From ADAS to Automated Driving Technical Session Schedule As of 10/15/2018 07:42 pm

Tuesday, October 9 Day 1 Session Code: ADAS100 Room 360 Session Time: ALL DAY Market Analysis and Outlook, Deployment Strategies, Safety Time Paper No. Title

9:00 a.m. ORAL ONLY Welcome and Introductions Tim A. Cavanaugh, SAE International 9:05 a.m. ORAL ONLY Keynote Presentation: Connected Automation: Vision, Planning and Initiatives at MDOT Kirk Steudle, Secretary, Michigan Dept. of Transportation 10:00 a.m. ORAL ONLY ADAS to Autonomy, a Means to an End? Car makers are scrambling to deploy safety systems and build the algorithms that will replace human driving. It’s time to assess the progress and reconsider the objective. Will autonomy be a feature, a service or an entirely new form of transportation? We’ll look at the promises, the projections and the data.

Roger Lanctot, Strategy Analytics Inc. 10:30 a.m. Networking Break

11:00 a.m. ORAL ONLY Vehicle Communications (V2X)

Vehicle to location services using satellite and cellular networks such as OnStar have been on vehicles since 1996.

Vehicle to Vehicle (V2V) communication is appearing on production vehicles. Investigative studies, test fleets and technical specifications have concentrated on Dedicated Shortrange Radio Communication (DSRC) technology at 5.9 GHz. A newcomer, funded by the telecom industry, is 5G cellular technology.

V2X Communication can provide over 100 immediate safety benefits (traffic congestion, early warning braking, construction zones, emergency vehicle and intersection safety). One of the enabling technologies for autonomous vehicles is V2X communication. Bob’s multimedia presentation will provide an overview of the technologies.

Robert L. Neff, Intrass 11:30 a.m. ORAL ONLY Current Simulation Shortcomings and How Revolutionary Solutions Will Make AVs a Reality The pursuit of autonomy is completely changing how vehicles are designed, developed and especially tested – in most cases, before official regulation or industry standards exist. Autonomous system software is overwhelmingly complex and constantly evolving, and the development pace and maintenance cost of highly and fully automated vehicles outpaces current development and testing methodologies. These factors all exacerbate the cost, resource and time requirements of testing AVs. Jeff Blackburn will discuss how the legacy simulation tools are limiting the testing that needs to take place. He'll also explain how scalable cloud based simulation is the enabling technology making the deployment of AVs possible by providing an immediate means to test the performance of autonomous systems in a risk-free environment, allowing automakers to stick to their 202X AV production goals. Jeffery Blackburn, Metamoto 12:00 p.m. Networking Lunch

1:00 p.m. ORAL ONLY Hardware-in-the-Loop (HIL) Implementation and Validation of SAE Level 2 Autonomous Vehicle with Subsystem Fault Tolerant Fallback Performance for Takeover Scenarios The advancement towards development of autonomy follows either the bottom- up approach of gradually improving and expanding existing Advanced Driver Assist Systems (ADAS) technology where the driver is present in the control loop or the top-down approach of directly developing Autonomous Vehicles (AV) hardware and software using alternative approaches without the driver present in the control loop. Most ADAS systems today fall under the classification of SAE Level 1 which is also referred to as the driver assistance level. The progression from SAE Level 1 to SAE Level 2 or partial automation involves the critical task of merging autonomous lateral control and autonomous longitudinal control such that the tasks of steering and acceleration/deceleration are not required to be handled by the driver under certain conditions [1]. However, the driver is still required to monitor the driving environment and handle scenarios where control is handed over to the driver due to subsystem faults of the autonomous system. Due to the disadvantages of vehicle testing being expensive, time-consuming and hazardous for testing such scenarios, an alternative method of development and validation is required. Therefore, the objectives of this research are two-fold. The first objective focuses on a real-time powertrain-based Hardware-in-the- Loop (HIL) implementation and validation of an SAE Level 2 autonomous vehicle. The second objective focuses on studying the performance of SAE Level 2 autonomous vehicles during takeover scenarios due to subsystem faults. To accomplish these objectives, an acceleration-based Adaptive Cruise Control (ACC) was combined with a path-following lateral control along with supervisory control for system mode transitions due to system deactivations and faults. This research presents system modes in which longitudinal control only and lateral control only are engaged as fallback states to the full autonomous system being faulted for lateral control and longitudinal control failures respectively. Simulations were conducted to evaluate the performance of the autonomous controls when subjected to these faults. A powertrain subsystem representative of the 2017 Hybrid was used as the hardware simulation platform using a dSPACE HIL simulator and CarSim RT. Adit Joshi, 1:30 p.m. ORAL ONLY The Two Pathways to Safe Mobility

Motor vehicle traffic fatalities had been steadily declining for a number of years, and between 2005 and 2011, there was an average decrease of over 1564 fatalities per year over that period. However, this trend has reversed since then, and an increase in these statistics has occurred through the first nine months of 2016, and NHTSA has estimated that the increase in 2016 may be as high as 8 percent over the previous year, based on the nine month’s data that are available for 2016 (NHTSA Traffic Safety Facts, January 2017). This increase, though significant, NHTSA has stated that in 2015 it was mainly due to fatality increases in vulnerable road users such as pedestrians, motorcyclists and pedal cyclists. However, it is not clear if there has been any attempt to understand the technical reasons for this increase beyond the increase in vehicle miles driven, except to rationalize it by saying that similar trends have been observed previously during a period of strong economic growth. Nevertheless, these vulnerable road user casualties need solutions as much as occupant casualties in traffic crashes.

Passive safety systems that are found in vehicles of today in the fleet were generally phased in over a period of three or four years through the National Highway Safety Administration (NHTSA) mandated regulations, or through New Car Assessment Program (NCAP) and other such rating programs, as the main driver for safety enhancement.

Joseph Kanianthra, retired, NHTSA 2:00 p.m. ORAL ONLY Observations on ADAS and Automated Driving in the Wild The concept of automating vehicles and removing the driver from direct control of the throttle, brake, and steering wheel was first explored nearly 100 years ago. Over the decades since, billons of dollars have been invested and automation has infiltrated the automobile in a multitude of ways. This talk will center on observations from the field, gathered through the Advanced Vehicle Technologies (AVT) consortium and other salient events. Highlighted will be observations on the use of production level automated driving features, the shifting nature of what we do in modern vehicles, and perceptions of the evolving automation technologies. Bryan Reimer, Massachusetts Institute of Technology 2:30 p.m. ORAL ONLY Integrated Safety As higher levels of automation evolve in future vehicles the goal will be to provide additional value to consumers. During the ride occupants may enjoy a higher flexibility by working, relaxing or interacting with other occupants. These additional use cases lead to challenges in designing an occupant safety system that maintains or enhances the level of safety during all modes of driving. This presentation defines a potential approach for an occupant safety system that supports alternative interior configurations and seat positions and discusses the role of enabling technologies such as advanced occupant sensing technologies that will help drive the integration of automated driving, occupant safety and consumer experience. Frank Laakmann, ZF Friedrichshafen AG 3:00 p.m. ORAL ONLY Challenges and Opportunities in Crashworthiness Evaluation of Autonomous Vehicles An accident free environment will take many years to attain, and it is hard to envision one even with the entire fleet consisting of autonomous, driverless vehicles. Some level of crashworthiness will still be required. Towards that end, the FMVSS 200 series of standards are expected to continue with changes to the regulatory language to enable testing of driverless vehicles to demonstrate compliance to the safety standards. Current compliance tests consist of specified velocities, direction of impacts and crash appropriately instrumented Anthropometric Test Devices. The relevance of the tests may be questionable as seating configurations evolve from the traditional configurations. This presentation will discuss the challenges associated with test devices, their bio- fidelity and associated injury criteria. The development of ATD’s is a time consuming, long term proposition. Till then a move to CAE oriented demonstration of compliance to safety standards is an opportunity to should be investigated. Priya Prasad, Prasad Engg LLC 3:30 p.m. Networking Break

4:00 p.m. ORAL ONLY Technical Breakout Sessions (Rooms: 360, 354 and 355) Moderators - Michael Carpenter, General Motors; Scott Craig, ON Semiconductor; Fred Huntzicker, GM; Joseph Kanianthra, Active Safety Engineering LLC; Oliver Spiess, ZF

Wednesday, October 10 Day 2 Session Code: ADAS200 Room 360 Session Time: ALL DAY Testing and Data Collection, AI & Machine Learning, Sensor Fusion, and Maintenance Time Paper No. Title

8:00 a.m. ORAL ONLY Keynote Presentation John P. Capp, General Motors 8:30 a.m. ORAL ONLY Evolving V&V Testing to Support an Automated Future The deployment of Advanced Driver Assistance Systems (ADAS) and Automated Vehicles (AVs) will transform the face of mobility and the transportation system as we know it. This transformation to an automated future, and the speed of this transformation, will depend not only on the collection and analysis of the abundant data that will be available but will also depend on robust and adaptable methods of conducting verification and validation (V&V) testing. This presentation will explore potential approaches to V&V testing, including data collection and analysis. Jeffrey Wishart, Exponent Inc. 9:00 a.m. ORAL ONLY Testing & Data Collection for ADS Safety Developing level 3 to 5 automated driving systems (ADS) is a quantum leap in complexity beyond ADAS. The key challenges include identifying the driving situations to be addressed, and building and demonstrating confidence in ADS performance in those situations, in the diversity of environments and conditions in the real world. How can testing and data collection, together with simulation, make this a feasible and defensible possibility? What roles can on-road data collection play, including post-deployment? David LeBlanc, UMTRI 9:30 a.m. Networking Break

10:00 a.m. ORAL ONLY AI-enabled Sensors While sensors are critical to helping autonomous vehicles understand their surroundings, the massive amount of data they produce takes time and significant computing power to process. A new generation of AI-enabled smart sensors solve this problem by pushing processing to the edge of the network. AEye will explore how software-configurable hardware enables vehicle control system software to selectively customize data collection in real-time, while optional edge processing reduces control loop latency for optimal performance and safety. Learn how to increase reliability, lower power requirements and reduce median system bandwidth requirements--advances that are key to high resolution, low latency mobility solutions. Barry Behnken, AEye Inc. 10:30 a.m. ORAL ONLY AI and Self Driving Cars Artificial Intelligence is transforming virtually every industry, including transportation. NVIDIA is building AI platforms for autonomous vehicles as well as the infrastructure to train and test them. Inside the vehicle, NVIDIA envisions the need to run numerous deep neural networks (DNNs) simultaneously in order to enable the vehicle’s artificial intelligence to drive the vehicle, while keeping the driver, its occupants, pedestrians, and other vehicles safe. NVIDIA has also created a simulation system for validation of these networks and other system software, that will enable AV developers to safely drive billions of miles in virtual reality and test rare edge cases, before putting actual vehicles on the road. This talk will provide insight into, and showcase demonstrations of the existing state of autonomous vehicles in development, as well as discuss the steps needed to bring these systems to production. Luke Harvey, NVIDIA 11:00 a.m. ORAL ONLY Managing System Complexity with the Rise of AI for Autonomous Systems

The era of the software-defined vehicle is upon us. Given the complexity of autonomous driving and with so much at risk, the issues of safety and security are more critical than ever. The highly connected nature of autonomous vehicles opens the door to new threat vectors. Not only must cars intelligently and instantaneously see, understand and communicate across its internal systems, but also with the external world around it.

This presentation will investigate how software in autonomous systems is increasing in complexity and scale, and the appropriate system architectures to address the resulting challenges. It will be critical to implement robust partitioning, especially as AI continues to evolve and there is an growing need for compute power and data flow to be increasingly fluid across automotive systems with mixed criticality. Maarten Koning, Wind River Systems 11:30 a.m. Networking Lunch

12:30 p.m. ORAL ONLY The new generation of 3D Solid State LiDAR for Autonomous Driving In order to reach the goal of autonomous driving, vehicles must be equipped with different kinds of sensing devices. The fusion of the 3D Solid State LiDAR together with radar, camera and HD maps will allow autonomous driving to have a high level of confidence and safety. In this presentation, the measurement principle of the new 3D Solid State LiDAR generation will be explained. Currently Ibeo is developing a solid sate scanner, called NEXT: collected data examples from the prototype will help you understanding how the system architecture provides centralized processing of the LiDARs, which offer variable field of views. The data processing has a 3D object tracking and classification algorithm, together with a grid based modeling of the environment. The first samples of the new ibeo 3D Solid Sate LiDAR will be available in 2019. Mario Brumm, IBEO Automobile Systems 1:00 p.m. ORAL ONLY Knocking Down the Barriers to Autonomous Driving: Enabling Technologies that will Define the Adoption Curve

When today’s autonomous vehicles are unable to recognize an obstacle or identify a threat, they are prone to disregard what they can’t “see.― This tendency to default to a do-nothing mode can put the vehicle’s occupants – and those around them – in grave danger.

This critical blind spot is caused by the industry’s lack of a comprehensive safety ecosystem in which the data the vehicle collects is reliable enough to render fail-safe decisions. By themselves, the tools that collect this data – cameras, motion sensors, RADAR and LiDAR – are unable to recognize and respond to all potential threats.

Chris Jacobs, Analog Devices Inc. 1:30 p.m. ORAL ONLY Technical and Market Evaluation of Sensor Fusion Concepts The race for autonomous driving has begun, with high expectations regarding timing, pricing and safety. To be successful, a seamless integration between all sensor inputs is critical. Sensor fusion need to balance the requirement for capability and robustness while optimizing component and computation costs. After an overview of market trends influencing sensor fusion, Modar will present technical considerations for designing and evaluating future concepts of sensor fusion. Modar will also present briefly on P3’s Autonomous Data and Analytics Platform for Testing (ADAPT). Modar Horani, P3 North America Inc. 2:00 p.m. ORAL ONLY Virtual Testing of Autonomous Vehicles with Simulation

Before viable autonomous vehicles can be brought to market their safety needs to be validated over virtually infinite different driving situations created by combinations of hundreds of variables such as lighting conditions, traffic patterns, weather, pedestrian behavior, road surface variations, and others. Simulation is key to validating such a wide range of driving scenarios.

This presentation will focus on high-fidelity physics-based simulations. Different use cases of simulation in the development of automated driving systems will be presented, including simulation of radar, lidar, camera and other sensors, electronics and semiconductors and closed-loop simulation with Software-In- Loop (SIL) as well as Hardware-In-Loop (HIL).

Sandeep Sovani, ANSYS Inc. 2:30 p.m. Networking Break

3:00 p.m. ORAL ONLY Predictive Analytics for Mobile Mining Equipment Mobile mining equipment is equipped with a wide variety of , transmissions, wheel groups or generators and wheel motors, as well as extensive hydraulic steering, braking, and hoist systems. These high dollar components are integrated into an electronic architectural backbone that provides machine command and control. These collective on-board components can be managed to achieve high reliability and low maintenance costs through the robust application of domain expertise, advanced analytics, and best in class maintenance practices. During this presentation, we will explore how Caterpillar uses dealer expertise coupled with best in class analytics inside Equipment Care Advisor, and world class maintenance practices to achieve optimum mobile mining fleet performance. Brian Schissler-Boichot, Caterpillar Inc. 3:30 p.m. ORAL ONLY Automotive Prognostics Emerges The advent of predictive analyt¬ics technology coupled with high-bandwidth telematic access to vehicles has led to advancements in automotive prognostics. This new para¬digm offers a path to enhance diagnostic development and imple¬mentation as well. Prognostics have demonstrated the ability to dramatically improve “perceived― reliability as well as other potential benefits. For example, the same type of information that is useful for detecting the onset of specific failure modes can also be used to enhance vehicle validation and to better manage field issues. The need becomes ever greater as vehicles are equipped with more sophisticated control systems and advanced safety features. In the future, automotive suppliers will need to play an increasingly vital role in the cost-effective implementation of prognostics by providing “health-ready components― to the OEMs. Emerging standards under development within SAE are expected to facilitate that eventuality. Steven W. Holland, General Motors 4:00 p.m. ORAL ONLY Technical Breakout Sessions (Rooms: 360, 354 and 355) Moderators - David LeBlanc, UMTRI; Rini Sherony, Toyota Motor Corp.; Scott Craig, ON Semiconductor; Oliver Spiess, ZF; Steve Holland, General Motors; Brian Schissler-Boichot, Caterpillar Inc.

Thursday, October 11 Day 3 Session Code: ADAS300 Room 360 Session Time: ALL DAY Car Architecture, Panel Discussion on System Integration, Human Experience/Factors, and Insurance/Legal Time Paper No. Title

9:00 a.m. ORAL ONLY Keynote Presentation: The Future of Mobility - From ADAS to Automated and beyond

Connectivity, automated driving, and vehicle electrification will help the industry reach the ‘triple zero’ target of zero crashes and fatalities, zero transportation carbon footprint, and zero stress. The rise of innovative mobility services, such as ride-hailing and carsharing, is part of the mobility evolution, a long-term gradual evolution of transportation preferences, toward on-demand shared mobility and a multimodal system that is less car-centric. Ultimately, deployment of automated vehicles promises increased safety and mobility and will have profound implications not only for people’s transportation choices and behavior but also for all players in the automotive sector.

Ms. Bailo’s talk will take a look at an industry view of 2025 and beyond including; how mobility, electrification, and automation will impact the industry. Carla Bailo, Center for Automotive Research 9:30 a.m. ORAL ONLY Power Distribution in Automated Driving Vehicles

Automated Driving and Driverless cars create a number of challenges for the Automotive Industry. While sensing the vehicle environment and computing a dedicated driving strategy are essential for such vehicles, it is equally important to supply Sensors, Controllers and Actuators with Electric Energy. Dependable Electric Power and Power Distribution are the key to make e.g. Robotaxis happen.

Today’s cars commonly only use a Fail Safe Power Supply system. In case of a power failure we depend on the driver to take over. In Highly Automated Cars this is not further possible and therefore we need to supply electric power fail operationally. The focus of this paper is based on research how to make this happen. In this presentation we will show different architectures and safety mechanisms potentially to be introduced into the power distribution in automated diving vehicles. We will explore options that there is always a reliable electric power supply. With the architecture examples it will lead to the proposals for required switching and fusing elements in the car circuitry.

Stefan Schumi, Infineon Technologies AG 10:00 a.m. ORAL ONLY Pathway to Decreasing ADAS & AV Sensor Complexity As the auto industry increases its capabilities into the various levels of autonomy so does the complexity and cost for all the sensors needed to accomplish it. Much of this complexity is driven by redundancy requirements and lack of performance from individual sensors. We’ll explore the different combinations of sensors that could help meet these requirements and dive into where current sensor capabilities are lacking. Brunno Moretti, General Motors 10:30 a.m. Networking Break 11:00 a.m. Panel Panel Discussion: System Integration Onboard vehicle ADAS and Safety Systems need to be defined, designed, produced and tested so that the integration of their functions operate in a manner expected and acceptable to their human passengers and other vehicles on the roadway. This panel will discuss how we can get there.

Learn more about Panel participants Moderators - Carla Bailo, Center for Automotive Research Panelists - Jace Allen, dSPACE Inc.; Ross Jatou, ON Semiconductor; Aaron M. Jefferson, ZF - TRW; Akash Maheshwari, Continental Automotive; Rini Sherony, Toyota Motor Corp.; Christopher Thibodeau, Ushr Inc.;

12:30 p.m. Networking Lunch

1:30 p.m. ORAL ONLY Keynote Presentation: Consumer Information Testing of Advanced Driver Assistance Systems Consumer information testing of vehicle safety is increasingly focused on crash avoidance and driver assistance systems. This focus presents unique challenges compared to crash testing, which has historically been the hallmark of vehicle safety information tests. Whereas crash test results are largely determined by physics, the effects of driving safety of driver assistance systems depends on variables including how and whether drivers will use the assistance, how the assistance changes the driving task and the variation in circumstances leading up to imminent crash circumstances. This presentation will provide and overview of current evaluation methods for crash avoidance and driver assistance and describe research being conducted to develop evaluations of systems that provide even more assistance. David S. Zuby, Insurance Institute for Highway Safety 2:00 p.m. ORAL ONLY Human Factors of Automated Driving Systems: How well do we understand these systems? Automated Driving Systems will induce disruptions in the way people have traditionally interacted with vehicles, and these disruptions will likely affect all road users. This is true even for currently available vehicles with ADAS. This talk will focus on the consequences that vehicle automation will have on human driving behavior. Gaps in our knowledge about users’ expectations, perceptions, and understanding of these systems limit our ability to anticipate the adoption of these technologies. Education, training, and effective HMI design can play critical roles in raising awareness to ensure clear understanding about appropriate use of and expectations about these systems. Anuj Pradhan, Univ. of Michigan 2:30 p.m. ORAL ONLY Consumers’ Perception, Understanding and Training with Advanced Driver Assistance Systems Consumers are faced with an increasingly complex decision process as novel safety technologies become commonplace in new vehicles. Consumers’ knowledge of these systems is potentially limited given the recent introduction and constant evolution of ADAS. This session will discuss data from a survey on consumer perception and understanding as well as look at research aimed at providing training for drivers to improve their understanding of advanced vehicle technologies. Implications for the design and implementation of ADAS will be discussed. David M. Cades, Exponent Failure Analysis 3:00 p.m. ORAL ONLY Computer Vision and AI – Aftermarket Applications in Commercial Fleets: Case study of risk reduction with ADAS and other in-cab notifications This presentation will discuss how computer vision coupled with artificial intelligence/machine learning is being practically applied to monitor and improve driving performance and safety outcomes in commercial fleets. Moving beyond legacy - inertially triggered video monitoring systems, the application of computer vision with real time in-cab ADAS alerts has dramatically improved commercial fleet operators' ability to both measure and reduce risky behaviors. Additionally, this vision-based driver safety & recognition system has been trained to measure positive driving habits to enable a more complete view of a fleet’s performance, providing the foundation for a balanced and meaningful driver rewards and recognition system. Dale Willis, Netradyne 3:30 p.m. Networking Break

4:00 p.m. ORAL ONLY Legal Challenges for Automated and Autonomous Driving

Automated and autonomous driving (AD) have significant ramifications and potential benefits for society. Such systems can ease our lives, contribute to road safety, and can provide a platform for numerous digital services. Companies launching AD systems may find promising new business opportunities but will also need to cope with increased uncertainty in terms of success factors, fragmentation of regulations, and liability risks.

In addition to the question of how automated vehicles will be deployed on global roadways, an evaluation of AD systems within a complex eco-system is required to determine the multi-faceted legal ramifications of these new systems.

Dagmar Thimm, ZF Global Electronics 4:30 p.m. ORAL ONLY Design for Legal Risk Mitigation: How Accidents, Disengagements and Regulatory Uncertainty Impact the Development of Autonomous Technology

Autonomous driving will ultimately save lives, but its development has experienced challenges that include heavily publicized fatalities. In these incidents, lawsuits are filed and settled. Root cause activities proceed on disengagements and other component failures. All the while, we look to NHTSA’s history to project where future regulatory actions might occur.

This presentation will discuss the likely path of regulation and guidance from federal, state and local governments and how those rules translate to specifications. It will address how and how product development teams can integrate lessons learned from fatalities and disengagement events into the product design cycle in an effort to reduce the risk of future events.

Jennifer Dukarski, Butzel Long 5:00 p.m. ORAL ONLY Closing Remarks Tim A. Cavanaugh, SAE International