x Volvo L3, 1 x vehicle L4 vehicles SAE Level(s) (per vehicle) 1 x L2, 1 x L3, 1 x L4 Environmental conditions (e.g. Good weather weather conditions) Traffic context (urban – mixed Urban context flow, rural, …) Traffic density (low – medium - Medium for L2, High for L3, Low for L4. high)
May 2020 88
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Training for road transport
Speed range (per vehicle) Low speed Specific digital infrastructure to Digital mapping for L4 vehicle be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure Still under investigation. requirements (e.g. 5G, G5, IoT) Training requirements and user Training of the 20 drivers between the first test in vehicles and second skills test in vehicles Key Risks (technical, • Operational: lack of L4 vehicles for tests; bad weather on the operational, legal, behavioural) weekend of the tests; COVID-19 measures still in place during the time of the tests (might be necessary to resign from passengers and plan for disinfections). • Technical: lack of digital mapping for L4 vehicle. • Legal: lack of the permit to conduct the L4 test. • Behavioural: how will the drivers without training behave while testing the vehicles. Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. In vehicle HMI & strategies for automated road vehicles related to optimal use of road HMI, common part of training of other modes, etc.) Future extensions/ Investigation of major influences on the acceptance of AVs. Training for transferability drivers can be used in other contexts. Relevant KPIs • KPI.1: User acceptance rating on UAS scale. Target: Overall mean user acceptance above 6 on UAS scale after the Pilot. • KPI.3: User acceptance after hands-on experience of AVs (Conflicts between automated vehicles and other traffic participants). Target: User acceptance after hands-on experience to increase by 50%. • KPI.4: Comparative WTH/WTP before/after the pilots. Target: Positive Comparative assessment before/after the pilot and WTH/WTP enhancement. • KPI.24: Assessment of comfort and stress levels of users by the use of wearables during AV manoeuvres (includes mainly bio signal capturing and camera sensors). Target: Increased comfort level and reduced stress level of users after Phase II. Starting point & target Test-weekend are being prepared, which is supposed to take place in innovation/added value September. To this day, no such demos have been organised in Poland.
Vienna, Austria (RO-5) – Responsible Partners: AIT, WL
Use Case title: Training for road transport
Instantiated research • Acceptance after hands-on experience of all levels of automation in hypotheses (see separate file urban, rural, highway and specific applications, such as tunnels, with research priorities from constructions and bridges, and environmental conditions (i.e. co- DoA and/or add new ones or pilot for adverse weather, unknown environments, unknown type of revise) vehicle, etc.) (HRO1).
May 2020 89
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Training for road transport
• Dependence of acceptance on age, gender, IT literacy, socioeconomic factors and understanding of automation (HRO2). Indicative step-wise scenario(s) • Assessing experience of 200 users with existing HMI options and (please descried the steps/ operation experience for identification of pros and cons. phases that your pilot will follow • Selection of good practices. in relation to this specific Use • Conduct of an experiment to investigate new forms of awareness and Case) intent communication based on advanced automated classification of other road users. • Further observation of the bus, where reactions of the participants and the local population will be analysed. Participating actors (primary - • Operators secondary) • Passengers • Other road users • Local population Type, model and number of NAVYA autonomous shuttle vehicles SAE Level(s) (per vehicle) 4 Environmental conditions (e.g. Mild central European weather conditions) Traffic context (urban – mixed Urban, open road flow, rural, …) Traffic density (low – medium - Low to medium high) Speed range (per vehicle) Up to 12 km/h Specific digital infrastructure to Vehicle tracing via website be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure GPS correction antenna requirements (e.g. 5G, G5, IoT) Training requirements and user • Driver License for Bus (D). skills • Operator training documents provided by NAVYA. Key Risks (technical, • Acceptance of both users and local population. operational, legal, behavioural) • Current state of technology does not fulfil the requirements. • Liability in case of accidents. • COVID-19 logistic impact and delays. Liability and operational issues 1 minor accident so far, no severe consequences (if any) Connected Use Cases (e.g. Operators-based HMI& strategies for road transport related to optimal use of road HMI, common part of training of other modes, etc.) Future extensions/ Currently being debated transferability Relevant KPIs KPI.1: Enhance AV user and VRUs acceptance thanks to Pilots. Target: Overall mean user acceptance above 6 after the Pilots and overall mean user acceptance above 7 for VRUs. Starting point & target • More efficient public transport (long-term goal) innovation/added value • User acceptance and public acceptance
May 2020 90
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Brussels/ Gent, Belgium (RO-6) – Responsible Partners: VUB, VIAS
Use Case title: Training for road transport
Instantiated research Increase of acceptance considering age, gender, IT literacy, hypotheses (see separate file socioeconomic factors and understanding of automation for all cohorts with research priorities from by Kansei/Citarasa methodologies (HRO2). DoA and/or add new ones or revise) Indicative step-wise scenario(s) An autonomous shuttle will serve as a first/last mile feeder at both (please descried the steps/ locations. In order to assess passenger acceptance, user surveys will be phases that your pilot will follow conducted at the location through an online survey. in relation to this specific Use Case) Participating actors (primary - • Passengers of the shuttle secondary) • Other road users Type, model and number of • Easymile EZ10, autonomous shuttle – 1 vehicle (Brussels pilot) vehicles • Local Motors Olli, autonomous shuttle – 1 vehicle (Gent pilot) SAE Level(s) (per vehicle) 4 Environmental conditions (e.g. Moderate continental climate. weather conditions) Traffic context (urban – mixed • Urban – mixed traffic on private and public road (Brussels pilot) flow, rural, …) • Urban – mixed traffic on private road (Gent pilot) Traffic density (low – medium - • High (Brussels pilot) high) • Low (Gent pilot) Speed range (per vehicle) 20km/h Specific digital infrastructure to The survey will be performed through the means of a tablet. be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure 4G requirements (e.g. 5G, G5, IoT) Training requirements and user No skills are required. Survey will be completed after the ride in the skills shuttle. Key Risks (technical, • Any technical issues connected to the direct functioning of the operational, legal, behavioural) shuttle. • COVID-19 logistic impact and delays. Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. AV conspicuity HMI & strategies for interaction of automated road related to optimal use of road vehicles with non-equipped other road users. HMI, common part of training of other modes, etc.) Future extensions/ At the Brussels location, the pilot might grow into a service depending transferability on whether the hospital would choose to extend it. At the Gent location, the pilot is part of a service and is intended to stay in place. Relevant KPIs • KPI.1: User acceptance rating on UAS scale. Target: Overall mean user acceptance above 6 after the Pilots and overall mean user acceptance above 7 for VRUs. • KPI.3: User acceptance after hands on experience of AVs. Target: User acceptance after hands-on experience to increase by 50%.
May 2020 91
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Training for road transport
• KPI.4: Comparative WTH/WTP before/after the pilots. Target: Positive Comparative assessment before/after the pilots and WTH/WTP enhancement. • KPI.22: User and public perception of privacy. Target: Overall mean perceived level of control unchanged after hands-on experience • KPI.24: Assessment of comfort and stress levels of users by the use of wearables during AV manoeuvres (includes mainly bio signal capturing and camera sensors). Target: Increased comfort level and reduced stress level of users after Phase II Starting point & target Stimulate users to using public transport combined with first/last mile innovation/added value feeder.
Rome, Italy (RO-7) – Responsible Partner: SWM
Use Case title: Training for road transport
Instantiated research • Increase of acceptance after hands-on experience of all levels of hypotheses (see separate file automation in urban, rural, highway and specific applications, such as with research priorities from tunnels, constructions and bridges, and environmental conditions DoA and/or add new ones or (i.e. co-pilot for adverse weather, unknown environments, unknown revise) type of vehicle, etc.) (HRO1). • Behaviour adaptation (“mimicking”, “flocking”) of non-equipped vehicles (HRO10). • Impact of mixed and automated flows to traffic flow (micro/macro) simulation, incl. big data analytics for scaling (HRO12). • Understanding of liability and operational issues per automation level and user cluster (HRO15). Indicative step-wise scenario(s) • Presentation of Levels of Automation. (please descried the steps/ • Presentation of the Infrastructure Roadmap supporting the Road phases that your pilot will follow Automation. in relation to this specific Use • Log-in on the TMC platform. Case) • Training of TMC’s to the monitoring of C-ITS services (Dashboard). • Presentation and individuation of benefits for traffic management. Participating actors (primary - • 5 TMC operators secondary) Type, model and number of Car sharing vehicles vehicles SAE Level(s) (per vehicle) SAE level 1 functions to be evaluated for Car sharing vehicles Environmental conditions (e.g. Normal weather conditions weather conditions) Traffic context (urban – mixed Mixed flows (Simulated) flow, rural, …) Traffic density (low – medium - In early stages the pilot can be performed under low traffic density high) conditions, to continue with a medium traffic density depending on the systems improvements. Speed range (per vehicle) (10 – 50) Km/h
May 2020 92
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Training for road transport
Specific digital infrastructure to Deployment of C-ITS services and deployment of a dashboard for real- be deployed (e.g. digital maps, time monitoring of the services, as well as the configuration of the C- booking applications, etc.) ITS APP for in-vehicle use on after-market devices Physical Infrastructure ITS infrastructure to support the extension of V2X services, Cellular requirements (e.g. 5G, G5, IoT) communication infrastructure Training requirements and user No specific user skills are requested; nevertheless having traffic skills management skills can support in understanding the application benefits Key Risks (technical, Any technical issues connected to the direct functioning of the operational, legal, behavioural) platform. Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. • Operators-based HMI & strategies for road transport related to optimal use of road • In vehicle HMI & strategies for automated road vehicles HMI, common part of training of other modes, etc.) Future extensions/ At TMC side, this type of service represents a first step towards road transferability automation, as static information (i.e. intersection topologies) and dynamic information (i.e. traffic signal phases) are being digitalized and shared real-time with the vehicles. The C-ITS services used for the training could be extended to the whole road network, training activities could be extended to a wider audience at road operator’s side. Relevant KPIs KPI.1: Enhance AV user and VRUs acceptance thanks to Pilots. Target: Overall mean user acceptance above 6 after the Pilots and overall mean user acceptance above 7 for VRUs. Starting point & target Assess the acceptance and operation capacity of the Traffic innovation/added value Management Centre (TMC) operators, as well as enabling the mind-set change towards autonomous vehicles and mixed flows.
Linköping, Sweden (RO-8) – Responsible Partner: VTI
Use Case title: Training for road transport
Instantiated research Training and dissemination with multi-platform tools for VR simulation hypotheses (see separate file (HRO14). with research priorities from DoA and/or add new ones or revise) Indicative step-wise scenario(s) • VR-based training scenario is developed by VTI with the selected HMI (please descried the steps/ from phase II. phases that your pilot will follow • Bus driver and passengers perform training scenario with automated in relation to this specific Use docking at bus stop. Case) o The bus driver drives along the route. o The bus approaches the bus stop. o The bus “asks” the driver to confirm giving over the control. o The passenger is informed that the bus is automated (how is one of the HMI questions).
May 2020 93
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Training for road transport
o The bus approach and stop. (both bus driver and bus passenger are now in the same environment) o The passenger enters the bus. o The bus asks the driver to confirm it is ok to leave the bus stop. o The bus asks the driver to take back control. o The bus driver accepts. o The bus gives back control (only if the driver is attentive and alert enough) • Evaluation of the VR-based training using questionnaires. Participating actors (primary - • Bus drivers secondary) • Bus passengers Type, model and number of VR based vehicles One bus with Automated docking at bus stop SAE Level(s) (per vehicle) SAE 3-4 (VR) Environmental conditions (e.g. No extremes. Cloud, good vision. weather conditions) Traffic context (urban – mixed Urban – mixed traffic flow, rural, …) Traffic density (low – medium - Low - Medium high) Speed range (per vehicle) 30-40 km/h Specific digital infrastructure to None required be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None required requirements (e.g. 5G, G5, IoT) Training requirements and user VR tool usage, bus driver competence for bus drivers might be needed. skills Key Risks (technical, • Problems with one of the two VR systems. operational, legal, behavioural) • Problems with lagging. Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. In vehicle HMI & strategies for automated road vehicles. related to optimal use of road HMI, common part of training of other modes, etc.) Future extensions/ With help of VR simulations, the insight and understanding of different transferability stakeholders’ perspective will be clearer and hence provide input to future developments of training but also on HMI solutions. The training with VR might be a starting point for training before going into real road. Relevant KPIs • KPI.1: User acceptance rating on UAS scale. Target: Overall mean user acceptance above 6 after the Pilots and overall mean user acceptance above 7 for VRUs. • KPI.2: Vehicle operators’ acceptance on UAS scale. Target: Mean vehicle operators’ acceptance above 7. • KPI.3 User acceptance after hands-on experience of AVs. Target: User acceptance after hands-on experience to increase by 50%.
May 2020 94
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Training for road transport
• KPI.4: Comparative WTH/WTP before/after the pilots. Target: Positive Comparative assessment before/after the pilots and WTH/WTP enhancement. • KPI.10 User opinion/rating of AVs. Users’ view of automated functions after Pilots to be much closer to actual performance than before. Target: Users’ view of automated functions after Pilots to be much closer to actual performance than before. • KPI.14: Consumer willingness to have and to pay for autonomous vehicles. Positive Comparative CEA from Pilots (for most modes/Pilots) and automated vehicle operators. Target: Positive Comparative CEA from Pilots (for most modes/Pilots) and automated vehicle operators. • KPI.21: User's perception of security levels while riding the vehicle (without operators on board). Target: Overall mean perceived level of security increased after hands-on experience. • KPI.22: User and public perception of privacy. Target: Overall mean perceived level of control unchanged after hands-on experience. • KPI.23: Improved User Experience of HMIs in pilots after phase II (user experience refers to a persuasive, affective, trusted and personalized HMI). Target: Improved User Experience of HMIs in pilots after Phase II. Starting point & target Insight and understanding of involved stakeholders’ perspective during innovation/ automation. added value
6.1.1 Training for rail transport 6.1.1.1 General description Title: Training for rail transport
Relevant WP/Activity WP4, WP5/ A1.7 Short description (output - This Use Case focuses on the training of users described below, objectives) concerning mainly the training of train drivers and operators, in order to enhance their awareness, performance and acceptance. Target user/stakeholder • Train driver students & train drivers as pilot users clusters • Signal box operators (signallers) as pilot users • Railway researchers as trainers, observers and assessors • Railway students as pilot users Mode(s) Rail transport Type of vehicles • Train simulator of Regina type • H0 scale model network, one train, one device (RailDriver and monitors) for remote control Target GoA levels 3-4 Key Research Hypothesis • Impact on training and education, ensuring safety culture in automated operations supervision (HRA3). • Increase of operators` acceptance and overall performance due to training (HRA3). May 2020 95
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Title: Training for rail transport
Relevant Pilot Sites • Linköping, Sweden (RA-1) • Berlin, Germany (RA-2)
6.1.1.2 Analysis per Pilot site Linköping, Sweden (RA-1) – Responsible Partner: VTI
Use Case title: Training for rail transport
Instantiated research Impact on training and education, ensuring safety culture in automated hypotheses (see separate file operations supervision (HRA3). with research priorities from DoA and/or add new ones or revise) Indicative step-wise scenario(s) • Train driving in theory (please describe the steps/ • Knowledge test phases that your pilot will follow • Subjective rating of knowledge level in relation to this specific Use • Practical training in simulator Case) • Subjective rating of knowledge level Participating actors (primary - Train driver students secondary) Type, model and number of Train simulator of Regina type vehicles GoA Level(s) (per vehicle) 3-4 Environmental conditions (e.g. All weather conditions weather conditions) Traffic context (urban – mixed Train track flow, rural, …) Traffic density (low – medium - Low high) Speed range (per vehicle) 40-180 km/h
Specific digital infrastructure to Train simulator software be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure Train simulator environment requirements (e.g. 5G, G5, IoT) Training requirements and user Students attending train driving education skills Key Risks (technical, Minor risk of technical problems operational, legal, behavioural) Liability and operational issues None required, as it is a simulator study. (if any) Connected Use Cases (e.g. • Operators HMI & strategies for rail transport related to optimal use of road • In vehicle HMI & strategies for rail vehicles HMI, common part of training of other modes, etc.)
May 2020 96
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Training for rail transport
Future extensions/ Training steps of theory and practice are useful in several areas. transferability Simulators enable this for various professions. Relevant KPIs • KPI.2: Vehicle operators’ acceptance on UAS scale. Target: Mean vehicle operators’ acceptance above 7. • KPI.5: Number of accidents caused by human errors. Target: Reduce overall accidents caused by AVs as opposed by conventional vehicles by 20%. Starting point & target From limited understanding to better understanding, leading to more innovation/added value knowledge and increased safety.
Berlin, Germany (RA-2) – Responsible Partner: TUB
Use Case title: Training for rail transport
Instantiated research Assessment of the impact of training and experience on operators` hypotheses (see separate file acceptance and overall performance (HRA3). with research priorities from DoA and/or add new ones or revise) Indicative step-wise scenario(s) Different user categories (received training and did not receive training) (please descried the steps/ for each user cluster. Details to be determined based on the input from phases that your pilot will follow WP4. in relation to this specific Use Case) Participating actors (primary - • Railway researchers as trainers, observers and assessors. secondary) • Railway students as pilot users. • Train drivers as pilot users. • Signal box operators (signallers) as pilot users. Type, model and number of H0 scale model network, one train, one device (RailDriver and monitors) vehicles for remote control GoA Level(s) (per vehicle) 3 and 4 Environmental conditions (e.g. All weather conditions) Traffic context (urban – mixed Fully-protected or mixed (with level crossing) flow, rural, …) Traffic density (low – medium - Applicable to all (test scenario to be determined later) high) Speed range (per vehicle) Initial plan: Max speed of 40 km/h during manual remote-driving (to be finalized later) Specific digital infrastructure to Based on the training schemes developed within WP4 be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure Based on the training schemes developed within WP4 requirements (e.g. 5G, G5, IoT) Training requirements and user Train driving skills Key Risks (technical, Lack of experience of the user to manage new technology / steep operational, legal, behavioural) learning curve May 2020 97
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Liability and operational issues None required, as it is a simulator study. (if any) Connected Use Cases (e.g. Operators HMI & strategies for rail transport related to optimal use of road HMI, common part of training of other modes, etc.) Future extensions/ Transferable to any rail transport system (also dependent on WP4 transferability training schemes) Relevant KPIs • KPI.2: Vehicle operators’ acceptance on UAS scale. Target: Mean vehicle operators’ acceptance above 7. • KPI.5: Number of accidents caused by human errors. Target: Reduce overall accidents caused by AVs as opposed by conventional vehicles by 20%. • Other related KPIs will be decided based on the training schemes developed in WP4. Starting point & target Target is to increase the user acceptance and performance as well as innovation/added value overall system performance through training
6.1.2 Training for air transport 6.1.2.1 General description Title: Training for air transport
Relevant WP/Activity WP4, WP5/ A1.7 Short description (output - This Use Case focuses on the training of drones’ operators in order to objectives) mainly enhance their perceptions on drones’ usefulness, safety, security and privacy issues. Target user/stakeholder • Drones pilots clusters • Drones operators Mode(s) Air transport Type of vehicles Various models from DJI, Parrot and Yuneec Target automation levels According the Level of Automation Taxonomy (https://seafile.dblue.it/f/48b8a6d1ca2e4dc08c8a/) used in the aviation domain the maximum level automation of drone HMIs can be defined as follows. It is worth noticing that this classification considers not manually piloted drones: • Information acquisition A5 Full - The system integrates data coming from different sources and filters and/or highlights the information items considered relevant for the user. The criteria for integrating, filtering and highlighting the info are predefined at design level and not visible to the user • Information Analysis B5 Full - The support is offered automatically, based on parameters predefined at design level. The system triggers visual and/or aural alerts if the analysis produces results requiring attention by the user • Decision making C2 Open - The user can select one of the alternatives proposed by the system or her/his own • Action implementation D3 Medium level - By executing a sequence of actions after activation by the user. The user
May 2020 98
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Title: Training for air transport
maintains full control of the sequence and can modify or interrupt it at any time Key Research Hypothesis • Simulated behaviour training in non-standard situations (HAV1). • Impact of the new HMIs on training (HAV8). Sites to be tested • Rome, Italy (AV-1)
6.1.2.2 Analysis per Pilot site Rome, Italy (AV-1) – Responsible Partner: DBL
Use Case title: Training for air transport
Instantiated research • Simulated behaviour training in non-standard situations (HAV1). hypotheses (see separate file • Impact of the new HMIs on training (HAV8). with research priorities from DoA and/or add new ones or revise) Indicative step-wise scenario(s) Phase III of the pilot will be dedicated to evaluate the new proposed (please descried the steps/ HMI and to discuss training requirements and implications. The step- phases that your pilot will follow wise scenario will be provided in later stages of the project. in relation to this specific Use Case)
Participating actors (primary - • 10 drones’ operators secondary) • 10 pilots Type, model and number of Various models from DJI, Parrot and Yuneec vehicles Automation Level(s) (per According the Level of Automation Taxonomy vehicle) (use the drones level) (https://seafile.dblue.it/f/48b8a6d1ca2e4dc08c8a/) used in the aviation domain the maximum level automation of drone HMIs can be defined as follows. It is worth noticing that this classification considers not manually piloted drones: • Information acquisition A5 Full - The system integrates data coming from different sources and filters and/or highlights the information items considered relevant for the user. The criteria for integrating, filtering and highlighting the info are predefined at design level and not visible to the user • Information Analysis B5 Full - The support is offered automatically, based on parameters predefined at design level. The system triggers visual and/or aural alerts if the analysis produces results requiring attention by the user • Decision making C2 Open - The user can select one of the alternatives proposed by the system or her/his own • Action implementation D3 Medium level - By executing a sequence of actions after activation by the user. The user maintains full control of the sequence and can modify or interrupt it at any time Environmental conditions (e.g. Bad weather as non-standard situation weather conditions) May 2020 99
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Training for air transport
Traffic context (urban – mixed Urban and rural flow, rural, …) With and without people in the drone operations area Traffic density (low – medium - N/A high) Speed range (per vehicle) To be defined Specific digital infrastructure to None required be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None required requirements (e.g. 5G, G5, IoT)
Training requirements and user Licensed pilots and operators with almost 1 year of experience skills Key Risks (technical, • Drone pilots and operators not available operational, legal, behavioural) • Drone pilots and operators not interested in the study Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. Operators HMI& strategies for air transport related to optimal use of road HMI, common part of training of other modes, etc.) Future extensions/ Outcomes related to training impact of new HMIs could be used in transferability further studies and also transferred to other sectors planning to introduce remotely controlled systems Success thresholds KPI.2: Vehicle operators’ acceptance on UAS scale. Target: Mean (see separate file of quality vehicle operators’ acceptance above 7. indicators and add/modify) Starting point & target Starting from currently available HMIs, the pilot intends to identify pros innovation/added value and cons of what already exists and use this information to design news HMIs that overcome current problems and bridges current gaps, while at the same time saving the positive aspects of already existing HMIs. The co-design approach adopted is supposed to bring to the design of HMIs in line with users’ needs and easy to use and train.
6.1.3 Training for maritime transport 6.1.3.1 General description Title: Training for maritime transport
Relevant WP/Activity WP4, WP5/ A1.7 Short description (output - This Use Case deals with the training of automated maritime vehicles objectives) both operators and passengers and it aims to facilitate their awareness but also acceptance of such vehicles and their operation, while also to facilitate them in the transition towards the systems monitoring phase. Target user/stakeholder • Operators of AV workboat clusters • Passengers of AV workboat
May 2020 100
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Title: Training for maritime transport
Mode(s) Maritime transport Type of vehicles AV workboat Target automation levels 3-4 Key Research Hypothesis • Increase of acceptance by passengers, pilots and operators (HMA1). • Impact on operators through spectrum of automation levels and quantitative prognosis of behavioural adaptations (HMA2). • Increase the users understanding of systems operation and deskilling issues (HMA3). • Enhancement of the perception of situation awareness vs. actual system status (HMA4). • Increase of vigilance and complacency issues in transition from operator to systems monitor (HMA5). • Cost efficiency of automated vs non-automated operation in a wide range of missions (HMA6). Relevant Pilot Sites • Faaborg, Denmark (MA-1)
6.1.3.2 Analysis per Pilot site Faaborg, Denmark (MA-1) – Responsible Partner: TUCO
Use Case title: Training for maritime transport
Instantiated research • Acceptance of passengers, pilots and operators (HMA1). hypotheses (see separate file • Impact on operators through spectrum of automation levels and with research priorities from quantitative prognosis of behavioural adaptations (HMA2). DoA and/or add new ones or • Deskilling issues and decreased system understanding (HMA3). revise) • Perceived situation awareness vs. actual system status (HMA4). • Vigilance and complacency issues in transition from operator to systems monitor (HMA5). • Cost efficiency of automated vs non-automated operation in a wide range of missions (HMA6). Indicative step-wise scenario(s) Training activities to workboat operators, following the training (please descried the steps/ schemes from WP4. phases that your pilot will follow in relation to this specific Use Case) Participating actors (primary - • 8 operators secondary) • 12 passengers Type, model and number of Autonomous ProZero Workboat vehicles Automation Level(s) (per 3-4 vehicle) Environmental conditions (e.g. The weather conditions are a mixture of all-weather conditions unless weather conditions) extreme conditions are met. Such as Iced up sea. Traffic context (urban – mixed Open sea flow, rural, …) Traffic density (low – medium - Low high)
May 2020 101
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Training for maritime transport
Speed range (per vehicle) 5-40 kts Specific digital infrastructure to None required be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None required requirements (e.g. 5G, G5, IoT) Training requirements and user Automated workboats operators: operators of automated workboats skills will be trained to better overlook, control and guide automated workboats among other equipped and non-equipped boats in a safe and confident manner. Key Risks (technical, COVID-19 logistic impact and delays. operational, legal, behavioural)
Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. • Operators HMI & strategies for maritime transport related to optimal use of road • AV Conspicuity HMI & strategies for interaction of automated ships HMI, common part of training of with other non-equipped vessels. other modes, etc.) Future extensions/ The training can be transferred to Harbour and freight industries, water transferability taxies, and other short sea transporting modes. Relevant KPIs • KPI.2: Vehicle operators’ acceptance on UAS scale. Target: Mean vehicle operators’ acceptance above 7. • KPI.19: % of job loss/growth of transport-related professions. Target: At least neutral impact, if not positive, to employability. Starting point & target The current vessel is at level 4 – 5. Currently we have not identified a innovation/added value need for new training, for the operators.
6.1.4 Operators-based HMI& strategies for road transport 6.1.4.1 General description Title: Operators-based HMI& strategies for road transport
Relevant WP/Activity WP3, WP5/ A1.7 Short description (output - The main role of this Use Case is to test the HMI principles that are objectives) related to the remote operation of road automated vehicles, focusing mainly on the increase of the operators readiness for the transition between manual and automated driving, while also test the operation of automated buses in an urban area and their combination to MaaS and other “feeder” transportation means. Target user/stakeholder • Public transport operators clusters • Shuttle operators • TMC operators Mode(s) Road transport Type of vehicles • Autonomous shuttles • Traffic management centre Target SAE Levels 4
May 2020 102
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Title: Operators-based HMI& strategies for road transport
Key Research Hypothesis • Increase of driver-readiness in transitions between manual and automated driving (HRO7). • Better understanding of liability and operational issues per automation level and user cluster (HRO15). • Decrease the number of vehicles thanks to AVs in combination with MaaS. • Facilitation in understanding autonomous bus operation principles for TMC operators (HRO4). Relevant Pilot Sites • Vienna, Austria (RO-5) • Brussels/Gent, Belgium (RO-6) • Rome, Italy (RO-7)
6.1.4.2 Analysis per Pilot site Vienna, Austria (RO-5) – Responsible Partners: AIT, WL
Use Case title: Operators-based HMI& strategies for road transport
Instantiated research • Increase of driver-readiness in transitions between manual and hypotheses (see separate file automated driving (HRO7). with research priorities from • Clarification and increase of understanding of liability and DoA and/or add new ones or operational issues per automation level and user cluster (HRO15). revise) Indicative step-wise scenario(s) • Phase I, e.g. the current state of operator interfaces of automated (please descried the steps/ buses (as represented by a leading supplier of such vehicles). phases that your pilot will follow • Test of at least 2 HMI configurations in relation to this specific Use Case) Participating actors (primary - Public transport operator secondary) Type, model and number of NAVYA autonomous shuttle vehicles State of actors (e.g. Very active, interested in technology drowsy/inattentive/…) SAE Level(s) (per vehicle) 4 Environmental conditions (e.g. Mid-European weather conditions weather conditions) Traffic context (urban – mixed Urban - mixed flow flow, rural, …) Traffic density (low – medium - Low to medium high) Speed range (per vehicle) Up to 12 km/h Specific digital infrastructure to Vehicle tracing via website be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure GPS correction antenna requirements (e.g. 5G, G5, IoT) Training requirements and user • Driver License for Bus (D). skills • Operator training documents provided by NAVYA. May 2020 103
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Operators-based HMI& strategies for road transport
Key Risks (technical, • Operating personnel is not willing to use the HMIs of the bus. operational, legal, behavioural) • liability in case of accidents • COVID-19 logistic impact and delays Liability and operational issues 1 minor accident so far, no severe consequences (if any) Connected Use Cases (e.g. • Training for road transport. related to optimal use of road • AV Conspicuity HMI& strategies for interaction of automated road HMI, common part of training of vehicles with non-equipped other road users. other modes, etc.) Future extensions/ Currently under investigation transferability Relevant KPIs • KPI.2: Vehicle operators’ acceptance on UAS scale. Target: Mean vehicle operators’ acceptance above 7. • KPI13: Number of sales of autonomous vehicles. P Target: positive Comparative CEA from Pilots (for most modes/Pilots) and automated vehicle operators. • KPI17: PT reliability. Target: Acceptance, infrastructure and operators’ efficiency enhancement by at least 1 in the UAS scale before/after Pilots. • KPI.21: User's perception of security levels while riding the vehicle (without operators on board). Target: Overall mean perceived level of security increased after hands-on experience. Starting point & target • More efficient public transport (long-term goal). innovation/added value • User acceptance and public acceptance.
Brussels/ Gent, Belgium (RO-6) – Responsible Partners: VUB, VIAS
Use Case title: Operators-based HMI& strategies for road transport
Instantiated research Increase of driver-readiness in transitions between manual and hypotheses (see separate file automated driving (HRO7). with research priorities from DoA and/or add new ones or revise) Indicative step-wise scenario(s) An autonomous shuttle will serve as a first/last mile feeder at both (please descried the steps/ locations. Semi-structured Interviews with shuttle operators will be phases that your pilot will follow conducted. in relation to this specific Use Case) Participating actors (primary - Shuttle operators secondary) Type, model and number of • Easymile EZ10, autonomous shuttle – 1 vehicle (Brussels pilot) vehicles • Local Motors Olli, autonomous shuttle – 1 vehicle (Gent pilot) State of actors (e.g. Attentive (the safety operator should be attentive at all times) drowsy/inattentive/…) SAE Level(s) (per vehicle) 4 Environmental conditions (e.g. Moderate continental climate weather conditions)
May 2020 104
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Operators-based HMI& strategies for road transport
Traffic context (urban – mixed • Urban – mixed traffic on private and public road (Brussels pilot) flow, rural, …) • Urban – mixed traffic on private road (Gent pilot) Traffic density (low – medium - • High (Brussels pilot) high) • Low (Gent pilot) Speed range (per vehicle) 20km/h Specific digital infrastructure to None required be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None required requirements (e.g. 5G, G5, IoT) Training requirements and user The operator is a skilled/trained safety operator by the shuttle company skills (either EasyMile or Local Motor) Key Risks (technical, • Any technical issues connected to the direct functioning of the operational, legal, behavioural) shuttle. • COVID-19 logistic impact and delays. Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. • Training for road transport related to optimal use of road • AV Conspicuity HMI & strategies for interaction of automated road HMI, common part of training of vehicles with non-equipped other road users other modes, etc.) Future extensions/ At the Brussels location, the pilot might grow into a service depending transferability on whether the hospital would choose to extend it. At the Gent location, the pilot is part of a service and is intended to stay in place. Relevant KPIs KPI.2: Vehicle operators’ acceptance on UAS scale. Target: Mean vehicle operators’ acceptance above 7. Starting point & target Provide added value for the standard operator training. innovation/added value
Rome, Italy (RO-7) – Responsible Partner: SWM
Use Case title: Operators-based HMI & strategies for road transport
Instantiated research • Increase of acceptance after hands-on experience of all levels of hypotheses (see separate file automation in urban, rural, highway and specific applications, such as with research priorities from tunnels, constructions and bridges, and environmental conditions DoA and/or add new ones or (i.e. co-pilot for adverse weather, unknown environments, unknown revise) type of vehicle, etc.) (HRO1). • Enhancement of conspicuity of automated vehicles and the mode they operate at (automated or not) (HRO5). • Behaviour adaptation (“mimicking”, “flocking”) of non-equipped vehicles (HRO10). • Impact of mixed and automated flows to traffic flow (micro/macro) simulation, incl. big data analytics for scaling (HRO12). • Understanding of liability and operational issues per automation level and user cluster (HRO15).
May 2020 105
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Operators-based HMI & strategies for road transport
Indicative step-wise scenario(s) • TLA C-ITS Service is enabled at TMC side (please descried the steps/ • A vehicle with GLOSA app approaches an intersection and request phases that your pilot will follow SPaT/Map information in relation to this specific Use • The TMC using 3G/4G sends to the vehicle the traffic light information Case) – SPaT/MAP • TMC operator implements a traffic management strategy involving the connected intersections. • The C-ITS service is being used as an actuation channel for traffic control. Participating actors (primary - 5 TMC operators secondary) Type, model and number of Car sharing vehicle and service vehicles vehicles State of actors (e.g. Attentive drowsy/inattentive/…) SAE Level(s) (per vehicle) Car sharing (SAE level 1 functions to be evaluated) and service vehicles Environmental conditions (e.g. Normal weather conditions weather conditions) Traffic context (urban – mixed Urban with mixed flow (in terms of equipped and non-equipped flow, rural, …) vehicles) Traffic density (low – medium - In early stages the pilot can be performed under low traffic density high) conditions, to continue with a medium traffic density depending on the systems improvements. Speed range (per vehicle) (10 – 50) Km/h Specific digital infrastructure to Deployment of C-ITS services and deployment of a dashboard for real- be deployed (e.g. digital maps, time monitoring of the services, as well as the configuration of the C- booking applications, etc.) ITS APP for in-vehicle use on after-market devices. Physical Infrastructure ITS infrastructure to support the extension of V2X services, Cellular requirements (e.g. 5G, G5, IoT) communication infrastructure Training requirements and user No specific user skills are requested; nevertheless having traffic skills management skills can support in understanding the application benefits Key Risks (technical, COVID-19 logistic impact and delays. operational, legal, behavioural) Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. • Training for road transport. related to optimal use of road • In vehicle HMI & strategies for automated road vehicles. HMI, common part of training of other modes, etc.) Future extensions/ At TMC side, this type of service represents a first step towards road transferability automation, as static information (e.g. intersection topologies) and dynamic information (e.g. traffic signal phases) are being digitalized and shared real-time with the vehicles. The C-ITS services used for the training could be extended to the whole road network, training activities could be extended to a wider audience at road operator’s side. Relevant KPIs • KPI2: Vehicle operators’ acceptance on UAS scale. Target: Mean vehicle operators’ acceptance above 7.
May 2020 106
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Operators-based HMI & strategies for road transport
• KPI13: Number of sales of autonomous vehicles. Target: Positive Comparative CEA from Pilots (for most modes/Pilots) and automated vehicle operators. • KPI17: PT reliability. Target: Acceptance, infrastructure and operators’ efficiency enhancement by at least 1 in the UAS scale before/after Pilots. Starting point & target Assess the acceptance and operation capacity of the Traffic innovation/added value Management Centre (TMC) operators, as well as enabling the mind-set change towards autonomous vehicles and mixed flows.
6.1.5 AV Conspicuity HMI & strategies for interaction of automated road vehicles with non-equipped other road users 6.1.5.1 General description Title: AV Conspicuity HMI & strategies for interaction of automated road vehicles with non-equipped other road users Relevant WP/Activity WP3, WP5/ A1.7 Short description (output - This Use Case touches upon issues that deal with the interaction of objectives) automated and non-automated road vehicles with pedestrians and other automation VRUs. Through this UC, behavioural issues will be also tested that might occur either on behalf of the road users or the vehicles. Target user/stakeholder • Cyclists clusters • PtW • Pedestrians / Other road users and VRUs • Drivers/riders of automated vehicles • Drivers/riders of non-automated vehicles Mode(s) Road transport Type of vehicles • Autonomous shuttle • Level 2 cars with varying control modes and interaction strategies • Abstracted vehicles (observed by intelligent infrastructure) • PtW simulator Target SAE Levels • Levels 0, 3 ,4 Key Research Hypothesis • Adaptation of an AV’s behaviour according to observed, abstracted behaviours so that the new AV behaviour is accepted by VRU’s in simulation and reality (HRO4). • Facilitation of conspicuity of automated vehicles and the mode they operate at (automated or not) (HRO5). • Impact of AV dissemination on VRU (PTW riders) safety: the AV conspicuity issue (HRO5). • Increase of acceptance of other vehicles’ drivers, passengers and VRUs (HRO4). • Behaviour adaptation (“mimicking”, “flocking”) of non-equipped vehicles (HRO10) Relevant Pilot Sites • Oslo, Norway (RO-1) • Karlsruhe, Germany (RO-2) • Versailles, France (RO-3C)
May 2020 107
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Title: AV Conspicuity HMI & strategies for interaction of automated road vehicles with non-equipped other road users • Brussels/Gent, Belgium (RO-6)
6.1.5.2 Analysis per Pilot site Oslo, Norway (RO-1) – Responsible Partner: TOI
Use Case title: AV Conspicuity HMI strategies for interaction of automated road vehicles with non-equipped other road users Instantiated research Regarding conspicuity the readability/conspicuity of the HMI/training hypotheses (see separate file combination as displayed on the AV-shuttle will be evaluated (HRO5). with research priorities from DoA and/or add new ones or revise) Indicative step-wise scenario(s) • Initial testing (Phase II) of HMI by video recordings will be conducted (please descried the steps/ at locations where safety relevant situations identified in Phase I are phases that your pilot will follow likely to happen, and all interactions between cyclists (and where in relation to this specific Use relevant also other road users). Case) • In Phase III, a final testing and evaluation of the HMI will be conducted. The location will depend on where the automated shuttle bus is then operating. Roadside interviews using the same questionnaire as in Phase I or video recordings using the same set-up as in Phase I/II will be conducted before and after the implementation of the HMI/training combination. Which of these two methods will be selected is dependent on and needs to be tailored to the situation at hand in 2021 (route, relevant locations). Participating actors (primary - • VRUs (Cyclists) secondary) • Other road users when relevant Type, model and number of AV shuttle bus (Navya Arma) vehicles State of actors (e.g. Attentive drowsy/inattentive/…) SAE Level(s) (per vehicle) 3-4 – steward on board Environmental conditions (e.g. Public road, route length is 1.4 kilometre (Ormøya), mixed traffic, urban weather conditions) environment Traffic context (urban – mixed Urban, mixed traffic (cyclists, pedestrians, cars, buses) flow, rural, …) Traffic density (low – medium - Low-medium high) Speed range (per vehicle) 8-18 km/h Specific digital infrastructure to Fixed time-table (part of Ruterbillett app) be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None required requirements (e.g. 5G, G5, IoT) Training requirements and user The HMI/training combination is aimed at informing, instructing other skills road users interacting with the shuttle to overtake at safe locations and how to overtake the shuttle. Interaction behaviour before and after implementation is observed and acceptance survey
May 2020 108
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: AV Conspicuity HMI strategies for interaction of automated road vehicles with non-equipped other road users Key Risks (technical, • Implementation of the HMI/training combination into existing display operational, legal, behavioural) in the shuttle might not be feasible. In that case a tailor made solution needs to be developed • COVID-19 logistic impact and delays. Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. Training for road transport related to optimal use of road HMI, common part of training of other modes, etc.) Future extensions/ The HMI/training combination can be used in similar AV-shuttles driving transferability elsewhere; if needed adapted to local route. Relevant KPIs • KPI.5: Number of accidents caused by human errors. Target: Reduce overall accidents caused by AVs as opposed by conventional vehicles by 20% • KPI.6: Number of accidents caused by machine errors. Target: Keep it below the current vehicle malfunction errors. Starting point & target HMI/training combination aimed at safe overtaking has not been innovation/added value implemented and evaluated before.
Karlsruhe, Germany (RO-2) – Responsible Partner: FZI
Use Case title: AV Conspicuity HMI& strategies for interaction of automated road vehicles with non-equipped other road users Instantiated research • Acceptance of other vehicles’ drivers, passengers and VRUs (HRO4). hypotheses (see separate file • Behaviour adaptation (“mimicking”, “flocking”) of non-equipped with research priorities from vehicles (HRO10). DoA and/or add new ones or revise) Indicative step-wise scenario(s) • Observe trajectories of interacting vehicles (both automated and (please descried the steps/ non-automated) & VRUs (i.e. pedestrians) in simulation and in reality. phases that your pilot will follow • Feed observed data (abstracted trajectories) into a behaviour model. in relation to this specific Use • Create and apply behaviour model in order to adapt AV’s behaviour Case) and generate non-automated vehicle behaviour • Perform tests with immersed (in simulation) and real pedestrians. Participating actors (primary - • Primary: Automated vehicle(s) and VRU (pedestrian), secondary) • Secondary: abstracted non-automated vehicles (observed by intelligent infrastructure) Type, model and number of • 1x Modified Audi Q5 vehicles • 1x Modified Smart electric drive • Abstracted vehicles (observed by intelligent infrastructure) State of actors (e.g. Pedestrian: Attentive drowsy/inattentive/…) SAE Level(s) (per vehicle) • SAE Level 3 (automated vehicles), • SAE Level 0 (abstracted vehicles, observed by intelligent infrastructure) Environmental conditions (e.g. Majority of experiments done at good weather conditions weather conditions)
May 2020 109
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: AV Conspicuity HMI& strategies for interaction of automated road vehicles with non-equipped other road users Traffic context (urban – mixed Urban flow, rural, …) Traffic density (low – medium - Low high) Speed range (per vehicle) 0 – 50 km/h (all vehicles) Specific digital infrastructure to Automated vehicles drive on high definition digital maps (lanelets) be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure Trajectories of non-automated vehicles are observed using intelligent requirements (e.g. 5G, G5, IoT) infrastructure within the Test Area Autonomous Driving Baden Württemberg Training requirements and user None specific skills Key Risks (technical, • Most of the tests done in simulation, therefore no risks. operational, legal, behavioural) • No legal risks at observing non-automated vehicles by intelligent infrastructure (vehicle trajectories are abstracted and anonymised). • COVID-19 logistic impact and delays. Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. Training for road transport (resulting interaction strategy, i.e. AV related to optimal use of road behaviour model is evaluated in training) HMI, common part of training of other modes, etc.) Future extensions/ Optionally create and publish dataset of virtually created, highly transferability interactive scenarios consisting of vehicle and VRU trajectories Relevant KPIs • KPI.1: User acceptance rating on UAS scale. Target: Overall mean user acceptance above 6 after the Pilots and overall mean user acceptance above 7 for VRUs. • KPI.3: User acceptance after hands-on experience of AVs (Conflicts between automated vehicles and other traffic participants). Target: User acceptance after hands-on experience to increase by 50% • KPI.10: User opinion/rating of AVs. Target: Users’ view of automated functions after Pilots to be much closer to actual performance than before. Starting point & target Increase pedestrian’s acceptance and confidence towards an innovation/added value automated vehicle
Versailles, France (RO-3C) – Responsible Partner: IFSTTAR
Use Case title: AV Conspicuity HMI& strategies for interaction of automated road vehicles with non-equipped other road users Instantiated research Impact of AV dissemination on VRU (PTW riders) safety: the AV hypotheses (see separate file conspicuity issue (HRO4 & HRO5). with research priorities from DoA and/or add new ones or revise) Indicative step-wise scenario(s) 1) Riders will be recruited to participate to an experiment on the PTW simulator, 2) They will fill pre-questionnaires before the experiment, 3)
May 2020 110
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: AV Conspicuity HMI& strategies for interaction of automated road vehicles with non-equipped other road users (please descried the steps/ their equilibrium will be tested, 4) they will have a familiarisation phases that your pilot will follow journey using the PTW simulator, 4) they will make a virtual journey in relation to this specific Use using a motorway and face dense traffic situations (mixed flow) that will Case) induce them to filter. During the manoeuvre they will face sections with specifically marked AV vehicles (by mean of colour) and sections where AV and non AV vehicles are not distinguishable, 5) they will fill post questionnaires, last 6) their equilibrium will be tested. Participating actors (primary - PTW simulator’s riders, simulated vehicular traffic (mixed / only secondary) regular) Type, model and number of • 1 simulator vehicles • 100 simulated vehicles State of actors (e.g. Regular drowsy/inattentive/…) SAE Level(s) (per vehicle) sections with AV (SAE 4 or 5) simulated vehicles, sections without (regular traffic without AV functions) Environmental conditions (e.g. Day or twilight, no rain, no dense fog weather conditions) Traffic context (urban – mixed Mixed / non mixed dense flow, motorway flow, rural, …) Traffic density (low – medium - High high) Speed range (per vehicle) Stop and go situation for the cars (0 up to 30 km/h) Specific digital infrastructure to None required as it is a simulator study. be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None required as it is a simulator study. requirements (e.g. 5G, G5, IoT) Training requirements and user Habituation to the PtW simulator, standard license, filtering practice as skills a normal behaviour in dense traffic situation Key Risks (technical, COVID-19 logistic impact and delays. operational, legal, behavioural) Liability and operational issues GDPR agreement (if any) Connected Use Cases (e.g. Training for road transport related to optimal use of road HMI, common part of training of other modes, etc.) Future extensions/ Towards real-life scenarios in a FoT context transferability Relevant KPIs • KPI.1: User acceptance rating on UAS scale. Target: Overall mean user acceptance above 6 after the Pilots and overall mean user acceptance above 7 for VRU. • KPI.3: User acceptance after hands-on experience of AVs (Conflicts between automated vehicles and other traffic participants). Target: User acceptance after hands-on experience to increase by 50%. • KPI.10: User opinion/rating of AVs. Target: Users’ view of automated functions after Pilots to be much closer to actual performance than before.
May 2020 111
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: AV Conspicuity HMI& strategies for interaction of automated road vehicles with non-equipped other road users Starting point & target Scientific knowledge improvement, proposal for specific AV marking innovation/added value (visual).
Brussels/ Gent, Belgium (RO-6) – Responsible Partners: VUB, VIAS
Use Case title: AV Conspicuity HMI& strategies for interaction of automated road vehicles with non-equipped other road users Instantiated research • Acceptance of other vehicles’ drivers, passengers and VRUs (HRO4). hypotheses (see separate file • Conspicuity of automated vehicles and the mode they operate at with research priorities from (automated or not) (HRO5). DoA and/or add new ones or revise) Indicative step-wise scenario(s) An autonomous shuttle will serve as a first/last mile feeder at both (please descried the steps/ locations. In order to assess other road users and VRU interaction, user phases that your pilot will follow surveys will be conducted at the location through an online survey. in relation to this specific Use Case) Participating actors (primary - Other road users and VRUs secondary) Type, model and number of • Easymile EZ10, autonomous shuttle – 1 vehicle (Brussels pilot) vehicles • Local Motors Olli, autonomous shuttle – 1 vehicle (Gent pilot) State of actors (e.g. Combination of attentive and inattentive (normal human state when drowsy/inattentive/…) walking/biking on the road, not particularly attentive of the AV). SAE Level(s) (per vehicle) 4 Environmental conditions (e.g. Moderate continental climate weather conditions) Traffic context (urban – mixed • Urban – mixed traffic on private and public road (Brussels pilot). flow, rural, …) • Urban – mixed traffic on private road (Gent pilot). Traffic density (low – medium - • High (Brussels pilot) high) • Low (Gent pilot) Speed range (per vehicle) 20km/h Specific digital infrastructure to The survey will be performed through the means of a tablet. be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure 4G requirements (e.g. 5G, G5, IoT) Training requirements and user No skills are required. Survey will be completed after the ride in the skills shuttle. Key Risks (technical, • Any technical issues connected to the direct functioning of the shuttle. operational, legal, behavioural) • COVID-19 logistic impact and delays.
Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. Training for road transport users. related to optimal use of road HMI, common part of training of other modes, etc.)
May 2020 112
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: AV Conspicuity HMI& strategies for interaction of automated road vehicles with non-equipped other road users Future extensions/ At the Brussels location, the pilot might grow into a service depending transferability on whether the hospital would choose to extend it. At the Gent location, the pilot is part of a service and is intended to stay in place. Relevant KPIs • KPI.9: Number of involved vulnerable road users in accidents. Target: In spite of their potential enhanced vulnerability in automated traffic flows, keep current numbers at least at today’s levels (no VRUs accidents enhancement) • KPI.10: User opinion/rating of AVs. Target: Users’ view of automated functions after Pilots to be much closer to actual performance than before. Starting point & target Make other road users more aware of AVs in traffic. innovation/added value
6.1.6 In vehicle HMI & strategies for automated road vehicles 6.1.6.1 General description Title: In vehicle HMI & strategies for automated road vehicles
Relevant WP/Activity WP3, WP5/ A1.7 Short description (output - This use Case examines the HMI principles inside road automated objectives) vehicles and its purpose is mainly to test functions that have to do with the transition from non-autonomous to autonomous driving. Target user/stakeholder • Automated cars drivers / experienced, non-professional drivers clusters • Pedestrians • TMC operators • Bus drivers • Bus passengers Mode(s) Road transport Type of vehicles • Renault Zoe (Wizard of Oz). • 3 cars with different levels of automation (Level 2, 3 and 4) and connectivity will be used. • Traffic management centre. Target SAE Levels • Levels 1- 2-3- 4 Key Research Hypothesis • Increase of acceptance after hands-on experience of all levels of automation in urban, rural, highway and specific applications, such as tunnels, constructions and bridges, and environmental conditions (HRO1). • Increase of acceptance of other vehicles’ drivers, passengers and VRUs (HRO4). • Facilitation of vigilance and complacency issues in Level 3 and Level 4 (HRO6). • Increase of driver-readiness in transitions between manual and automated driving (HRO7). • Facilitation of autonomous bus operation principles for TMC operators • Dependence of AV acceptance on HMI. Which HMI strategy is the most optimal (HRO17). May 2020 113
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Title: In vehicle HMI & strategies for automated road vehicles
Relevant Pilot Sites • Versailles, France (RO-3A) • Warsaw, Poland (RO-4) • Rome, Italy (RO-7) • Linköping, Sweden (RO-8)
6.1.6.2 Analysis per Pilot site Versailles, France (RO-3A) – Responsible Partner: VEDECOM
Use Case title: In vehicle HMI & strategies for automated road vehicles
Instantiated research • Increase of acceptance after hands-on experience of all levels of hypotheses (see separate file automation in urban, rural, highway and specific applications, such as with research priorities from tunnels, constructions and bridges, and environmental conditions DoA and/or add new ones or (HRO1). revise) • Facilitation of vigilance and complacency issues in Level 3 and Level 4 (HRO6). • Increase of driver-readiness in transitions between manual and automated driving (HRO7). Indicative step-wise scenario(s) Participants are assigned to one experimental group (with (please descried the steps/ training/without training). Then, all participants will be asked to drive phases that your pilot will follow on public roads as if they had acquired/rent the vehicle for the first in relation to this specific Use time. While the current pilot does not manipulate HMI content, that is, Case) test different versions of HMI, it explores the experience and comprehension of drivers of a generic HMI as a function of receiving training or not. Participating actors (primary - • Participants: experienced, non-professional drivers. secondary) • Experimenters: human factor researchers & expert driver. Type, model and number of Main vehicle: Renault Zoe (Wizard of Oz). For security reasons, a vehicles following vehicle could be involved. State of actors (e.g. Participants may be inattentive or distracted during the automated drowsy/inattentive/…) driving phase SAE Level(s) (per vehicle) • During manual driving: SAE level 1 • During automated driving: SAE level 3 (simulated by the wizard) Environmental conditions (e.g. Depending on the local weather (except in case of low-visibility during weather conditions) heavy rain/fog) Traffic context (urban – mixed Urban & peri-urban public roads flow, rural, …) Traffic density (low – medium - Low – medium. For safety reasons rush hours will be excluded high) Speed range (per vehicle) 0 to 50 km/h Specific digital infrastructure to None required due to WoZ vehicle type. be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None required due to WoZ vehicle type. requirements (e.g. 5G, G5, IoT) Training requirements and user Non-professional drivers with a valid driver’s license for a minimum of skills 3 years and regular driving experience (at least 2-3 days/week). Some
May 2020 114
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: In vehicle HMI & strategies for automated road vehicles
of the participants will receive a training before driving, while the others will not Key Risks (technical, Malfunctioning of the HMI due to technical problems may occur. operational, legal, behavioural) Any kind of road hazard + a possible confusion due to the fake automated conditions Liability and operational issues VEDECOM already obtained the authorization from the French (if any) Government for the WOZ to go on public roads. Connected Use Cases (e.g. • Training for road transport. related to optimal use of road • AV Conspicuity HMI & strategies for interaction of automated road HMI, common part of training of vehicles with non-equipped other road users. other modes, etc.) Future extensions/ Improving the flow and identifying user’s needs in terms of transferability understandability of the HMI. Relevant KPIs KPI 23: Higher acceptance ratings and lower reaction times among the participants who received the training protocol. Target: Improved User Experience of HMIs in pilots after Phase II. Starting point & target This experiment builds up on the outcomes of a previous experiment innovation/added value carried out as part of the AutoConduct project by VED, which tested an earlier version of the training protocol.
Warsaw, Poland (RO-4) – Responsible Partner: PZM
Use Case title: In vehicle HMI & strategies for automated road vehicles
Instantiated research Dependence of AV acceptance on HMI. Which HMI strategy is the most hypotheses (see separate file optimal (HRO17). with research priorities from DoA and/or add new ones or revise) Indicative step-wise scenario(s) Questionnaire to check the current acceptance level, test in vehicles (please descried the steps/ with wearables, focus group discussion, after few months: training, test phases that your pilot will follow in vehicles, focus group discussion, questionnaire to check the final in relation to this specific Use acceptance level and opinion on HMI. Case) Participating actors (primary - • Primary: 20 drivers. secondary) • Secondary: 40 passengers, public authorities, interested institutes, universities, media. Type, model and number of 1 x Volvo L2 (model to be announced in June), 1 x Volvo L3 (model to vehicles be announced in June), 1 x TBC L4 State of actors (e.g. Attentive drowsy/inattentive/…) SAE Level(s) (per vehicle) 1 x L2, 1 x L3, 1 x L4 Environmental conditions (e.g. Good weather weather conditions) Traffic context (urban – mixed Urban context flow, rural, …) Traffic density (low – medium - Medium for L2, High for L3, Low for L4. high)
May 2020 115
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: In vehicle HMI & strategies for automated road vehicles
Speed range (per vehicle) Low speed Specific digital infrastructure to Digital mapping for L4 vehicle be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure To be confirmed in a later phase requirements (e.g. 5G, G5, IoT) Training requirements and user Training of the 20 drivers between the first test in vehicles and second skills test in vehicles Key Risks (technical, • Operational: lack of L4 vehicles for tests; bad weather on the operational, legal, behavioural) weekend of the tests; COVID-19 measures still in place during the time of the tests (might be necessary to resign from passengers and plan for disinfections). • Technical: lack of digital mapping for L4 vehicle. Legal: lack of the permit to conduct the L4 test. • Behavioural: how will the drivers without training behave while testing the vehicles. Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. • Training for road transport. related to optimal use of road • AV Conspicuity HMI & strategies for interaction of automated road HMI, common part of training of vehicles with non-equipped other road users. other modes, etc.) Future extensions/ We will find out what influences the most the acceptance of AVs. HMI transferability feedback can be used in other contexts. Relevant KPIs • KPI.4: Comparative WTH/WTP before/after the pilots. Target: Positive Comparative assessment before/after the pilots and WTH/WTP enhancement. • KPI.24: Assessment of comfort and stress levels of users by the use of wearables during AV manoeuvres. Target: Increased comfort level and reduced stress level of users after Phase II. Starting point & target We are preparing for the test-weekend which is supposed to take place innovation/added value in September. To this day, no such demos have been organised in Poland.
Rome, Italy (RO-7) – Responsible Partner: SWM
Use Case title: In vehicle HMI & strategies for automated road vehicles
Instantiated research • Increase of acceptance of other vehicles’ drivers, passengers and hypotheses (see separate file VRUs. with research priorities from • Increase of driver-readiness in transitions between manual and DoA and/or add new ones or automated driving. revise) • Transfer of expertise from rail, water, air sectors. • Behaviour adaptation (“mimicking”, “flocking”) of non-equipped vehicles. • Impact of mixed and automated flows to traffic flow (micro/macro) simulation, incl. big data analytics for scaling. Indicative step-wise scenario(s) • TLA C-ITS Service is enabled at TMC side
May 2020 116
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
(please descried the steps/ • A vehicle with GLOSA app approaches an intersection and request phases that your pilot will follow SPaT/Map information in relation to this specific Use • The TMC using 3G/4G sends to the vehicle the traffic light information Case) – SPaT/MAP • The vehicle receives the SPaT/Map and calculates the best speed to approach the traffic light • In addition, since all the time traffic lights are not green, when the traffic light is red the app also gives time to green at the intersection Participating actors (primary - 5 TMC operators secondary) Type, model and number of Car sharing vehicle and service vehicles vehicles State of actors (e.g. Attentive drowsy/inattentive/…) SAE Level(s) (per vehicle) SAE level 1 functions to be evaluated for Car sharing and service vehicles. Environmental conditions (e.g. Normal weather conditions. weather conditions) Traffic context (urban – mixed Urban with mixed flow (in terms of equipped and non-equipped flow, rural, …) vehicles) Traffic density (low – medium - In early stages the pilot can be performed under low traffic density high) conditions, to continue with a medium traffic density depending on the systems improvements. Speed range (per vehicle) (10 – 50) Km/h Specific digital infrastructure to Deployment of C-ITS services and deployment of a C-ITS APP for in- be deployed (e.g. digital maps, vehicle use on after-market devices. booking applications, etc.) Physical Infrastructure ITS infrastructure to support the extension of V2X services, Cellular requirements (e.g. 5G, G5, IoT) communication infrastructure. Training requirements and user No specific user skills are requested. skills Key Risks (technical, COVID-19 logistic impact and delays. operational, legal, behavioural) Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. Operators-based HMI and strategies for road transport. related to optimal use of road HMI, common part of training of other modes, etc.) Future extensions/ The C-ITS services used for the training could be extended to more transferability users, thus enabling a higher penetration rate of connected vehicles and therefore support the TMC in the transition towards road automation. Relevant KPIs KPI.23: Higher acceptance ratings and lower reaction times among the participants who received the training protocol. Target: Improved User Experience of HMIs in pilots after Phase II. Starting point & target Enable acceptance of the drivers on the path to road automation. innovation/added value
Linköping, Sweden (RO-8) – Responsible Partner: VTI
May 2020 117
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: In vehicle HMI & strategies for automated road vehicles
Instantiated research • Increase of acceptance of other vehicles’ drivers, passengers and hypotheses (see separate file VRUs (HRO4). with research priorities from • Increase of driver-readiness in transitions between manual and DoA and/or add new ones or automated driving (HRO7). revise) Indicative step-wise scenario(s) • Two in-vehicle HMI solutions are developed and implemented in a (please descried the steps/ VR-scenario. phases that your pilot will follow • Bus drivers and passengers perform VR driving scenario with in relation to this specific Use automated docking at bus stops: Case) o The bus driver drives along the route. o The bus approaches the bus stop. o The bus “asks” the driver to confirm giving over the control. o The passenger is informed that the bus is automated (how is one of the HMI questions). o The bus approaches and stops (both bus driver and bus passenger are now in the same environment). o The passenger enters the bus. o The bus asks the driver to confirm it is ok to leave the bus stop. o The bus asks the driver to take back control. o The bus driver accepts. o The bus gives back control (only if the driver is attentive and alert enough). • Evaluation of the two HMI solutions using acceptance and HMI questionnaires Participating actors (primary - • Bus drivers secondary) • Bus passengers Type, model and number of VR – 1 AV bus vehicles State of actors (e.g. The drivers will not be manipulated, but the instructed scenario will be drowsy/inattentive/…) that the drivers have to be attentive and alert and have at least 1 hand at steering wheel to get back the control. SAE Level(s) (per vehicle) 3-4 Environmental conditions (e.g. Normal weather without heavy rain or snow/ ice on the road. weather conditions) Traffic context (urban – mixed Urban - mixed flow, rural, …) Traffic density (low – medium - Low - medium high) Speed range (per vehicle) 40-50 km/h Specific digital infrastructure to None required be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None required requirements (e.g. 5G, G5, IoT) Training requirements and user VR tool usage, bus driver competence for bus drivers might be needed. skills Passengers NA Key Risks (technical, Motion sickness might happen. operational, legal, behavioural)
May 2020 118
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: In vehicle HMI & strategies for automated road vehicles
Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. • Training for road transport. related to optimal use of road • AV Conspicuity HMI & strategies for interaction of automated road HMI, common part of training of vehicles with non-equipped other road users. other modes, etc.) Future extensions/ The results might be possible to transfer in terms of guidelines also for transferability interactions with other type of AV vehicles (Shuttles, vessels, trains etc) Relevant KPIs • KPI.1: User acceptance rating on UAS scale. Target: Overall mean user acceptance above 6 after the Pilots and overall mean user acceptance above 7 for VRUs. • KPI.2: Vehicle operators’ acceptance on UAS scale. Target: Mean vehicle operators’ acceptance above 7. • KPI.3: User acceptance after hands-on experience of AVs (Conflicts between automated vehicles and other traffic participants). Target: User acceptance after hands-on experience to increase by 50%. • KPI4: Comparative WTH/WTP before/after the pilots. Target: Positive Comparative CEA from Pilots (for most modes/Pilots) and automated vehicle operators. • KPI.10: User opinion/rating of AVs. Target: Users’ view of automated functions after Pilots to be much closer to actual performance than before. • KPI.21: User's perception of security levels while riding the vehicle (without operators on board). Target: Overall mean perceived level of security increased after hands-on experience. • KPI.22: User and public perception of privacy. Target: Overall mean perceived level of control unchanged after hands-on experience. • KPI.23: Improved User Experience of HMIs in pilots after phase II (user experience refers to a persuasive, affective, trusted and personalized HMI). Target: Improved User Experience of HMIs in pilots after Phase II. Starting point & target Guideline for HMI during interactions with AV and passengers. innovation/added value
6.1.7 Operators HMI & strategies for rail transport 6.1.7.1 General description Title: UC7: Operators HMI & strategies for rail transport
Relevant WP/Activity WP3, WP5/ A1.7 Short description (output - This Use Case concerns the application of HMI principles to the rail objectives) vehicles of GoA 3 and 4, aiming at the optimisation of an HMI for remote control (signaller/train operator perspective). Target user/stakeholder • Signaller/train operator students clusters • Railway researchers as trainers, observers and assessors • Railway students as pilot users • Train drivers as pilot users • Signal box operators (signallers) as pilot users
May 2020 119
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Title: UC7: Operators HMI & strategies for rail transport
• Public (demonstration of the pilot) Mode(s) Rail Transport Type of vehicles • Train simulator of Regina type • H0 scale model network, one train, one device (RailDriver and monitors) for remote control Target GoA Levels 3-4 Key Research Hypothesis • Development and examination of HMI for GoA3/4 operation (signaller/train operator perspective) (HRA2). • Impact on training and education, ensuring safety culture in automated operations supervision (HRA3). Relevant Pilot Sites • Linköping, Sweden (RA-1) • Berlin, Germany (RA-2)
6.1.7.2 Analysis per Pilot site Linköping, Sweden (RA-1) – Responsible Partner: VTI
Use Case title: Operators HMI & strategies for rail transport Instantiated research • Development and examination of HMI for GoA3/4 operation hypotheses (see separate file (signaller/train operator perspective) (HRA2). with research priorities from • Impact on training and education, ensuring safety culture in DoA and/or add new ones or automated operations supervision (HRA3). revise) Indicative step-wise scenario(s) • Background theory of train driving (please descried the steps/ • Presentation of questions to discuss afterwards phases that your pilot will follow • Presentation of think-aloud video clip with train driver in relation to this specific Use • Discussion Case) • Evaluation Participating actors (primary - Signaller/train operator students secondary) Type, model and number of Train simulator of Regina type vehicles State of actors (e.g. Attentive drowsy/inattentive/…) GoA Level(s) (per vehicle) 3-4 Environmental conditions (e.g. Normal weather conditions weather conditions) Traffic context (urban – mixed Train track flow, rural, …) Traffic density (low – medium - Low high) Speed range (per vehicle) 40-180 km/h Specific digital infrastructure to Video clips be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure Computer and monitor requirements (e.g. 5G, G5, IoT)
May 2020 120
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Operators HMI & strategies for rail transport Training requirements and user Students attending signaller/train operator education skills Key Risks (technical, Minor risk of students not being motivated for discussions operational, legal, behavioural) Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. Training for rail transport related to optimal use of road HMI, common part of training of other modes, etc.) Future extensions/ The teaching method using think-aloud video clips may be useful in transferability several areas where an understanding of the view and context associated with a specific profession is necessary. Relevant KPIs KPI.5: Number of accidents caused by human errors. Target: Reduce overall accidents caused by AVs as opposed by conventional vehicles by 20%. Starting point & target From limited understanding to better understanding, leading to more innovation/added value knowledge and increased safety.
Berlin, Germany (RA-2) – Responsible Partner: TUB
Use Case title: Operators HMI & strategies for rail transport
Instantiated research • Examination and optimization of an HMI for remote control of hypotheses (see separate file GoA3/4 operation (signaller/train operator perspective). with research priorities from • Examination of suitable user groups. DoA and/or add new ones or revise) Indicative step-wise scenario(s) • Examining and benchmarking existing practices. (please descried the steps/ • Optimisation of HMIs for better user experience. phases that your pilot will follow in relation to this specific Use Case) Participating actors (primary - • Railway researchers as trainers, observers and assessors. secondary) • Railway students as pilot users. • Train drivers as pilot users. • Signal box operators (signallers) as pilot users. • Public (demonstration of the pilot). Type, model and number of H0 scale model network, one train, one device (RailDriver and monitors) vehicles for remote control State of actors (e.g. Attentive drowsy/inattentive/…) GoA Level(s) (per vehicle) 3 and 4 Environmental conditions (e.g. Clear weather-good visibility weather conditions) Traffic context (urban – mixed Fully-protected or mixed (with level crossing) flow, rural, …) Traffic density (low – medium - Applicable to all (test scenario to be determined later) high)
May 2020 121
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Operators HMI & strategies for rail transport
Speed range (per vehicle) Initial plan: Max speed of 40 km/h during manual remote-driving (to be finalised later) Specific digital infrastructure to A software to allow remote control. be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure Real-world implementations require robust communication and signal requirements (e.g. 5G, G5, IoT) transmission technology such as 5G or radio control. Training requirements and user Train driving skills skills Key Risks (technical, • Technical: Cyberattack, hardware/transmission failure operational, legal, behavioural) • Behavioural: User bias during tests (self-report), fear of unemployment for real-world implementations • Legal/operational: Standardisation Liability and operational issues None required, since it is a simulation study. (if any) Connected Use Cases (e.g. • Operators-based HMI& strategies for road transport related to optimal use of road • Operators HMI & strategies for air transport HMI, common part of training of • Operators HMI & strategies for maritime transport other modes, etc.) Future extensions/ Transferable to any rail transport system. transferability Relevant KPIs • KPI.2: Vehicle operators’ acceptance on UAS scale. Target: Mean vehicle operators’ acceptance above 7. • KPI.23: Improved User Experience of HMIs in pilots after phase II (user experience refers to a persuasive, affective, trusted and personalized HMI. Target: Improved User Experience of HMIs in pilots after Phase II. • Safe level of operator workload (threshold to be determined later). • Reliable incident management (threshold to be determined later). Starting point & target Current automated trains usually have a train attendant on-board or innovation/added value staff stationed in certain stations/facilities who are directed by dispatchers in cases of incident or emergency. Target innovation/added value is the increased resilience in the system and faster response to incidents. Additionally, maintaining safe operation and incident management with lower number of staff.
6.1.8 In vehicle HMI & strategies for rail vehicles 6.1.8.1 General description Title: In vehicle HMI & strategies for rail vehicles
Relevant WP/Activity WP3, WP5/ A1.7 Short description (output - This Use Case concerns the application of HMI principles to the rail objectives) vehicles of GoA 3 and its purpose is mainly to test functions that have to do with the transition from knowledge about in vehicle HMI (ERTMS) compared to line-side signalling (ATC)
May 2020 122
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Title: In vehicle HMI & strategies for rail vehicles
Target user/stakeholder Train drivers clusters Mode(s) Rail Transport Type of vehicles Train simulator of Regina type Target GoA Levels 3-4 Key Research Hypothesis • Development and examination of HMI for GoA3/4 operation (signaller/train operator perspective) (HRA2). • Impact on training and education, ensuring safety culture in automated operations supervision (HRA3). Relevant Pilot Sites • Linköping, Sweden (RA-1)
6.1.8.2 Analysis per Pilot site Linköping, Sweden (RA-1) – Responsible Partner: VTI
Use Case title: In vehicle HMI & strategies for rail vehicles
Instantiated research • Development and examination of HMI for GoA3/4 operation hypotheses (see separate file (signaller/train operator perspective) (HRA2). with research priorities from • Impact on training and education, ensuring safety culture in DoA and/or add new ones or automated operations supervision (HRA3). revise) Indicative step-wise scenario(s) • Familiarisation with simulator (please descried the steps/ • Initiating eye tracking phases that your pilot will follow • Driving with ATC – train protection system (line-side signalling) in relation to this specific Use • Driving with ERTMS train protection system (in vehicle HMI) Case) • Comparison of attentiveness and subjective workload with the two train protection systems Participating actors (primary - Train drivers secondary)
Type, model and number of Train simulator of Regina type vehicles State of actors (e.g. Attentive/inattentive drowsy/inattentive/…) GoA Level(s) (per vehicle) 3-4 Environmental conditions (e.g. Not applicable weather conditions) Traffic context (urban – mixed Train track flow, rural, …) Traffic density (low – medium - Low high) Speed range (per vehicle) 40-180 km/h Specific digital infrastructure to Train simulator software be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure Train simulator environment requirements (e.g. 5G, G5, IoT)
May 2020 123
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: In vehicle HMI & strategies for rail vehicles
Training requirements and user Professional train driver skills Key Risks (technical, • Minor risk of technical problems. operational, legal, behavioural) • COVID-19 logistic impact and delays. Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. • Training for rail transport. related to optimal use of road • Operators HMI & strategies for rail transport. HMI, common part of training of other modes, etc.) Future extensions/ With the switch from ATC to ERTMS follows a major need for training transferability to learn the new system. Simulators can facilitate this training. Knowledge of the effects of system change on attentiveness can be useful also in other areas. Relevant KPIs • KPI.5: Number of accidents caused by human errors. Target: Reduce overall accidents caused by AVs as opposed by conventional vehicles. • KPI.10: User opinion/rating of AVs. Target: Users’ view of automated functions after Pilots to be much closer to actual performance than before. Starting point & target Increasing the knowledge about in vehicle HMI (ERTMS) compared to innovation/added value line-side signalling (ATC).
6.1.9 Operators HMI & strategies for air transport 6.1.9.1 General description Title: Operators HMI & strategies for air transport
Relevant WP/Activity WP3, WP5/ A1.7 Short description (output - This Use Case aims to test operators’ and pilots’ perceptions of different objectives) HMIs for different levels of automation. Target user/stakeholder • Operators clusters • Pilots Mode(s) Air transport Type of vehicles Various models from DJI, Parrot and Yuneec Target automation levels According the Level of Automation Taxonomy used in the aviation domain the maximum level automation of drone HMIs can be defined as follows. It is worth noticing that this classification considers not manually piloted drones: • Information acquisition A5 Full - The system integrates data coming from different sources and filters and/or highlights the information items considered relevant for the user. The criteria for integrating, filtering and highlighting the info are predefined at design level and not visible to the user • Information Analysis B5 Full - The support is offered automatically, based on parameters predefined at design level. The system triggers visual and/or aural alerts if the analysis produces results requiring attention by the user May 2020 124
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Title: Operators HMI & strategies for air transport
• Decision making C2 Open - The user can select one of the alternatives proposed by the system or her/his own • Action implementation D3 Medium level - By executing a sequence of actions after activation by the user. The user maintains full control of the sequence and can modify or interrupt it at any time Key Research Hypothesis • Impact of adaptive HMI on drone flight planning and execution (HAV2). • Public acceptance of drones (HAV3). • Drone purpose of use correlation to its HMI (HAV4). • Risk of drone accidents (HAV5). • Situational awareness for the drone operator and the supervising controllers (HAV7). • Liability and operational issues (HAV9). Sites to be tested • Rome, Italy (AV-1)
6.1.9.2 Analysis per Pilot site Rome, Italy (AV-1) – Responsible Partner: DBL
Use Case title: Operators HMI& strategies for air transport
Instantiated research • Impact of adaptive HMI on drone flight planning and execution hypotheses (see separate file (HAV2). with research priorities from • Public acceptance of drones (HAV3). DoA and/or add new ones or • Drone purpose of use correlation to its HMI (HAV4). revise) • Risk of drone accidents (HAV5). • Situational awareness for the drone operator and the supervising controllers (HAV7). • Liability and operational issues (HAV9). Indicative step-wise scenario(s) • Drone pilots and operators will be asked to explore pros and cons of (please descried the steps/ their current HMIs with respect to the specific activity (mission) they phases that your pilot will follow have to perform. in relation to this specific Use • We will identifies current gaps and areas of improvements. Case) • New HMIs will be proposed and co-designed to overcome the identified problems and meet new opportunities that can show up. • The final version of these HMIs will be evaluated with the users in order to check impacts in terms of HF, safety, operability and liability. Participating actors (primary - • 10 drones’ operators secondary) • 10 pilots Type, model and number of Various models from DJI, Parrot and Yuneec vehicles State of actors (e.g. Vigilant and suitably trained drowsy/inattentive/…) Automation level(s) (per According the Level of Automation Taxonomy used in the aviation vehicle) domain the maximum level automation of drone HMIs can be defined as follows. It is worth noticing that this classification considers not manually piloted drones:
May 2020 125
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Operators HMI& strategies for air transport
• Information acquisition A5 Full - The system integrates data coming from different sources and filters and/or highlights the information items considered relevant for the user. The criteria for integrating, filtering and highlighting the info are predefined at design level and not visible to the user • Information Analysis B5 Full - The support is offered automatically, based on parameters predefined at design level. The system triggers visual and/or aural alerts if the analysis produces results requiring attention by the user • Decision making C2 Open - The user can select one of the alternatives proposed by the system or her/his own • Action implementation D3 Medium level - By executing a sequence of actions after activation by the user. The user maintains full control of the sequence and can modify or interrupt it at any time Environmental conditions (e.g. Bad weather as non-standard situation weather conditions) Traffic context (urban – mixed • Urban and rural flow, rural, …) • With and without people in the drone operations area Traffic density (low – medium - To be defined high) Speed range (per vehicle) To be defined Specific digital infrastructure to None required be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None required requirements (e.g. 5G, G5, IoT) Training requirements and user Licensed pilots and operators with almost 1 year of experience. skills Key Risks (technical, • Drone pilots and operators not available. operational, legal, behavioural) • Drone pilots and operators not interested in the study. Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. Training for air transport related to optimal use of road HMI, common part of training of other modes, etc.) Future extensions/ Outcomes related to HMIs could be used in further studies and also transferability transferred to other sectors planning to introduce remotely controlled systems Relevant KPIs • KPI.2: Vehicle operators’ acceptance on UAS scale. Target: Mean vehicle operators’ acceptance above 7. • KPI.10: User opinion/rating of AVs. Target: Users’ view of automated functions after Pilots to be much closer to actual performance than before. • Pilot/operator workload Starting point & target Starting from currently available HMIs, the pilot intends to identify pros innovation/added value and cons of what already exists and use this information to design news HMIs that overcome current problems and bridges current gaps, while
May 2020 126
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Operators HMI& strategies for air transport
at the same time saving the positive aspects of already existing HMIs. The co-design approach adopted is supposed to bring to the design of HMIs in line with users’ needs and easy to use and train.
6.1.10 Operators HMI & strategies for maritime transport 6.1.10.1 General description Title: Operators HMI & strategies for maritime transport
Relevant WP/Activity WP3, WP5/ A1.7 Short description (output - This Use Case aims to test project’s alternative HMIs for workboats objectives) operators as proposed by WP3 and through an iterative process to optimise and finalise them. Target user/stakeholder • Operators clusters • Passengers Mode(s) Maritime transport Type of vehicles Autonomous workboat Target automation Levels 3 & 4 Key Research Hypothesis • Increase of acceptance of pilots and operators (HMA1). • Impact on operators through spectrum of automation levels and quantitative prognosis of behavioural adaptations (HMA2). • Understanding of deskilling issues and decreased system understanding (HMA3). • Enhancement of perceived situation awareness vs. actual system status (HMA4). • Facilitation of vigilance and complacency issues in transition from operator to systems monitor (HMA5). • Cost efficiency of automated vs non-automated operation in a wide range of missions (HMA6). Relevant Pilot Sites • Faaborg, Denmark (MA-1)
6.1.10.2 Analysis per Pilot site Faaborg, Denmark (MA-1) – Responsible Partner: TUCO
Use Case title: Operators HMI & strategies for maritime transport
Instantiated research • Increase of acceptance of pilots and operators (HMA1). hypotheses (see separate file • Impact on operators through spectrum of automation levels and with research priorities from quantitative prognosis of behavioural adaptations (HMA2). DoA and/or add new ones or • Understanding of deskilling issues and decreased system revise) understanding (HMA3). • Enhancement of perceived situation awareness vs. actual system status (HMA4). • Facilitation of vigilance and complacency issues in transition from operator to systems monitor (HMA5).
May 2020 127
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Operators HMI & strategies for maritime transport
• Cost efficiency of automated vs non-automated operation in a wide range of missions (HMA6).
Indicative step-wise scenario(s) Phase II: Test (with 20 operators) of alternative HMI for workboats (please descried the steps/ operators as proposed by WP3 and through an iterative process to phases that your pilot will follow optimise and finalise them. in relation to this specific Use Case) Participating actors (primary - • 8 operators secondary) • 12 passengers Type, model and number of Autonomous ProZero Workboat vehicles State of actors (e.g. Aware drowsy/inattentive/…) Automation Level(s) (per 3-4 vehicle) Environmental conditions (e.g. The weather conditions are a mixture of all-weather conditions unless weather conditions) extreme conditions are met. Such as Iced up sea. Traffic context (urban – mixed Open sea flow, rural, …) Traffic density (low – medium - Low high) Speed range (per vehicle) 5-40 kts Specific digital infrastructure to None required be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None required requirements (e.g. 5G, G5, IoT) Training requirements and user Automated workboats operators: operators of automated workboats skills will be trained to better overlook, control and guide automated workboats among other equipped and non-equipped boats in a safe and confident manner. Key Risks (technical, • Legal risks and technology acceptance are the main risks for operational, legal, behavioural) automated transport modes, also on sea. Furthermore, we can add the need for an infrastructure which is ready to accommodate large amounts of electrical driven vessels, as these are the future. • COVID-19 logistic impact and delays. Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework. Connected Use Cases (e.g. Training for maritime transport related to optimal use of road HMI, common part of training of other modes, etc.) Future extensions/ The technology can be transferred to harbour and freight industries, transferability water taxies, and other short sea transporting modes. Relevant KPIs • KPI.2: Enhance AV vehicle operators' acceptance thanks to Pilots. Target: Mean vehicle operators’ acceptance above 7.
May 2020 128
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: Operators HMI & strategies for maritime transport
• KPI.19: Same or increased employability thanks to project innovative training schemes and material. Target: At least neutral impact, if not positive, to employability. Starting point & target HMI, such as icons or other signaling between humans (operators, innovation/added value Passengers) and the autonomous system, have not indicated that operators need new training, in order to understand the icons used in the current state of the autonomous vessel. As such, they fully understood the indicators of say, obstacles, collision paths, autonomous state of the vessel, etc.
6.1.11 AV Conspicuity HMI & strategies for interaction of automated ships with other non-equipped vessels 6.1.11.1 General description Title: AV Conspicuity HMI & strategies for interaction of automated ships with other non-equipped vessels
Relevant WP/Activity WP3, WP5/ A1.7 Short description (output - This Use Case touches upon issues that deal with the interaction of objectives) automated and non-automated vessels. Through this UC, behavioural issues will be also tested concerning operators and required behavioural adaptations. Target user/stakeholder • Operators clusters • Passengers
Mode(s) Maritime transport
Type of vehicles Autonomous Workboat
Target automation Levels Levels 3 & 4 Key Research Hypothesis • Increase of acceptance of pilots and operators. • Impact on operators through spectrum of automation levels and quantitative prognosis of behavioural adaptations. • Understanding of deskilling issues and decreased system understanding. • Enhancement of perceived situation awareness vs. actual system status. • Facilitation of vigilance and complacency issues in transition from operator to systems monitor. • Cost efficiency of automated vs non-automated operation in a wide range of missions. Relevant Pilot Sites • Faaborg, Denmark (MA-1)
6.1.11.2 Analysis per Pilot site Faaborg, Denmark (MA-1) – Responsible Partner: TUCO
May 2020 129
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: AV Conspicuity HMI& strategies for interaction of automated ships with other non-equipped vessels
Instantiated research • Increase of acceptance of pilots and operators. hypotheses (see separate file • Impact on operators through spectrum of automation levels and with research priorities from quantitative prognosis of behavioural adaptations. DoA and/or add new ones or • Understanding of deskilling issues and decreased system revise) understanding. • Enhancement of perceived situation awareness vs. actual system status. • Facilitation of vigilance and complacency issues in transition from operator to systems monitor. • Cost efficiency of automated vs non-automated operation in a wide range of missions. Indicative step-wise scenario(s) Test (with 20 operators) of alternative HMI for workboats operators as (please descried the steps/ proposed by WP3 and through an iterative process to optimise and phases that your pilot will follow finalise them. in relation to this specific Use Case) Participating actors (primary - • 8 operators secondary) • 12 passengers Type, model and number of Autonomous ProZero Workboat vehicles State of actors (e.g. Aware drowsy/inattentive/…) Automation level(s) (per 3 & 4 vehicle) Environmental conditions (e.g. The weather conditions are a mixture of all-weather conditions unless weather conditions) extreme conditions are met. Such as Iced up sea. Traffic context (urban – mixed Open sea flow, rural, …) Traffic density (low – medium - Low high) Speed range (per vehicle) 5-40 kts Specific digital infrastructure to None required be deployed (e.g. digital maps, booking applications, etc.) Physical Infrastructure None required requirements (e.g. 5G, G5, IoT) Training requirements and user Automated workboats operators: operators of automated workboats skills will be trained to better overlook, control and guide automated workboats among other equipped and non-equipped boats in a safe and confident manner. Key Risks (technical, • Legal risks and technology acceptance are the main risks for operational, legal, behavioural) automated transport modes, also on sea. Furthermore, we can add the need for an infrastructure which is ready to accommodate large amounts of electrical driven vessels, as these are the future. • COVID-19 logistic impact and delays. Liability and operational issues Pilot realisation is already according to existing local legal and (if any) operational framework.
May 2020 130
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Use Case title: AV Conspicuity HMI& strategies for interaction of automated ships with other non-equipped vessels
Connected Use Cases (e.g. Training for maritime transport related to optimal use of road HMI, common part of training of other modes, etc.) Future extensions/ The technology can be transferred to harbour and freight industries, transferability water taxies, and other short sea transporting modes. Relevant KPIs • KPI.2: Vehicle operators’ acceptance on UAS scale. Target: Mean vehicle operators’ acceptance above 7. • KPI.19: % of job loss/growth of transport-related professions. Target: At least neutral impact, if not positive, to employability. Starting point & target HMI, such as icons or other signaling between humans (operators, innovation/added value Passengers) and the autonomous system, have not indicated that operators need new training, in order to understand the icons used in the current state of the autonomous vessel. As such, they fully understood the indicators of say, obstacles, collision paths, autonomous state of the vessel, etc.
May 2020 131
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance 7 Conclusions
Current developments in automated driving require rethinking not only of terms regarding technologies used roadside or in the vehicles but also what happens with and to people indirectly connected to these developments. This Deliverable provides a common terminology and novel user clustering, including a redefinition of VRUs and a transfer of terms across modes. This is essential as the number of terms related to AVs increases and the traditional term “VRU” does not cover relevant connectivity aspects. To get insights on the opinions on AVs, a wide voice of customers’ survey, that also includes focused scenarios, has been performed. Based on the previous working steps, initial research hypotheses have been formulated. Research has been also performed on a structured transferability of AV concepts across modes and a taxonomy of future AV operators’ knowledge and skills for each mode, as basis to project’s developed training and HMI development. Detailed UCs, combining HMI concepts and training schemes towards holistic future interventions are structured to be developed and tested during the project. When planning autonomous vehicles in any transport mode (road, rail, air, maritime) it is necessary to provide a more detailed view on possible users and stakeholders in the whole process. With this approach, it is possible to provide solutions that are not only modular and inclusive, but are also in line to real-life technical and legal restrictions. Within this document, all affected user groups and stakeholders are listed and selected UCs address them accordingly. Often, when talking about autonomous vehicles and autonomous driving a common vocabulary is missing, as there are multiple implementations of different types of solutions in the same environment and constant development of technologies. Also, user preferences and fears have to be taken into consideration when it comes to the implementation of autonomous transport. Thus, relevant security and safety risks were highly ranked but also quantified (e.g. as level of accidents % to be achieved to consider automated vehicles as acceptable). Risks to employment remain a more controversial issue. It is important, that the use cases and pilot tests are wide spread over Europe and the different transport modes. Besides the often-communicated pilots with autonomous road vehicles like cars and bus shuttles, autonomous vessels are not very often seen in operation. It also shows that the rail and aviation sector both have developed various good solutions in the last years, thus paving the way to the other sector too. This document provides a blueprint when it comes to deploying autonomous transportation systems. It may be used as reference to selecting key deployment factors. It generally shows that the topic of autonomous driving in all transport is developing seamlessly, but with these developments come new questions, fears and possible bottlenecks.
May 2020 132
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance References
Alonso Raposo, M., Ciuffo, B., Makridis, M., Thiel, C. (2017): The r-evolution of driving: from Connected Vehicles to Coordinated Automated Road Transport (C-ART). Part I: Framework for a safe & efficient Coordinated Automated Road Transport (C-ART) system. Eur Comm. Andritsos, F., Perez-Prat, J. (2000): The Automation and Integration of Production Processes in Shipbuilding. For the: DG Enterprise, unit E.6. ATKearney (2016): How Automakers Can Survive the Self-Driving Era. URL: https://www.atkearney.com/documents/10192/8591837/How+Automakers+Can+Survive+the+Self- Driving+Era+%282%29.pdf/1674f48b-9da0-45e8-a970-0dfbd744cc2f Avenoso, A., Beckmann, J. (2005): The Safety of Vulnerable Road Users in the Southern, Eastern and Central European Countries. European Transport Safety Council, Brussels, 2005 AVENUE Project. (n.d.). AVENUE – Autonomous Vehicles to Evolve to a New Urban Experience : an H2020 funded project opening the way for adoption of Autonomous Vehicles in public transportation. Retrieved May 4, 2020, from https://h2020-avenue.eu/ Bieg, H. J., Daniilidou, C., Michel, B., & Sprung, A. (2020). Task load of professional drivers during level 2 and 3 automated driving. D. de Waard, A. Toffetti, L. Pietrantoni, T. Franke, JF. Petiot, C. Dumas, A. Botzer, L. Onnasch, I. Milleville, and F. Brunel (n.a.): Autonome Schiffe: Der Kapitän von morgen ist ein Hightech-Experte. URL: https://www.brunel.de/de-de/autonome-schiffe Council Directive of 25 July 1985 on the approximation of the laws, regulations and administrative provisions of the Member States concerning liability for defective products. (1985, August 7). Retrieved March 24, 2020, from https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:31985L0374&from=EN
Council of the European Union. (2019, March 28). Proposal for a Regulation of the European Parliament and of the Council on type-approval requirements for motor vehicles and their trailers, and systems, components and separate technical units intended for such vehicles, as regards their general safety and the protection of vehicle occupants and vulnerable road users, amending Regulation (EU) 2018/... and repealing Regulations (EC) No 78/2009, (EC) No 79/2009 and (EC) No 661/2009. Retrieved March 24, 2020, from https://data.consilium.europa.eu/doc/document/ST-7673-2019-INIT/en/pdf CS MARE (2019): Workshop SIMUSAFE. URL: https://www.csmare.eu/en/workshop-simusafe-autonomous- vehicles-and-road-safety Directive 2003/88/EC of the European Parliament and of the Council of 4 November 2003 concerning certain aspects of the organisation of working time. (2003, November 18). Retrieved May 4, 2020, from https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32003L0088&from=EN
Directive 2009/103/EC of the European Parliament and of the Council of 16 September 2009 relating to insurance against civil liability in respect of the use of motor vehicles, and the enforcement of the obligation to insure against such liability. (2009, October 7). Retrieved May 5, 2020, from https://eur- lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32009L0103&from=EL Deloitte (2017): Deloitte global automotive consumer studies, 2017. Deloitte (2018): Deloitte global automotive consumer studies, 2018.
May 2020 133
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Dörfelt, S., Scherf, J. (2017): Logistikjobs der Zukunft – Was sich durch Digitalisierung ändert. Artikel vom 30.08.2017 in: MMLogistik. URL: https://www.mm-logistik.vogel.de/logistikjobs-der-zukunft-was-sich- durch-digitalisierung-aendert-a-635582/ European Commission. (2018a, May 17). On the road to automated mobility: An EU strategy for mobility of the future. Retrieved March 25, 2020, from https://ec.europa.eu/transport/sites/transport/files/3rd- mobility-pack/com20180283_en.pdf
European Commission. (2018b, July 10). Europe on the Move: Commission completes its agenda for safe, clean and connected mobility. Retrieved March 25, 2020, from https://ec.europa.eu/transport/modes/road/news/2018-05-17-europe-on-the-move-3_en
European Commission (2020): Expectations and Concerns from a Connected and Automated Mobility. https://ec.europa.eu/commfrontoffice/publicopinion/index.cfm/Survey/getSurveyDetail/instruments/S PECIAL/surveyKy/2231 European Road Transport Research Advisory Council (ERTRAC). (2019a). Connected Automated Driving Roadmap. Retrieved from https://www.ertrac.org/uploads/documentsearch/id57/ERTRAC-CAD- Roadmap-2019.pdf
European Road Transport Research Advisory Council (ERTRAC). (2019b). Long Distance Freight Transport: A roadmap for System integration of Road Transport. Retrieved from https://www.ertrac.org/uploads/documentsearch/id57/ERTRAC-CAD-Roadmap-2019.pdf
European Union (2010): Directive 2010/40/EU of the European Parliament and of the Council of 7 July 2010 on the framework for the deployment of Intelligent Transport Systems in the field of road transport and for interfaces with other modes of transport. European Union (2011): White Paper on transport. Roadmap to a single European transport area - Towards a competitive and resource-efficient transport system. Retrievable at: https://ec.europa.eu/transport/themes/strategies/2011_white_paper_en Fraedrich, E., Cyganski, R., Wolf, I., Lenz, B. (2016): User Perspectives on Autonomous Driving: a Use-case- driven Study in Germany. Arbeitsberichte, 187, Geographisches Institut, Humboldt-Universität zu Berlin. Frey, T. (2017): Lokführermangel in Deutschland. Traumberuf Lokführer – das war mal. URL: https://www.swr.de/marktcheck/lokfuehrermangel-in-deutschland-traumberuf-lokfuehrer-das-war- mal//id=100834/did=20112094/nid=100834/1cdrpp9/index.html Frison, A. K., Aigner, L., Riener, A., Wintersberger P. (2017): Senior Drivers: Using the Benefits of Automated Driving for the Elderly. Mensch und Computer (MuC), pp. 6. Geneva Convention on Road Traffic. (1949, September 19). Retrieved May 5, 2020, from https://treaties.un.org/doc/Treaties/1952/03/19520326 03-36 PM/Ch_XI_B_1_2_3.pdf
Germany: Road Regulations Amended to Allow Autonomous Vehicles | Global Legal Monitor. (2016, December 15). Retrieved May 5, 2020, from http://www.loc.gov/law/foreign-news/article/germany- road-regulations-amended-to-allow-autonomous-vehicles/
Hartwich, F., Beggiato, M., & Krems, J. F. (2018). Driving comfort, enjoyment and acceptance of automated driving–effects of drivers’ age and driving style familiarity. Ergonomics, 61(8), 1017-1032. IGPmagazin (2019): Selbstfahrende Autos: Senioren als lukrative Zielgruppe. Article from Roland Losch. URL: https://www.igp-magazin.de/selbstfahrende-autos-senioren-als-lukrative-zielgruppe/
May 2020 134
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
International Transport Forum - OECD. (2017). Managing the Transition to Driverless Road Freight Transport. Retrieved from https://www.itf-oecd.org/managing-transition-driverless-road-freight-transport
IRU - International Road Union. (2019). Tackling Driver Shortage in Europe. Retrieved from https://www.ecta.com/resources/Documents/Other%20publications/Tackling%20the%20European%20 Driver%20Shortage%20-%20IRU%20report.pdf Li, T., and K. Kockelman (2016): Valuing the Safety Benefits of Connected and Automated Vehicle Technologies. Proceedings of the 95th Annual Meeting of the Transportation Research Board. Litman, T. (2015): Autonomous Vehicle Implementation Predictions: Implications for Transport Planning, Victoria Transport Policy Institute, s.l. Martek Marine (2018): Automation for Maritime: The Future of Shipping and Ports. URL: https://www.martek-marine.com/blog/automation-maritime-future/ Methorst, R. (2003): Vulnerable Road Users: New Approaches Needed? 15th ICTCT workshop. https://www.ictct.net/migrated_2014/ictct_document_nr_187_Methorst.pdf Mobility4EU (2016): D2.1 Societal needs and requirements for future transportation and mobility as well as opportunities and challenges of current solutions. Retrievable at: https://www.mobility4eu.eu/?wpdmdl=1245 Mobility4EU (2018a): Vision map 2030 for the European transport system. Retrievable at: https://www.mobility4eu.eu/?wpdmdl=2160 Mobility4EU (2018b): European Action plan for user-centric and cross modal transport in 2030. Retrievable at: https://www.mobility4eu.eu/?wpdmdl=3115 OECD (1998): Safety of vulnerable road users. DSTI/DOT/RTR/RS7(98)1/FINAL, 1998 Porsche Consulting (2018): The Future of Vertical Mobility. Sizing the market for passenger, inspection, and goods services until 2035. A Porsche Consulting Study. URL: https://fedotov.co/wp- content/uploads/2018/03/Future-of-Vertical-Mobility.pdf Regulation (EC) No 561/2006 of the European Parliament and of the Council of 15 March 2006 on the harmonisation of certain social legislation relating to road transport and amending Council Regulations (EEC) No 3821/85 and (EC) No 2135/98 and repealing Council Regulation (EEC) No 3820/85. (2006, March 15). Retrieved March 24, 2020, from https://eur-lex.europa.eu/legal- content/EN/TXT/PDF/?uri=CELEX:02006R0561-20150302&from=EN
Regulation (EU) No 165/2014 of the European Parliament and of the Council of 4 February 2014 on tachographs in road transport, repealing Council Regulation (EEC) No 3821/85 on recording equipment in road transport and amending Regulation (EC) No 561/2006 of the European Parliament and of the Council on the harmonisation of certain social legislation relating to road transport. (2014, February 28). Retrieved March 25, 2020, from https://eur-lex.europa.eu/legal- content/EN/TXT/PDF/?uri=CELEX:32014R0165&from=EN
Rice, S., Winter, S. (2019): The future of aviation? Even more automation. URL: https://www.fastcompany.com/90324699/the-future-of-aviation-even-more-automation Roland Berger (2018): Urban air mobility. The rise of a new mode of transport. URL: https://www.rolandberger.com/en/Publications/Passenger-drones-ready-for-take-off.html Safety4Sea (2019): 5 digital innovations that are changing the maritime industry. URL: https://safety4sea.com/5-digital-innovations-that-are-changing-the-maritime- industry/?__cf_chl_jschl_tk__=66507954a7e1c8d553fcd5635126a7d8fc696fe9-1589975093-0- May 2020 135
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
AW0R3SwZUyLdzYsiCy2CNFaWjLj_CAcOzp6AgcCAKHFXSfgRwq4Yi0Nu2svf7kQ5kdvKFGVcw4DhoYoE7aS Dh_0loc0mV3bIhjKeLhva5s86zB6KaxEEBFNFCUvMGu0d7o0UYpYrT9swWl- OGz8EAzoqiaGakj6nv9OuHZqAVelQIJC8ENDgMvOCX9BordeCp5B6ZQkr3Jokeyx1wX-lk1bt8yGyk0-f- thwfmICCjZF-mPH4SnlFKLanPDrfchiip2psuIOkkLh-1ky5rtiGehISyEVDkSy0hg6LyI10X- m6zyzsO5YoG461aw4K-TWXfwLXV8Eq2u72DUA5kDgXjI3Y3Uylc-bztQ9xgLaGIlF Siulagi, A., Antin, J. F., Molnar, L. J., Bai, S., Reynolds, S., Carsten, O., Greene-Roesel, R. (2016): Vulnerable Road Users: How Can Automated Vehicle Systems Help to Keep Them Safe and Mobile? In: Meyer/Beiker: Road Vehicle Automation 3. Springer. Switzerland. Société des transports intercommunaux de Bruxelles (STIB). (2019, September 20). La STIB teste des navettes autonomes. Retrieved May 5, 2020, from https://www.stib-mivb.be/article.html?l=fr&_guid=00a66cc7- 0769-3710-0e97-803f4095ace7
Sucha, M. (2014): Road Users. URL: http://psych.upol.cz/wp-content/uploads/2014/02/Presentation3_ Road_users.pdf SWOV (2012): Vulnerable road users. SWOV Fact sheet, July 2012 Trommer, S., Kolarova, V., Fraedrich, E., Kröger, L., Kickhöfer, B., Kuhnimhof, T., Lenz, B., Phleps, P. (2016): Autonomous Driving, The Impact of Vehicle Automation on Mobility Behaviour. Institute for Mobility Research. UBER Elevate (2016): Fast-Forwarding to a Future of On-Demand Urban Air Transportation. URL: https://www.uber.com/elevate.pdf United Nations Economic Commission for Europe (UNECE). (2018a, October 3). Report of the Global Forum for RoadTraffic Safety on its seventy-seventhsession. Retrieved May 4, 2020, from http://www.unece.org/fileadmin/DAM/trans/doc/2018/wp1/ECE-TRANS-WP1-165e.pdf
United Nations Economic Commission for Europe (UNECE). (2018b, October 9). UNECE adopts resolution on the deployment of highly and fully automated vehicles in road traffic. Retrieved March 25, 2020, from https://www.unece.org/info/media/presscurrent-press-h/transport/2018/unece-adopts-resolution-on- the-deployment-of-highly-and-fully-automated-vehicles-in-road-traffic/doc.html
US DOT FHWA (2019): Safety Culture and the Zero Deaths Vision. URL: https://safety.fhwa.dot.gov/zerodeaths/ Vienna Convention on Road Traffic. (1968, November 8). Retrieved May 4, 2020, from https://www.unece.org/fileadmin/DAM/trans/conventn/crt1968e.pdf
WELT (2018): An manchen Flughäfen sitzt bald kein Fluglotse mehr. Article from 7/3/18. URL: https://www.welt.de/wirtschaft/article174290604/Fluglotsen-Mit-der-Fernueberwachung-sollen- Fluglotsen-kleine-Airports-steuern.html Wolf, I. (2015): Wechselwirkung Mensch und autonomer Agent. In: Maurer, M., Gerdes, J. C., Lenz, B., Winner, H. (Eds.): Autonomes Fahren. Technische, rechtliche und gesellschaftliche Aspekte. Springer Open.
May 2020 136
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance ANNEX 1: Common terminology for automated driving in Drive2theFuture
Term Abbreviation Definition Source Active Safety ASS Active safety systems are vehicle systems that sense SAE-J3063 System and monitor conditions inside and outside the vehicle for the purpose of identifying perceived present and potential dangers to the vehicle, occupants, and/or other road users, and automatically intervene to help avoid or mitigate potential collisions via various methods, including alerts to the driver, vehicle system adjustments, and/or active control of the vehicle subsystems (brakes, throttle, suspension, etc.) Automated ADS The hardware and software that are collectively SAE-J3016 Driving System capable of performing the entire DDT on a sustained basis, regardless of whether it is limited to a specific operational design domain (ODD); this term is used specifically to describe a level 3, 4, or 5 driving automation system. ADS-dedicated ADS-DV A vehicle designed to be operated exclusively by a level SAE-J3016 vehicle 4 or level 5 ADS for all trips within its given ODD limitations (if any). [driverless An entity that dispatches an ADS-equipped vehicle(s) in SAE-J3016 operation] driverless operation. dispatching entity conventional A vehicle designed to be operated by a conventional SAE-J3016 vehicle driver during part or all of every trip. dispatch [in To place an ADS-equipped vehicle into service in SAE-J3016 driverless driverless operation by engaging the ADS. operation] driving The performance by hardware/software systems of SAE-J3016 automation part or all of the DDT on a sustained basis. driving The hardware and software that are collectively SAE-J3016 automation capable of performing part or all of the DDT on a system or sustained basis; this term is used generically to describe technology any system capable of level 1-5 driving automation. [driving A level 1-5 driving automation system’s design-specific SAE-J3016 automation functionality at a given level of driving automation system] feature within a particular ODD, if applicable. or application driver support A general term for level 1 and level 2 driving SAE-J3016 [driving automation system features. automation system] feature driverless Operation of an ADS-equipped vehicle in which either SAE-J3016 operation [of an no on-board user is present, or in which on-board users ADS-equipped are not drivers or fallback-ready users. vehicle]
May 2020 137
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Term Abbreviation Definition Source [ADS-equipped] A type of ADS-equipped vehicle designed for both SAE-J3016 dual-mode driverless operation and operation by a conventional vehicle driver for complete trips. dynamic driving DDT All of the real-time operational and tactical functions SAE-J3016 task required to operate a vehicle in on-road traffic, excluding the strategic functions such as trip scheduling and selection of destinations and waypoints [dynamic driving The response by the user to either perform the DDT or SAE-J3016 task] fallback achieve a minimal risk condition after occurrence of a DDT performance-relevant system failure(s) or upon operational design domain (ODD) exit, or the response by an ADS to achieve minimal risk condition, given the same circumstances. lateral vehicle The DDT subtask comprising the activities necessary for SAE-J3016 motion control the real-time, sustained regulation of the lateral component of vehicle motion longitudinal The DDT subtask comprising the activities necessary for SAE-J3016 vehicle motion the real-time, sustained regulation of the longitudinal control component of vehicle motion Minimal Risk MRC A condition to which a user or an ADS may bring a SAE-J3016 Condition vehicle after performing the DDT fallback in order to reduce the risk of a crash when a given trip cannot or should not be completed. [DDT A malfunction in a driving automation system and/or SAE-J3016 performance- other vehicle system that prevents the driving relevant] system automation system from reliably performing the failure portion of the DDT on a sustained basis, including the complete DDT, that it would otherwise perform. monitor the user The activities and/or automated routines designed to SAE-J3016 assess whether and to what degree the user is performing the role specified for him/her. monitor the The activities and/or automated routines that SAE-J3016 driving accomplish real-time roadway environmental object environment and event detection, recognition, classification, and response preparation (excluding actual response), as needed to operate a vehicle. monitor vehicle The activities and/or automated routines that SAE-J3016 performance [for accomplish real-time evaluation of the vehicle DDT performance, and response preparation, as needed to performance- operate a vehicle. relevant system failures] monitor driving The activities and/or automated routines for evaluating SAE-J3016 automation whether the driving automation system is performing system part or all of the DDT appropriately. performance Object and Event OEDR The subtasks of the DDT that include monitoring the SAE-J3016 Detection and driving environment (detecting, recognizing, and Response classifying objects and events and preparing to respond as needed) and executing an appropriate response to
May 2020 138
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Term Abbreviation Definition Source such objects and events (i.e., as needed to complete the DDT and/or DDT fallback). operate [a motor The activities performed by a (human) driver (with or SAE-J3016 vehicle] without support from one or more level 1 or 2 driving automation features) or by an ADS (level 3-5) to perform the entire DDT for a given vehicle during a trip. Operational ODD Operating conditions under which a given driving SAE-J3016 Design Domain automation system or feature thereof is specifically designed to function, including, but not limited to, environmental, geographical, and time-of-day restrictions, and/or the requisite presence or absence of certain traffic or roadway characteristics. receptivity [of the An aspect of consciousness characterized by a person’s SAE-J3016 user] ability to reliably and appropriately focus his/her attention in response to a stimulus. request to Notification by an ADS to a fallback-ready user SAE-J3016 intervene indicating that s/he should promptly perform the DDT fallback, which may entail resuming manual operation of the vehicle (i.e., becoming a driver again), or achieving a minimal risk condition if the vehicle is not drivable. supervise [driving The driver activities, performed while operating a SAE-J3016 automation vehicle with an engaged level 1 or 2 driving automation performance] system feature, to monitor that feature’s performance, respond to inappropriate actions taken by the feature, and to otherwise complete the DDT. sustained Performance of part or all of the DDT both between SAE-J3016 operation [of a and across external events, including responding to vehicle] external events and continuing performance of part or all of the DDT in the absence of external events. trip The traversal of an entire travel pathway by a vehicle SAE-J3016 from the point of origin to a destination. usage A particular level of driving automation within a SAE-J3016 specification particular ODD. human driver A user who performs in real-time part or all of the DDT SAE-J3016 and/or DDT fallback for a particular vehicle. conventional A driver who manually exercises in-vehicle braking, SAE-J3016 driver accelerating, steering, and transmission gear selection input devices in order to operate a vehicle. remote driver A driver who is not seated in a position to manually SAE-J3016 exercise in-vehicle braking, accelerating, steering, and transmission gear selection input devices (if any) but is able to operate the vehicle. passenger A user in a vehicle who has no role in the operation of SAE-J3016 that vehicle. [DDT] fallback- The user of a vehicle equipped with an engaged level 3 SAE-J3016 ready user ADS feature who is able to operate the vehicle and is receptive to ADS-issued requests to intervene and to evident DDT performance-relevant system failures in
May 2020 139
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Term Abbreviation Definition Source the vehicle compelling him or her to perform the DDT fallback. driverless A user(s) who dispatches an ADS-equipped vehicle(s) in SAE-J3016 operation driverless operation. dispatcher vehicle A machine designed to provide conveyance on public SAE-J3016 streets, roads, and highways. Automation Level The performance of the entire DDT by the driver, even SAE-J3016 0 (No Driving when enhanced by active safety systems Automation) Automation Level The sustained and ODD-specific execution of either the SAE-J3016 1 (Driver lateral or the longitudinal vehicle motion control Assistance) subtask of the DDT (but not both simultaneously) by a driving automation system with the expectation that the driver performs the remainder of the DDT. Automation Level The sustained and ODD-specific execution of both the SAE-J3016 2 (Partial Driving lateral and longitudinal vehicle motion control subtasks Automation) of the DDT by a driving automation system with the expectation that the driver completes the OEDR subtask and supervises the driving automation system. Automation Level The sustained and ODD-specific performance of the SAE-J3016 3 (Conditional entire DDT by an ADS with the expectation that the Driving DDT fallback-ready user is receptive to ADS-issued Automation) requests to intervene, as well as to DDT performance- relevant system failures in other vehicle systems, and will respond appropriately. Automation Level The sustained and ODD-specific performance of the SAE-J3016 4 (High Driving entire DDT and DDT fallback by an ADS , without any Automation) expectation that a user will respond to a request to intervene. Automation Level The sustained and unconditional (i.e., not ODD-specific) SAE-J3016 5 (Full Driving performance of the entire DDT and DDT fallback by an Automation) ADS without any expectation that a user will respond to a request to intervene. baseline reference to which the series of tests in a study are DINSAE- compared 91381 control factors influential variables that are kept constant within a DINSAE- series of tests 91381 derived measurement calculated from a direct measurement DINSAE- measurement (e. g. by applying mathematical or statistical 91381 operations) or a combination of one or more direct or derived measurements direct measurement logged directly from a sensor, without DINSAE- measurement further manipulations except linear transformations 91381 (e.g. m/s to kph) external measurement provided by sensors outside of the log DINSAE- measurement equipment used in the study 91381 field operational FOT study to evaluate functions or vehicles under typical DINSAE- test operating conditions in uncontrolled environments 91381 encountered by the vehicle under test
May 2020 140
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Term Abbreviation Definition Source internal measurement provided by sensors of the system under DINSAE- measurement test in the study 91381 metrics algorithm to calculate an indicator based on DINSAE- measurements applied to a concrete scenario 91381 naturalistic unobtrusive observation of human drivers in DINSAE- driving study uncontrolled test environments 91381 pilot test study to evaluate prototype functions or vehicles under DINSAE- typical operating conditions in uncontrolled 91381 environments encountered by the vehicle under test treatment phase period of testing during which the variables/system DINSAE- under study are manipulated 91381 closed testbed test environment without public traffic DINSAE- 91381 controlled test setting under which variables external to the vehicle DINSAE- environment under test are determined 91381 cyber physical representation of objects using a combination of DINSAE- model physical and virtual models for interaction with each 91381 other object fidelity quality of representation of a real-world subject's DINSAE- relevant characteristics by a virtual or physical model 91381 open testbed test environment with public traffic DINSAE- 91381 ordinary driver person untrained in the system under test or similar DINSAE- systems and whose behavior is expected to represent 91381 an average human driver physical model tangible representation of a real-world object DINSAE- 91381 uncontrolled test setting under which all variables external to the vehicle DINSAE- environment under test are not determined 91381 virtual model software representation of a real-world object DINSAE- 91381 concrete scenario parameterised model of the time sequence of scenes DINSAE- (logical scenario) which begins with an initial scene and 91381 defined point in time; the behaviour of the main actor (vehicle under test) is not further specified. corner case scenario in which two or more parameter values are DINSAE- each within the capabilities of the system, but together 91381 constitute a rare condition that challenges its capabilities driver takeover action by the driver to regain manual control of the DINSAE- vehicle 91381 edge case scenario in which the extreme values or even the very DINSAE- presence of one or more parameters results in a 91381 condition that challenges the capabilities of the system event influencing change of state or condition within a DINSAE- scenario 91381 handover controlled transition of the vehicle control from the DINSAE- system to the driver and vice versa 91381 influencing actor scenario participant that either requires the vehicle DINSAE- under test to take action or limits its action 91381
May 2020 141
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Term Abbreviation Definition Source initial condition state of the environment and vehicle under test at the DINSAE- beginning of a scenario 91381 logical scenario beginning with an initial scene, a model of the time DINSAE- sequence of scenes whose parameters are defined as 91381 ranges; at a defined point in time, the behaviour of the main actor (vehicle under test) is not further specified maneuver physical movement of an actor in a scenario DINSAE- 91381 near crash event requiring a rapid, evasive maneuver to avoid a DINSAE- collision 91381 replay scenario recorded, unmodified, real-world data representing the DINSAE- experienced test 91381 resulting state of the environment and vehicle under test at the DINSAE- condition end of a scenario 91381 safety critical SCE event with increased collision risk that might lead to a DINSAE- event near crash or a crash 91381 scenario abstraction and general description of a temporal and DINSAE- spatial traffic constellation without any specification of 91381 the parameters scene snapshot that includes the moving and non-moving DINSAE- elements of the traffic environment, the self- 91381 representation of all actors and observers and the relations between those elements supporting actor required scenario participant that does not directly DINSAE- influence the vehicle under test but limits the action of 91381 others system takeover temporary assumption of driving control by the vehicle DINSAE- 91381 vehicle under test VUT scenario participant whose behavior is of primary DINSAE- interest 91381 augmented AR an interactive experience of a real-world environment WiPed- reality simulation simulation where the objects that reside in the real world are 201910, enhanced by computer-generated perceptual modified information (from simulation) virtual reality VR an interactive experience of an environment where the WiPed- simulation simulation objects are generated by simulation 201910, modified digital twin [of digitalized version of pilot site ProjectDef pilot site] Wizard of Oz a testing or iterative design methodology wherein an WiPed- vehicle experimenter (the “wizard”) simulates the behavior of 201910, a theoretical intelligent computer application in a modified laboratory setting automatic train ATC general class of train protection systems for railways WiPed- control that involves a speed control mechanism in response to 201910 external inputs -> as the term "control" is not clearly defined (see SAE-J3016, deprecated terms), replace by term "driving automation system"
May 2020 142
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Term Abbreviation Definition Source automatic train ATO operational safety enhancement device used to help WiPed- operation automate operations of trains -> replace by "driving 201910 automation" grade of GoA synonym for automation level -> prefer to use synonym automation "automation level" Lane Change LCA The system monitors the areas to the left and right of ERTRAC- Assist the car, including the blind spot detection, and up to 50 2017 metres behind it and warns you of a potentially hazardous situation by means of flashing warning lights in the exterior mirrors. Park Distance PDC The Park Distance Control supports the driver to ERTRAC- Control manoeuvre into tight spaces and reduce stress by 2017 informing him of the distance from obstacles by means of acoustic or, depending on vehicle, optical signals. Lane Departure LDW Lane Departure Warning helps to prevent accidents ERTRAC- Warning caused by unintentionally wandering out of lane, and 2017 represents a major safety gain on motorways and major trunk roads. If there is an indication that the vehicle is about to leave the lane unintentionally, the system alerts the driver visually and in some cases by means of a signal on the steering wheel. Front Collision FCW "The Front Collision Warning monitoring system uses a ERTRAC- Warning radar sensor to detect situations where the distance to 2017 the vehicle in front is critical and helps to reduce the vehicle’s stopping distance. In dangerous situations, the system alerts the driver by means of visual and acoustic signals and/or with a warning jolt of the brakes. Front Collision Warning operates independently of the ACC automatic ERTRAC- ERTRAC- distance control." 2017 2017 Adaptive Cruise ACC The cruise control system with “automatic distance ERTRAC- Control control ACC” uses a distance sensor to measure the 2017 distance and speed relative to vehicles driving ahead. The driver sets the speed and the required time gap with buttons on the multifunction steering wheel or with the steering column stalk (depending on model). The target and actual distance from following traffic can be shown as a comparison in the multifunction display. Park Assist PA "Park Assist automatically steers the car into parallel ERTRAC- and bay parking spaces, and also out of parallel parking 2017 spaces. The system assists the driver by automatically carrying out the optimum steering movements in order to reverse-park on the ideal line. The measurement of the parking space, the allocation of the starting position and the steering movements are automatically undertaken by the Park Assist: all the driver has to do is to operate the accelerator and the brake. This means that the driver retains control of the car at all times."
May 2020 143
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Term Abbreviation Definition Source ACC including Adaptive cruise control with stop & go function ERTRAC- Stop & Go includes automatic distance control (control range 0– 2017 250 km/h) and, within the limits of the system, detects a preceding vehicle. It maintains a safe distance by automatically applying the brakes and accelerating. In slow-moving traffic and congestion, it governs braking and acceleration. Lane Keeping LKA Lane Assist automatically becomes active from a ERTRAC- Assist specific speed (normally from 50 km/h) and upwards. 2017 The system detects the lane markings and works out the position of the vehicle. If the car starts to drift off lane, the LKA takes corrective action. If the maximum action it can take is not enough to stay in lane, or the speed falls below 50 km/h, the LKA function warns the driver (e.g. with a vibration of the steering wheel). Then it is up to the driver to take correcting action. Park Assist (Level Partial Automated Parking into and out of a parking ERTRAC- 2) space, working on public parking area or in private 2017 garage. Via smartphone or key parking process is started, vehicle accomplishes parking manoeuver by itself. The driver can be located outside of the vehicle, but has to constantly monitor the system, and stops the parking manoeuver if required. Parking Garage Highly Automated parking including manoeuvring to ERTRAC- Pilot (Level 4) and from parking place. In parking garage the driver 2017, does not have to monitor the system constantly and modified may leave once the system is active. E.g. via smartphone or key, parking manoeuvre and return of the vehicle is initiated. The parking garage may take over part of the functionality, so that early introduction is supported. Automated Valet Highly Automated parking including manoeuvring in a ERTRAC- Parking (Level 4) limited area with limited speed to and from most 2017, parking spaces. The driver can leave the vehicle and modified initiates the manoeuvring to the parking space and the parking itself by e.g. smartphone or key. He does not have to monitor the system constantly and may initiate the parking-out manoeuvre the same way when coming back. Traffic Jam Assist The function maintains the longitudinal and lateral ERTRAC- (Level 2) control of the vehicle to follow the traffic flow in low 2017, speeds (<60km/h for passenger cars, <30km/h for modified freight vehicles). The system can be seen as an extension of the ACC with Stop&Go functionality without lane change support. Traffic Jam Conditional automated driving in traffic jams up to 60 ERTRAC- Chauffeur (Level km/h on motorways and roads similar to motorways. 2017 3) The system can be activated in case of a traffic jam scenario. It detects slow driving vehicles in front and then maintains the longitudinal and lateral control of the vehicle. Later version of this functionality might May 2020 144
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Term Abbreviation Definition Source include lane change functionality. The driver must deliberately activate the system, but does not have to monitor the system constantly. The driver can at all times override or switch off the system. In case of a takeover request to the driver from the system, the driver has sufficient time reserve to orientate himself and take over the driving task. In case the driver does not take over, the system will go to a reduced risk condition, i.e. bring the vehicle to a safe stop. Highway Conditional Automated Driving (up to 130 km/h for ERTRAC- Chauffeur (Level passenger cars, up to 90 km/h for freight vehicles) on 2017 3) motorways or roads similar to motorways from entrance to exit, on all lanes, including overtaking. The driver must deliberately activate the system, but does not have to monitor the system constantly. The driver can at all times override or switch off the system. In case of a takeover request to the driver from the system, the driver has sufficient time reserve to orientate himself and take over the driving task. In case the driver does not take over, the system will go to a reduced risk condition, i.e. bring the vehicle to a safe stop. Urban and Highly Automated Driving up to limitation speed, in ERTRAC- Suburban Pilot urban and suburban areas. The system can be activated 2017 (Level 4) by the driver on defined road segments, in all traffic conditions. The driver can at all time override or switch off the system. Autonomous The fully automated vehicle to handle all driving from ERTRAC- private vehicles point A to B, without any input from the passenger. The 2017 on public roads driver can at all-time override or switch off the system. (Level 5) Note: only a rough time estimation can be given for this system at the moment. C-ACC Platooning Partially automated truck platooning, in which trucks ERTRAC- (Level 1) are coupled by Cooperative ACC (C-ACC), through 2017 speed control keeping a short but safe distance to the lead vehicle, while the drivers remain responsible for all other driving functions. Automated Truck This function enables platooning in both dedicated ERTRAC- Platooning (Level lane/road and on open roads in mixed traffic. The 2017 2) vehicle should be able to keep its position in the platoon with a safe distance between the vehicles. The driving behaviour of the leading vehicle is transmitted by V2V communication to the following vehicle taking vehicle characteristics into consideration, such as braking capacity, load. The function will also handle platooning management of forming, merging and dissolving platoons together with interaction with other road users and road infrastructure requirements. Highway Pilot Automated Driving on motorways or highways from ERTRAC- platooning (Level entrance to exit, on all lanes, incl. overtaking and lane 2017 4) change. The driver must deliberately activate the May 2020 145
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Term Abbreviation Definition Source system, but does not have to monitor the system constantly. The driver can at all-time override or switch off the system. There is no request from the system to the driver to take over when the system is in normal operation area (i.e. on the motorway). Highly automated Automated freight transport carriers in confined areas ERTRAC- freight vehicles in (e.g. harbour, mining and work-site) for potentially un- 2017 confined areas manned freight transport. Vehicles can be designed (Level 4) without cab for driver. Highly automated Automated freight transport carriers on dedicated and ERTRAC- freight vehicles in controlled lanes/roads/areas and for potentially un- 2017 dedicated manned freight transport. Vehicles can be designed lanes/roads/areas without cab for driver. Operation could be done during (Level 4) night in lower speed to safe fuel. Highly automated High automated trucks for automated operation on ERTRAC- freight vehicles public roads in mixed traffic handling all typical 2017 on open roads scenarios without driver intervention on planned (Level 4) freight transport operations hub-to-hub on approved roads according to planned routes. Remote fleet and transport management and monitoring are required. Fully automated The fully automated vehicle should be able to handle all ERTRAC- freight vehicles driving from point A to B, without any input from the 2017 (Level 5) driver or passenger in all operation environments. Urban Bus Assist Automated assist functions for city-buses to increase ERTRAC- (Level 2) productivity and safety for city bus operation such as 2017 bus-stop manoeuvring, short-distance following and narrow-lane manoeuvring. Automated The automated PRT/Shuttle drives in mixed traffic with ERTRAC- PRT/Shuttles in the same speed as other traffic. 2017 mixed traffic (Level 4) Automated Buses The automated bus operates in dedicated bus lanes ERTRAC- on dedicated lane together with non-automated buses in normal city bus 2017 (Level 4) speed. Functions may include bus-trains, following and bus-stop automation for enhanced productivity, safety, traffic flow and network utilization. Automated Buses The automated bus operates in mixed traffic on open ERTRAC- in Mixed Traffic roads together in normal city traffic speed. Functions 2017 (Level 4) may include bus-trains, following and bus-stop automation for enhanced productivity, safety, traffic flow and network utilization. Fully Automated Fully automated vehicles without driver that can bring ERTRAC- Urban Vehicles passengers to any destination as “robotaxis”, 2017 (Level 5) “cybercars” or as fully automated shuttles and city buses.
May 2020 146
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance ANNEX 2: Drive2theFuture Voice of Customers Survey (full)
Consumer Survey on Autonomous Vehicles Acceptance
Welcome to our survey! By participating in this survey you will help us understand you and your preferences better! About the survey: This survey is part of an EU-funded project called Drive2theFuture (http://www.drive2thefuture.eu/).
The purpose of the survey is to explore the opinion and acceptance of customers and users regarding Autonomous Vehicles for the different transport modes.
The survey is structured in a modular way: • of four main areas corresponding to each transport domain (aviation, maritime, rail, road); • Every section begins with a simple description of what automation means for that mode, with simple illustrative examples and a short list of multiple choice questions.
It shouldn’t take you more than 10-15 minutes to answer the questions, and you can withdraw from the questionnaire at any time. There are no right or wrong answers. Your opinion as a transport user is what matters to us the most!
If you are interested in the results of the survey, do not hesitate to let us know, we will be happy to share them with you.
All the information collected will be treated in an anonymous and confidential manner by the researchers. Your demographic information will be used only to contextualize the statistical analysis of the aggregated results, and they will not be published or used in any form, other than the above mentioned statistical analysis.
All the data will be securely stored and used only for the purpose of the present research, in accordance to the ethical requirements defined within the project.
If you agree, continue to the next section. For any further information about the survey you can contact: [email protected]
Thank you for your participation.
Drive2theFuture Consortium
A. General Information May 2020 147
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
1. What is your gender? ☐ Male ☐ Female ☐ Other ☐ Prefer not to say
2. Which is your age group? ☐ 18-24 ☐ 25-35 ☐ 36-45 ☐ 46-60 ☐ >60
3. What is your country of residence? ______
4. What is your educational background (including ongoing education)? ☐ Primary/Elementary/High School Degree ☐ Trade/technical training ☐ Bachelor Degree ☐ M.Sc. ☐ PhD ☐ Other (please specify)
5. What is your employment status (dropdown menu): ☐ Employed full-time ☐ Employed part-time ☐ Self-employed ☐ Platform worker (i.e. at UBER, Airbnb, etc.) ☐ Unemployed ☐ Retired ☐ Student
6. Is operating a vehicle an aspect of your work? ☐ Yes ☐ Sometimes ☐ Rarely ☐ No ☐ Please specify which type of vehicle …………………………..
7. How much is your annual gross income? ☐ Below €10,000 ☐ €10,000 – €25,000 ☐ €26,000 – €40,000 ☐ €41,000 – €60,000 ☐ €61,000 – €80,000 ☐ €81,000 – €100,000 ☐ more than €100,000 ☐ prefer not to say
May 2020 148
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
B. Travel Behaviour
8. What mode of transport do you typically use for the following trip types? (Select ALL that apply)
ther ther
urban rail) urban
-
Passenger car Passenger Public transport Taxi Motorcycle scooter or walking or Bicycle Train Airplane Ship UBER and/or o services sharing I do not take such trips (including inter (including
Commuting
Business travel
Leisure/social
Errands (incl.
groceries)
Vacation
9. When it comes to trying a new technology product, I am generally….
☐ among the last ☐ in the middle ☐ among the first
10. Have you heard about autonomous vehicles?
☐ Yes
☐ No
11. Do you have any experience with autonomous vehicles?
☐ Yes
☐ No
Automation is the use of control systems and information technologies reducing the need for human intervention.
Several trends have been evolving across sectors around digitalisation, increased interconnectivity levels in production processes and advanced automation levels. Implementation of these technologies has already begun in many transport chain areas and will keep impacting upon all transport modes in the future.
AIR Transport
May 2020 149
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
The illustrations below present, in a simplified way, an indicative example of the Air Transport Automation levels developed exclusively for the scope of this survey.
Please read the description of levels below to decide which you prefer more.
If you wish to check the official description of the Air Transport Automation levels, please click here. Level 1
In level 1, the pilot has to react based on his/ her own observations and the information he/ she gets from the Air Traffic Control.
Level 2
In level 2, the reaction of the pilot is guided and assisted by the computer on board of the plane. The pilot has still all control.
Level 3
May 2020 150
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
In level 3, the plane itself can react automatically to each situation. The pilot can intervene at all time.
Question 1: Please rank the levels of air automation as described above from the one you prefer the most to the one you prefer the least.
1st place: ______
2nd place: ______
3rd place: ______
The illustrations below, present in addition the different levels of automation and operation concerning Drones, developed exclusively for the scope of this survey: Level 1
VLOS is a flight within Visual Line Of Sight, which is 500 metres from the pilot. The drone is operated by the pilot (from ground), who is always in visual contact with the drone.
Level 2
May 2020 151
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
In level 2, the drone is programmed to follow a certain route. It does not stay all the time in the visual line of sight of the pilot. The pilot can follow the drone on a monitor and, if needed can intervene to correct movements.
Level 3
In this level, swarms of drones autonomously make decisions based on shared information. The use of drones for real-time data collection is becoming common practice in the areas of (indicatively) precision agriculture and civil defence, such as firefighting.
Question 1: Please rank the levels of air automation as described above from the one you prefer the most to the one you prefer the least..
1st place: ______
2nd place: ______
3rd place: ______Comments (if any):
______
May 2020 152
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
______
Question 2: Your greatest concerns regarding air automation are related to:
Concern Criticality level Comments (1: not critical at all – 5: very critical) Safety: fear that it will fall out of the sky. 1 2 3 4 5 Cybersecurity: fear that it will be hijacked and used for a terrorist 1 2 3 4 5 attack. Employment risk: loss of jobs, especially in the logistics delivery 1 2 3 4 5 sector.
Other (please specify) 1 2 3 4 5
Question 3: I would accept automated air vehicles if they: have the same level of accidents as non-automated ones today. have somehow fewer accidents than non-automated ones today. have much fewer accidents (i.e. reduction of 50% or more) than non-automated ones today. have close to zero accidents.
Comments (if any):
______
Question 4: Automated flying vehicles should be used as (please rate your agreement level for the following suggestions):
Suggestion Agreement Comments (1: not at all – 5 very much) Drones to carry medicine, food, etc. during emergencies. 1 2 3 4 5
May 2020 153
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Suggestion Agreement Comments (1: not at all – 5 very much) Drones for urban deliveries with limited carrying capacity (i.e. up 1 2 3 4 5 to 5 kg).
Drones for urban deliveries without a capacity limitation. 1 2 3 4 5
Drones for deliveries outside cities. 1 2 3 4 5
Urban air vehicles for passenger transport within cities. 1 2 3 4 5
Urban air vehicles for cargo transport within cities. 1 2 3 4 5
Air transport vehicles for cargo inter-city flights. 1 2 3 4 5
Air transport vehicles for passenger inter-city flights. 1 2 3 4 5
Question 5: In your opinion, will automation in air transport affect the sector’s labour force? Yes, automation will cause job losses in the air transport sector Yes, automation, will bring new jobs in the air transport sector No significant changes will be caused in the air transport sector Other (please specify)
Question 6: In your opinion, will autonomous air vehicles facilitate the mobility of persons with disabilities? Yes No Under conditions (please specify) Question 7: What is your general opinion regarding autonomous air vehicles? 1 2 3 4 5 (Not good at (Very good) all)
Comments (if any): May 2020 154
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
______
May 2020 155
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance MARITIME Transport
The illustrations below present, in a simplified way, an indicative example of the Maritime Transport Automation levels developed exclusively for the scope of this survey.
Please read the description of levels below to decide which you prefer more.
If you wish to check the official description of the Ships Transport Automation degrees, please click here. Level 1
At level 1, the ship is equipped with extra sensors integrated into the bridge systems. The captain and the crew members have to react when these sensors detect a possible conflict.
Level 2
At this level, the on-board systems are connected through a satellite to a command centre, which helps the captain and his/ her crew to make the right decisions.
Level 3
May 2020 156
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
At level 3, the control of the ship is being realised by a remote control.
Level 4
At level 4, the ship is fully automated to the point that, in many cases, there is no bridge at all. PCs and algorithms control its operation, while vision cameras around the ship expand its location awareness capabilities and detect other vessels’ proximity.
Question 1: Please rank the levels of automation as described above from the one you prefer the most to the one you prefer the least..
1st place: ______
2nd place: ______
3rd place: ______
4th place: ______
Comments (if any):
May 2020 157
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
______
Question 2: I believe that autonomous ships should be used for (please rate your agreement level for the following suggestions):
Suggestion Agreement Comments (1: not at all – 5 very much) Rescue boats and other special purpose vessels operating in 1 2 3 4 5 harsh environments (i.e. offshore wind farm support vessels, ice breakers etc.)
Small barges transferring cargoes in inland waterways and to other 1 2 3 4 5 destinations near the coast line (i.e. from fjord to fjord).
Cargo ships of any type (i.e. Tanker vessels, Container 1 2 3 4 5 Carriers etc.). Passenger services for short sea trips (i.e. for less than 1 hour) 1 2 3 4 5
Passenger services for medium sea trips (i.e. for less than 24 1 2 3 4 5 hour)
Passenger ships of all types. 1 2 3 4 5
Question 3: Should an automated ship be marked or not? Please rate your agreement level for the following suggestions:
Suggestion Agreement Comments (1: not at all – 5 very much) Not marked at all. 1 2 3 4 5 Specifically marked at Variable Message Signs (VMS) and/ or at 1 2 3 4 5 the ticketing counters.
Announced at the ship’s arrival/ departure. 1 2 3 4 5
Question 4:
May 2020 158
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
I would accept automated ships if they: have the same level of accidents as non-automated ones today. have somehow less accidents than non-automated ones today. have much less accidents (i.e. reduction of 50% or more) than non-automated ones today.
have close to zero accidents.
Comments (if any):
______
Question 5: In your opinion, will automation in maritime transport affect the sector’s labour force? Yes, automation will cause job losses in the maritime transport sector Yes, automation, will bring new jobs in the maritime transport sector No significant changes will be caused in the maritime transport sector Other (please specify)
Question 7: What is your general opinion regarding autonomous ships? 1 2 3 4 5 (Not good at (Very good) all)
Comments (if any):
______
May 2020 159
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance RAIL Transport
The illustrations below present, in a simplified way, an indicative example of the Rail Transport Automation levels developed exclusively for the scope of this survey.
Please read the description of levels below to decide which you prefer more.
If you wish to check the official description of the Rail Transport Automation levels, please click here. Level 1
At level 1, the train operator is in full control. He/ she oversees the tracks and decides when to speed-up or slow down. He/she operates the doors and watches over the passengers’ safe disembarkation while the train is at the station.
Level 2
At level 2, the locomotive is controled remotely but the operator is still at the driver’s seat, overseeing the tracks and deciding wether or not to intervene. While at the station, he/she operates the doors and watches over the passengers’ safe disembarkation. Level 3
May 2020 160
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
At level 3, the train is controled remotely. While at the station, the doors are being operated by the station’s personnel. There is no conductor (driver) in the train.
Level 4
At level 4, the train’s control is fully automated and its operation is being monitorted remotely.
Question 1: Please rank the levels of automation as described before from the one you prefer the most to the one you prefer the least.
1st place: ______
2nd place: ______
3rd place: ______
4th place: ______
Comments (if any):
May 2020 161
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
______
Question 2: Your greatest concerns regarding rail automation are related to:
Concern Criticality level Comments (1: not critical at all – 5: very critical) Safety: fear that a technological failure will produce an accident 1 2 3 4 5
Cybersecurity: fear that it will be hijacked and used for a terrorist 1 2 3 4 5 attack.
Employment risk: loss of jobs 1 2 3 4 5
Other (please specify) 1 2 3 4 5
Question 3: In your opinion, will automation in rail transport affect the sector’s labour force? Yes, automation will cause job losses in the rail transport sector Yes, automation, will bring new jobs in the rail transport sector No significant changes will be caused in the rail transport sector Other (please specify)
Question 4: Please rate your acceptance level for the following situations:
Situation Acceptance Comments (1: not at all – 5 very much) A driverless passenger train, controlled fully by an automated 1 2 3 4 5 system.
A driverless passenger train, under the co-supervision of a 1 2 3 4 5 human remote controller.
May 2020 162
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Situation Acceptance Comments (1: not at all – 5 very much) A driverless passenger train, under the co-supervision of a 1 2 3 4 5 human attendant in train
A driverless freight train, controlled fully by an automated 1 2 3 4 5 system.
A driverless freight train, under the co-supervision of a human 1 2 3 4 5 remote controller.
An automated signalling system for trains that are operating 1 2 3 4 5 without human supervision.
An automated signalling system for trains that are operating with 1 2 3 4 5 remote human supervision.
Question 5: Should an automated train be marked or not? Please rate your agreement level for the following suggestions:
Suggestion Agreement Comments (1: not at all – 5 very much) Not marked at all. 1 2 3 4 5 Specifically marked at the Variable Message Signs (VMS) 1 2 3 4 5 and/ or at the ticketing counters.
Announced at the train’s arrival/ departure. 1 2 3 4 5
Question 6: I would accept autonomous rail vehicles if they: have the same level of accidents as non-automated ones today. have somehow fewer accidents than non-automated ones today.
have much fewer accidents (i.e. reduction of 50% or more) than non-automated ones today. have close to zero accidents.
Comments (if any):
May 2020 163
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
______
Question 7 What is your general opinion regarding autonomous and self-driving rail vehicles? 1 2 3 4 5 (Not good at (Very good) all)
Comments (if any):
______
May 2020 164
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance ROAD Transport
The illustrations below present, in a simplified way, an indicative example of the Road Transport Automation levels developed exclusively for the scope of this survey.
Please read the description of levels below to decide which you prefer more, as a road transport user.
If you wish to check the official description of the Road Transport Automation levels, please click here. Level 1
At level 1, the system informs/warns the driver on any potential danger. The driver decides on the necessary action. Thus, the car is always under driver’s control .
Level 2
At level 2, the driver explicitely allows the vehicle to be in control for specific tasks, i.e. in order to self-park the car, to drive at low speed in case of traffic-jam, etc.. Thus, the car is responsible only for a limited number of manoeuvres or driving scenario.
Level 3
May 2020 165
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
At level 3, the driver can ask the car to fully drive by itself under specific scenaria (i.e. on the motorway) and then the car tells the driver to take back control whenever necessary. Thus, there is constant shared responsibility between the driver and the car. The driver still assumes responsibility as he/she can regain control whenever he/she wants.
Level 4
At level 4, the car drives by itself. The driver is still there to react only in case of an unforeseen scenario, i.e. an emergency. The driver can also intervene whenever he/she wishes; but is not supposed to do so.
Level 5
May 2020 166
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
At level 5, no driver is present in the vehicle, which is fully automated and monitored remotely. There is not even a steering wheel in the car.
Question 1: Please rank the levels of automation as described before from the one you prefer the most to the one you prefer the least, as a road user.
1st place: ______
2nd place: ______
3rd place: ______
4th place: ______
5th place: ______
Comments (if any):
______
Question 2: Under which conditions would you prefer automation support and to what extend? Please check your preference in the scales below for the different types of road users ➢ Driver/ Passenger of autonomous road vehicle Condition Preference Comments (1: not at all – 5 very much) Adverse weather (i.e. heavy rain, wind, heavy snow, etc.) 1 2 3 4 5
In unknown environment (i.e. foreign country, unknown city, 1 2 3 4 5 rural area or countryside) May 2020 167
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Condition Preference Comments (1: not at all – 5 very much)
Different traffic environment (i.e. left or right driving) 1 2 3 4 5
Unknown vehicle type (i.e. with or without automated gearbox, 1 2 3 4 5 electric vehicle, etc.)
Other (please specify): 1 2 3 4 5
➢ Driver/passenger of non-autonomous road vehicle, in mixed flow with Autonomous Vehicles (AVs) Condition Preference Comments (1: not at all – 5: very much) Adverse weather (i.e. heavy rain, wind, heavy snow, etc.) 1 2 3 4 5
In unknown environment (i.e. foreign country, unknown city, 1 2 3 4 5 rural area or countryside)
Different traffic environment (i.e. left or right driving) 1 2 3 4 5
Unknown vehicle type (i.e. with or without automated gearbox, 1 2 3 4 5 electric vehicle, etc.)
Other (please specify) 1 2 3 4 5
➢ Vulnerable Road Users (VRU) (i.e. pedestrian, cyclist, etc.) in mixed traffic with AVs and non AVs Condition Preference level Comments (1: not at all – 5 very much) Adverse weather (i.e. heavy rain, wind, heavy snow, etc.) 1 2 3 4 5
In unknown environment (i.e. foreign country, unknown city, 1 2 3 4 5 rural area or countryside)
Different traffic environment (i.e. left or right driving) 1 2 3 4 5
Unknown vehicle type (i.e. with or without automated gearbox, 1 2 3 4 5 electric vehicle, etc.)
May 2020 168
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Condition Preference level Comments (1: not at all – 5 very much) Other (please specify) 1 2 3 4 5
Question 3: In which way do you think that autonomous road vehicles should be marked in traffic? Other, please specify
With a sign on top of them With a flashing siren on top of them With a flashing siren and sound on top of them
Question 4: Your greatest concerns regarding road automation are related to:
Concern Criticality level Comments (1: not critical at all – 5: very critical) Security: what if my child travels in an automated bus, without 1 2 3 4 5 driver and some stranger kidnaps it?
Safety: what if the autonomous road vehicle makes a choice that 1 2 3 4 5 leads to an accident?
Cybersecurity: what if someone takes control of the vehicle and 1 2 3 4 5 uses it in a terrorist attack?
Data Privacy: what if someone gets access to your data? 1 2 3 4 5
Employment risk: what will happen to the public transport 1 2 3 4 5 and other vehicle drivers? Will many jobs be lost? May 2020 169
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Concern Criticality level Comments (1: not critical at all – 5: very critical)
Cost of AVs: what will be the cost of the new automated cars? Will it 1 2 3 4 5 be affordable? Other (please specify) 1 2 3 4 5
Question 5: In your opinion, will automation in road transport affect the sector’s labour force? Yes, automation will cause job losses in the road transport sector Yes, automation, will bring new jobs in the road transport sector No significant changes will be caused in the road transport sector Other (please specify)
Question 6: Which of the following statements you agree with?
Statement Agreement level Comments (1: I don’t agree at all – 5: I fully agree) No driver or human supervision will be required in future 1 2 3 4 5 automated buses.
A remote supervisor is required but does not need to be seen or 1 2 3 4 5 felt by the passengers.
A remote supervisor is required, with the possibility to talk to, 1 2 3 4 5 hear passengers and react to their requests.
A remote supervisor is required, with the possibility to also be 1 2 3 4 5 presented as an avatar in a vehicle screen.
A remote supervisor is required, also presented as a real-time 1 2 3 4 5 video-audio connection in a vehicle screen.
May 2020 170
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Statement Agreement level Comments (1: I don’t agree at all – 5: I fully agree) A driver should always be there, as a back-up to the system. 1 2 3 4 5
Fully autonomous public transport vehicles (bus, tram, 1 2 3 4 5 etc.) should have speed restrictions (i.e. up to 20km/h).
Private cars and should have speed restrictions (i.e. up to 1 2 3 4 5 30km/h) only in towns.
Private cars and should have speed restrictions (i.e. up to 1 2 3 4 5 70km/h) in motorways also
Fully autonomous road vehicles should have the same 1 2 3 4 5 speed limits as non- autonomous vehicles
Question 7: I would accept autonomous road vehicles if they: have the same amount of accidents as non-automated ones today. have fewer accidents than non-automated vehicles today have much fewer accidents (i.e. reduction of 50% or more) than non-automated vehicles today. have close to zero accidents.
Comments (if any):
______
Question 8: I would accept autonomous road vehicles if they: have higher price than the non-automated ones of today have the same cost as the non-automated ones of today. have somehow less cost than the non-automated ones of today. have much less cost than the non-automated ones of today.
May 2020 171
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Comments (if any):
______
Question 9: In your opinion, road users would need to be reskilled for using autonomous vehicles? Yes, somehow reskilled Yes, significantly reskilled No Question 9a:
If yes, for which skills related to autonomous road vehicles, would you like to receive training for?
______
Question 10: In your opinion, will autonomous road vehicles facilitate the mobility of persons with disabilities? Yes No Under conditions (please specify)
Question 11: What is your general opinion regarding autonomous and self-driving road vehicles? 1 2 3 4 5 (Not good at (Very good) all)
Comments (if any):
______
May 2020 172
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
Thank you for your participation!
If you wish to receive our news and the outcomes of the Drive2TheFuture project, you may go to our website and register in our Newsletter (http://www.drive2thefuture.eu/newsletter/)
We’ll be very glad to have you on board!
The Drive2theFuture Consortium
May 2020 173
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
ANNEX 3: Literature Review Template for Open Research Issues and Hypotheses
Title: ______
Authors: ______
Year:
Source: ______
Type of source: □ Scientific Journal □ Conference Proceedings □ Book/ Chapter □ Other (please specify) ______
Short description (up to 200 words) ______May 2020 174
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
______
Relevance to D2F (up to 3-5 lines) Acceptance/ Awareness issues User clusters/ user needs VRUs Use Cases/ Scenarios Simulation System development Research Priorities Other (please specify): ______
Comments:______
Specific user needs/ acceptance information (up to half a page) ______May 2020 175
D1.1: User clusters, opinion, research hypotheses and use cases towards future autonomous vehicle acceptance
______
Conclusions to be transferred to Drive2theFuture (up to 10 bullets)
May 2020 176