Preparation of Road Safety Inspection and Audit Plans for Core/Comprehensive Road Network in Western Balkans (WB6) and Pilots

TA to Connectivity in the Western Balkans

EuropeAid/137850/IH/SER/MULTI

Sub-Project

Code: CONNECTA-TRA-CRM-REG-01 Area: Connectivity Transport Reform Measures

Preparation of Road Safety Inspection and Audit Plans for core/comprehensive road network in Western Balkans (WB6) and Pilots

FINAL REPORT (Consolidated)

30 July 2018

Page 1

Issue and revision record

Revision Date Originator Checker Approver Description 1 30 June 2018 Jesper Mertner Kostas Chris Final Report (consolidated) (Proj. Manager) Georgiou Germanacos 2 24 July 2018 Jesper Mertner Kostas Chris Final Report (consolidated) (Proj. Manager) Georgiou Germanacos 3 30 July 2018 Jesper Mertner Liljana Çela Liljana Çela Final Report (consolidated) (Proj. Manager) (Task manager) (Task manager)

Information Class: EU Standard

The contents of this document are the sole responsibility of the Mott MacDonald Connecta Consortium and can in no way be taken to reflect the views of the European Union.

This document is issued for the party, which commissioned it and for specific purposes connected with the above-captioned project only. It should not be relied upon by any other party or used for any other purpose.

We accept no responsibility for the consequences of this document being relied upon by any other party, or being used for any other purpose, or containing any error or omission which is due to an error or omission in data supplied to us by other parties.

This document contains confidential information and proprietary intellectual property. It should not be shown to other parties without consent from us and from the party, which commissioned it.

This report has been pr epared sol el y for use by the party which commissi oned it (the ‘Client’) i n connecti on with the capti oned proj ect. It should not be used for any other purpose. N o person other than the Client or any party who has expressl y agreed ter ms of r eliance with us (the ‘Reci pient(s)’) may rel y on the content, i nformati on or any views expressed i n the repor t. We accept no duty of care, responsi bility or liability to any other r eci pient of thi s document. T his r eport is confi denti al and contains pr opri etar y intell ectual property.

No representati on, warranty or under taking, expr ess or i mplied, is made and no responsi bility or liability is accepted by us to any party other than the Cli ent or any Reci pient(s), as to the accuracy or completeness of the i nformati on contai ned i n this r eport. For the avoidance of doubt this r eport does not in any way purport to i nclude any legal , insur ance or fi nanci al advice or opi nion.

We disclai m all and any liability whether arising i n tort or contract or other wise which it might otherwise have to any party other than the Cli ent or the Reci pient(s), in r espect of this report, or any infor mation attri buted to i t.

We accept no r esponsibility for any error or omission i n the r eport which is due to an error or omission i n data, infor mation or statements supplied to us by other par ties incl udi ng the client (‘D ata’). We have not i ndependentl y verified such D ata and have assumed it to be accurate, complete, reli abl e and current as of the date of such infor mation. Forecasts presented i n this document were pr epared usi ng Data and the report is dependent or based on D ata. Inevitabl y, some of the assumptions used to develop the for ecasts will not be realised and unantici pated events and circumstances may occur. C onsequentl y M ott MacDonal d does not guarantee or warr ant the concl usi ons contained i n the repor t as there are li kel y to be differ ences between the for ecasts and the actual results and those di ffer ences may be material. Whil e we c onsi der that the infor mation and opini ons gi ven i n this r eport are sound all parti es must rel y on their own skill and j udgeme nt when making use of it. Under no circumstances may this report or any extr act or summar y ther eof be used in connection with any public or pri vate sec urities offering i ncluding any rel ated memorandum or prospectus for any securities offering or stock exchange listing or announcement.

Page 2

Cooperation and Acknowledgements

We would like to thank the Task Manager Liljana Cela, monitoring this Connecta sub-project, but also General Director Dejan Lasica from SEETO for helping to coordinate data collection and arranging meetings with stakeholders with the Regional Participants, and also providing necessary advice.

In the course of this sub-project, the Connecta team would like to acknowledge the kind support and good collaboration from the Road Safety Working Group members, chaired by Said Dahdah from the World Bank, from institutions from all Regional Participants acting also as focal points and devoting time for meetings and interviews as well as providing data and reviewing reports.

Page 3

List of Abbreviations

ALB/AL Albania AIS Accident Information System AO Administrative Order ARA Albanian Roads Authority ARCMAP GIS Mapping program AusRAP Australian Road Assessment Programme AVL Automatic Vehicle Location system BiH Bosnia and Herzegovina CA Contracting Authority CADaS Common Road Accident Data Framework In Europe ChinaRAP China Road Assessment Programme CNC Core Network Corridor Connecta The MMD led Consortium implementing Connecta Connecta Technical Assistance to Connectivity in the Western Balkans CRM Connectivity Reform Measures CRMMP Connectivity Reform Measures Management Plan DfT Department for Transport UK DG MOVE Directorate-General for Mobility and Transport DG NEAR Directorate-General for Neighbourhood and Enlargement Negotiations EB Empirical Bayes EBRD European Bank for Reconstruction and Development EC European Commission EU European Union EUR Euro (currency) EuroRAP The European Road Assessment Programme FBiH Federation of Bosnia and Herzegovina (entity) FR Final Report GDP Gross Domestic Product GIS Geographical Information Systems GPS Global Positioning System ICD International Classification for Diseases ICJ International Court of Justice IFI International Financing Institution IFICO International Financing Institution Coordination Office INSTAT Albanian Institute of Statistics IPA Instrument for Pre-accession Assistance IPF Infrastructure Project Facility ITS Intelligent Transport Systems IR Inception Report iRAP The International Road Assessment Programme IRTAD International Road Traffic and Accident Database ISS Information Sharing Systems ITE Institute of Transportation Engineers

Page 4

ITS Intelligent Transport Systems KE Key Expert KfW Kreditanstalt fur Wiederaufbau (Bank) KiwiRAP Kiwi Road Assessment Programme KOS Kosovo* Police Information System KoM Kick-off-Meeting KOS Kosovo (hereinafter referred to as Kosovo) MAIS Maximum Abbreviated Injury Scale MAP Multi Annual Plan MED Mediterranean (corridor) MIE Ministry of Infrastructure and Energy MKD the former Yugoslav Republic of Macedonia MMD Mott MacDonald MMUAC Model Minimum Uniform Crash Criteria MNE/MON Montenegro MOI Ministry of Infrastructure MONSTAT statistical office of Montenegro MoTC/MoI/MoCTI Ministry related to Transport and Infrastructure MoT Ministry of Transport MoU Memorandum of Understanding MTI Ministry of Transport and Infrastructure NIPAC National IPA Coordinator NKE Non-Key Expert OEM Orient East Mediterranean (corridor) PD Preliminary Design PDF Project Description Form PE Public Enterprise PERS/PE RoS Public Enterprise Roads of Serbia PM Project Manager REG Regional RFA Request for Approval RPS Road Protection Score RS Republic of Srpska (entity of Bosnia and Herzegovina) RS Road Safety RSA Road Safety Audit RSI Road Safety Inspection RSWG Road Safety Working Group SAFEGE Monitoring of the Road Safety Strategies in SEETO Members and Draft a Regional Short-term Action Plan – September 2015 SEETIS South East Europe Transport Information System SEETO South East Europe Transport Observatory SNKE Senior Non Key Expert SPF Safety Performance Functions

 This designation is without prejudice to positions on status and is in line with UNSCR 1244/1999 and the ICJ Opinion on the Kosovo declaration of independence.

Page 5

SQL Structured Query Language. SRB/SER Serbia TA Technical Assistance TAIEX Technical Assistance and Information Exchange instrument of the European Commission TEN-T Trans-European Network – Transport TL Team Leader ToR Terms of Reference TRA Transport UNSCR United Nations Security Council Resolutions usRAP United States Road Assessment Programme ViDA cloud-based software developed for iRAP WB Western Balkan WB (G) World Bank (Group) WB6 Western Balkans 6 Regional Participants WHO World Health Organisation

Page 6

Contents

Cooperation and Acknowledgements ...... 3

List of Abbreviations ...... 4 SYNOPSIS ...... 11 Summary ...... 12 1 Introduction and background ...... 19 1.1 Background ...... 19

1.1.1 TEN-T and Comprehensive and Core Network ...... 19

1.1.2 Road safety in the region ...... 20

1.2 Project Purpose and Objectives ...... 22

1.3 Activities carried out ...... 22

1.4 The Team of Non-Key Experts (NKE) ...... 23

2 Component 1 - Road Safety Inspections ...... 25 2.1 Existing Core and Comprehensive road network in Western Balkans ...... 25

2.2 Road Safety Inspections implemented in the past ...... 26

2.3 Three-Year Plan for Road Safety Inspections ...... 29

2.3.1 Methodology scheme ...... 29

2.3.2 Requirements - Qualifications ...... 36

2.3.3 The 3-year plan proposal ...... 37

2.4 Pilot Road Safety Inspections ...... 47

2.4.1 The selection – planning process...... 47

2.4.2 General findings ...... 53

2.5 RSI reporting ...... 64

2.6 Conclusion ...... 64

3 Component 2 - Road Safety Audit ...... 66 3.1 List of Rehabilitation and Construction Projects ...... 66

3.1.1 Albania - List of Rehabilitation and Construction Projects identified ...... 68

3.1.2 Bosnia and Herzegovina (Federation of Bosnia and Herzegovina and Republic of Srpska) - List of Rehabilitation and Construction Projects identified ...... 70

Page 7

3.1.3 The former Yugoslav Republic of Macedonia - List of Rehabilitation and Construction Projects identified ...... 71

3.1.4 Kosovo - List of Rehabilitation and Construction Projects identified ...... 72

3.1.5 Montenegro - List of Rehabilitation and Construction Projects identified ...... 73

3.1.6 Serbia - List of Rehabilitation and Construction Projects identified ...... 74

3.2 Road Safety Audit Plan for identified projects ...... 75

3.2.1 Albania – Plan for Road Safety Audits ...... 76

3.2.2 Bosnia and Herzegovina – Plan for Road Safety Audits ...... 77

3.2.3 The former Yugoslav Republic of Macedonia– Plan for Road Safety Audits ...... 78

3.2.4 Kosovo – Plan for Road Safety Audits ...... 79

3.2.5 Montenegro – Plan for Road Safety Audits ...... 80

3.2.6 Serbia – Plan for Road Safety Audits...... 81

3.2.7 Three-year plan by Corridor and Route ...... 82

3.2.8 Yearly allocation of resources ...... 85

3.3 Pilot Road Safety Audits per WB6 Regional Participant ...... 85

3.3.1 Format for Missions to WB6 Regional Participants ...... 87

3.3.2 RSA Team Mission Reports ...... 87

3.4 Conclusions ...... 107

4 Component 3 - Road Map for establishing system for continuous road crash data collection .... 108 4.1 Background and Terminology of Crash Data Base Systems ...... 108

4.1.1 Data Collection ...... 109

4.1.2 Form Content and Quality ...... 109

4.1.3 International Recommendations ...... 110

4.1.4 Data Capture ...... 111

4.1.5 Data Collection Processes ...... 111

4.1.6 Under Reporting ...... 112

4.1.7 Data Linking ...... 112

4.1.8 Geographical Information Systems (GIS) Data ...... 113

4.1.9 Additional Data for Road Safety Analyses ...... 114

4.1.10 In Depth Crash Data ...... 114

Page 8

4.2 Findings on crash databases systems from previous studies ...... 116

4.2.1 Albania ...... 116

4.2.2 Bosnia and Herzegovina ...... 116

4.2.3 The former Yugoslav Republic of Macedonia ...... 117

4.2.4 Kosovo ...... 117

4.2.5 Montenegro ...... 118

4.2.6 Serbia ...... 118

4.3 Current road crash data collection-analysis systems ...... 119

4.3.1 Albania ...... 119

4.3.2 Bosnia and Herzegovina ...... 120

4.3.3 The former Yugoslav Republic of Macedonia ...... 120

4.3.4 Kosovo ...... 121

4.3.5 Montenegro ...... 121

4.3.6 Serbia ...... 122

4.4 Concept for a common system in WB6 based on EU practice ...... 123

4.4.1 Data collection ...... 123

4.4.2 Data linking and sharing ...... 123

4.4.3 Data Analysis ...... 124

4.5 Road Map ...... 125

4.5.1 Standardised Data...... 125

4.5.2 CADaS Dataset ...... 126

4.5.3 Regional Variations and Omissions by CADaS ...... 136

4.5.4 Additional Data ...... 136

4.5.5 Data Encoding ...... 137

4.5.6 Mandatory Fields ...... 138

4.5.7 Data Quality ...... 139

4.5.8 Data Sharing ...... 139

4.5.9 Data Analysis ...... 140

4.5.10 Computer programs ...... 143

4.5.11 Recommendations ...... 144

Page 9

4.6 Conclusions and recommendations ...... 146

APPENDIX A: Relevant Documentation ...... 150 APPENDIX B: Crash Report Forms ...... 153 APPENDIX C: Questionnaire ...... 165 APPENDIX D: Memorandum of Understanding...... 170 APPENDIX E: Example of medical forms ...... 173

Page 10

SYNOPSIS

Project (sub-project) Title: Preparation of Road Safety Inspection (RSI) and Audit (RSA) Plans for core/comprehensive network in Western Balkans (WB6) and Pilots Final Report (consolidated)

Project Code: CONNECTA-TRA-CRM-REG-01

Area: Connectivity Transport Reform Measures in WB6

Contracting Authority: European Commission - DG NEAR

Main Beneficiary/Monitoring: SEETO

End Beneficiaries: Albania, Bosnia and Herzegovina, the former Yugoslav Republic of Macedonia, Kosovo, Montenegro and Serbia

Context: Regional

Consultant: Mott MacDonald Ltd. (UK) in Consortium with COWI A/S, WYG, CeS COWI, TRENECON, SYSTEMA

Administrative Order: 31 May 2017 (supplementary on 26 June 2017)

Mobilisation of NKEs: 05 June 2017 (Kick-off Meeting with SEETO at 13 June 2017)

Sub-Project Duration: 12 months

Anticipated completion: 30 June 2018

Responsible Transport KE: Kostas Georgiou

Page 11

Summary

The road safety directive is implemented on the Trans-European Network in EU and is recommended to be implemented on the remaining national networks as well. The SEETO Regional Participants should, as part of the accession process, implement the directive as well. Since many of the SEETO Regional Participants will implement and improve the road network in the coming years, it would make sense to implement road safety audits as soon as possible in the SEETO Regional Participants. Additionally, in a meeting in Ljubljana, on 25 April 2018 a Road Safety Declaration was endorsed by Transport Ministers / representatives of the Western Balkans.

The purpose of this TA was to prepare short-term plans (2018-2020) for road safety inspections and audits for the Core and Comprehensive Road Network in the Western Balkans. This consultancy also delivered a part of these plans, as pilots, in 2018. The specific objectives of this TA were to:  Prepare three-year RSI plan for the core and comprehensive network and pilot RSIs on high accident sections  Help to ensure that road safety audits are carried out according to the Directive 2008/96/EC on all projects on the core and comprehensive network and undertake sample audits  Support Regional Participants in establishment of a national system for continuous road crash data collection (by 2018).

Road safety inspections

The RSI component covers four activities:

 Map existing core and comprehensive road network in Western Balkans.  Compile a list of all Road Safety Inspections that have been implemented during the last 3 years (2014-2016), including those that followed the EuroRAP/iRAP inspection methodology.  Prepare a three-year Plan (2018-2020) for road safety inspection for the core and comprehensive road network.  Undertake road safety inspections using SEETO’s road safety inspection guidelines on 10% (about 550 km (actually 580 km were carried out)) of the core and comprehensive road network that is considered highest risk portion of the network based on fatal crash data.

The total length of the SEETO comprehensive and core network is 5,462 km of which the core road network is 3,522 km and of the Comprehensive road network is 1,940 km long.

A list of road safety inspection carried out in the region was prepared and used as basis for preparation of the three-year Plan (2018-2020) for road safety inspection for the core and comprehensive road network. The 3-year RSI plan incorporated EuroRAP-IRAP and traditional RSI (as per SEETO Guidelines), as appropriate.

The following table summarises the required financial resources on a yearly basis, expressed in Euro for complementing inspections in the SEETO core/comprehensive road network. The necessary budget is indicative according to average cost values per road length and per methodology followed.

Page 12

Year 1 Year 2 Year 3 Year 4 Total EuroRAP 21,500 21,500 ALB iRAP 0 ‘traditional' RSI 33,000 33,000 34,000 100,000 EuroRAP 26,500 26,500 BiH iRAP 99,000 99,000 ‘traditional' RSI 135,000 100,000 58,000 293,000 EuroRAP 9,000 9,000 MKD iRAP 0 ‘traditional' RSI 34,000 35,000 35,000 104,000 EuroRAP 21,000 21,000 KOS iRAP 0 ‘traditional' RSI 110,000 85,000 52,000 247,000 EuroRAP 16,000 16,000 MNE iRAP 64,500 64,500 ‘traditional' RSI 95,000 65,000 30,000 190,000 EuroRAP 43,500 43,500 SRB iRAP 160,500 160,500 ‘traditional' RSI 210,000 165,000 99,000 474,000

EuroRAP 137,500 0 0 0 137,500 Total iRAP 324,000 0 0 0 324,000 ‘traditional' RSI 177,000 593,000 451,000 187,000 1,408,000

A total of 580 km Road Safety Inspections were carried out in the WB6 Regional Participants leading to preparation of 24 individual pilot reports.

The RSI reports produced as part of this assignment may also be a useful tool for participants to give some guidance on methodology, types of safety hazards, etc.

The Road Safety Inspections carried out in the WB6 Regional Participants showed overall lack of maintenance leading to many of the road safety problems.

In the RSIs carried out, there were many common issues shared within all Regional Participants. This includes the overall lack of maintenance which require urgent action. Crash barriers (missing, inadequate, damaged, etc.), property accesses and high operating speed vehicles passing through villages are the most common hazards identified in all Regional Participants.

Consistent use of unsafe barrier terminals, short barrier lengths, missing barriers, gaps in barrier and outdated bridge parapet implementation were common issues identified in all inspections. Commercial and residence accesses, especially at the single carriageway roads (the non-motorway), are safety hazards that need urgent confrontation. Junctions and interchanges are usually with many problems, outdated design and missing elements. Roads passing through built-up areas or areas with concentrated pedestrian movements do not have adequate infrastructure elements for vulnerable road users.

Following the completion of the pilot RSIs missions undertaken in each of the regional participants, the opinion from each participant is that the missions were very valuable in raising awareness of the inspection process and its benefits.

Page 13

Road safety audits As many roads are being upgraded and constructed on the SEETO core and comprehensive network, among others to ensure improved road safety, it is important to include road safety audits in the design and construction process. In addition, all roads being part of the SEETO core and comprehensive network may become part of TEN-T and then RSA is mandatory according to EU Directive2008/96/EC. RSA should be done on both upgrading projects and new projects.

The three activities contained within the RSA component were:  Compile a list of all expected rehabilitation and new construction road projects on the core and comprehensive road network that are currently at the concept or preliminary design stage in the six Western Balkans Regional Participants  Prepare a plan to undertake road safety audits, at various stages as per Directive 2008/96/EC and SEETO’s Road Safety Audit Guidelines for the list of projects prepared  Undertake Road Safety Audits for a sample of six projects, one in every SEETO Regional Participant.

A list of all expected rehabilitation and new construction road projects on the core and comprehensive road network that are currently at the concept or preliminary design stage in the six Western Balkans Regional Participants was compiled. This was based on input from e.g. Regional Participants, SEETO reports, WBIF, etc. This formed the basis for the plan for audits.

A plan to undertake road safety audits at various stages according to Directive 2008/96/EC and SEETO’s Road Safety Audit Guidelines for the list of projects identified was prepared. The plan includes the Audit stages that are required, as well as the required Auditors inputs and indicative costs, The following table summarises the required financial resources (in Euros) on a yearly basis for implementing RSAs in each Regional Participant according to the plan.

Year 1 2018 Year 2 2019 Year 3 2020 Year 4 2021 Total ALB 300,000 - - 450,000 750,000 BiH 144,000 18,000 - - 162,000 MKD 27,000 9,000 - - 36,000 KOS 33,000 9,000 18,000 18,000 78,000 MNE 332,400 256,200 90,000 18,000 696,600 SRB 291,000 180,000 - - 471,000 Total 1,127,400 472,200 108,000 486,000 2,193,600

Road Safety Audits for a sample of six projects were carried out in each SEETO Regional Participant. The audits carried out at preliminary design phase also demonstrated that potentially unsafe design options can be modified at an early stage and will deliver the largest potential safety benefits. Therefore, the audit team were able to show participants that early engagement of auditors in the design process is the most effective.

In the six audits carried out, there were some common issues shared within all Regional Participants. This has highlighted that certain areas of design require urgent review. Crash barrier design is one area where all participants agreed new standards and guidance for designers was urgently needed. Consistent use of unsafe barrier terminals, short barrier lengths, gaps in barrier and outdated bridge parapet design were common problems in all audits. Junction and interchange design was also a common issue and is outdated and not in line with international best practice.

Following the completion of the missions to undertake pilot RSAs in each of the regional participants, the opinion from each participant is that the missions were very valuable in raising awareness of the audit process and its benefits. In several Regional Participants, there was the view that audits result in large increases in project costs. Through the pilots (6 projects, one per Regional Participant in design stage), the audit team successfully demonstrated that this is not the case and very often, recommendations from audits may involve cost savings in design and construction.

Page 14

Although at the time of writing no designers’ responses have been received from the submitted audit reports, verbal feedback from all participants was that every effort would be made to incorporate the recommendations from the audit reports into the designs.

The missions also identified that there is real enthusiasm to try and develop internal audit capabilities within the region. Serbia and Kosovo were particularly interested in how they might best manage the audit process in the future and how they might select, train and certify auditors in the future.

The audit plans produced as part of this assignment will also will be a useful tool for participants to give some guidance on timing and appropriate costs for the different stages.

Road Map for establishing system for continuous road crash date collection

This activity is intended to present the components and progression that must be implemented in the development of a comprehensive crash database system. The phrase “Crash Database Systems” covers all the elements which constitute the methods and arrangements to collect, store and analyse any systematic report or information collected on road collisions and those injured in them (WHO 2010). This definition therefore includes the stakeholders, which are any persons involved with the system in any capacity. Generally, when Crash Data Systems are considered the focus tends to be on the IT systems primarily (associated computer hardware and software). The three activities contained within this component are:  Assess current road crash data collection-analysis systems.  Set up a concept for a common system in WB6 based on EU practice.  Prepare road map for establishing national system for continuous road crash data collection and analysis.

A number of previous reports and pilot studies have been undertaken, the most recent being the SAFEGE study which provided a preliminary assessment of the crash database systems within the WB6 Region. The study used the basic CADaS datasets to assess the current status of the crash data collection; an overview of the findings has been included within this report for clarity.

An updated assessment of the current status of crash data collection within the WB6 Region is provided within the report. It was found that some of the Regional Participants have significantly developed their data collection beyond that found during the SAFEGE study. All Regional Participants reported funding was the main hindrance to progress.

The core focus of the component is to develop a road map for the sequence of activities that must be undertaken by the WB6 Regional Participants to enable them to achieve a common approach to crash data collection, analysis and dissemination of statistical crash to all interested parties. The core components that must be addressed, in order of priority are:

1. Standardise data collection a. Achieve full compliance with the Advanced CADaS datasets

Page 15

d. Develop a review protocol for the encoded data to ensure there are no errors or omissions present. e. Work towards achieving a basic requirement that 100% of the data fields must be completed before the crash report can be assigned as complete and closed.

4. Data Quality a. Develop a quality assessment protocol to rectify errors and omissions 5. Data sharing within a Regional Participant a. Develop a ‘real time’ sharing capability of all the CADaS statistical data sets with all the relevant Ministries and Road Safety Agencies within Regional Participants. b. The signing of a memorandum of understanding (MOU) between all the parties involved outlining the exchange mechanism, security protocols and timelines. c. Procurement of a computer linking database or ‘HUB’ which will facilitate the exchange of data between each of the data sources d. Procurement of an advanced analytical and GIS database to permit all the end users the ability to view and analysis the statistical data. 6. Data sharing with WB6 a. It is proposed all the CADaS statistical datasets is shared with neighbouring WB6 Regional Participants and SEETO. To achieve this sharing will require: i. The drafting of a memorandum of understanding (MOU) between all the parties involved outlining the exchange mechanism, security protocols and timelines. ii. Each recipient will require a computer linking database or ‘HUB’ which will facilitate the exchange of data between each of the data sources. iii. An advanced analytical database to permit the end user to view and analysis the crash records provided 7. Data Analysis a. The need to procure an ‘off the shelf’ advanced analytical and GIS based system.

The table below presents a summary of the actions and recommendations required to be addressed to achieve a common approach to creating a quality crash data recording system.

Activity Actions Components An agreement that all the advanced CADaS It is proposed the WB6 Reginal Standardise statistical version 3.6 2017 datasets will be adopted Representatives sign a formal agreement to crash datasets within WB6 regional as the de-facto crash data use the Advanced CADaS datasets as a recording convention. requirement It is proposed a separate statistical crash data Enhancement of paper reporting form in line reporting form should be created to with Advanced CADaS datasets. encompass all the advanced CADaS datasets Use CADaS reference codes as de-facto Translation of advanced CADaS datasets and dataset identification and develop an agreed manual into local language translation for each dataset Elements:

Standardise statistical  Identification of lead agency responsible for crash reporting form completion of statistical crash report  Identification of agency /officer / department responsible for completing Produce protocol for completion statistical each sections of report. report form  Maximum time periods permitted when form / sections should be completed  Protocol to quality audit paper form and require lead agency / officer / department to correct omissions and errors identified

Page 16

Activity Actions Components Develop a reference guide for the completion of the crash report form in the local language Produce training manual in local language that based on the manuals provided by EU. provides an explanation of the CADaS dataset Provide training to personnel on how to complete the crash report form There will be a need to either:  Upgrade the existing database datasets to Enhancement of computer database datasets match statistical crash data form or to enable encoding of new statistical datasets  Procure / develop a separate crash database capable of encoding the statistical crash data form Elements:  Identification of lead agency responsible for encoding the statistical crash data Encoding statistical  Maximum time periods permitted when crash data into Produce a protocol outlining the methodology initial record is encoded – within 24 hours computer database for:  Maximum time periods when record should  Encoding the statistical crash data into the be completed database. o damage only 7 days  Quality audit of data recorded o Serious injury 30 days  Protocol for marking record as complete o Fatal defined on an incident by incident and closed bases at least 30 days.  Protocol to identify omissions and errors  Protocol to rectify errors and omissions  Protocol to close a record as complete. Develop a multiagency approach to providing It is proposed each Region within the WB6 information for the statistical crash report form: produce a formal agreement with the various  Medical Ministries to agree on a protocol for the o Ambulance dissemination of data associated with a road o Trauma centre crash.  Fire and Rescue  Identify roles and responsibilities of each  Engineering agency o GIS mapping o Traffic flow and speed data  Develop a protocol for the provision of o iRAP data data. Medical data associated with a casualty involved in a crash is referenced to the casualty’s name and date of birth. Multi-Agency approach Ministry of Health to develop a protocol with The Police database also contains the names the Ministry of Interior (police) with respect to and date of births of the casualties involved in the provision of trauma data from ambulance a collision. and trauma centre for casualties. Proposal is to encode the medical casualty data into the police database to enable easy linking of data. Regional agreement on adopting the Maximum It is proposed each Region within the WB6 Abbreviated Injury Scale (MAIS) casualty formal agree to adopt MAIS injury coding coding system within WB6 Region system. This will require a computer system capable of Ministry of Infrastructure to develop a protocol displaying crash data as an overlay within a to provide access to GIS mapping GIS map Achieve a real time sharing of all the CADaS statistical data sets with all the relevant

Ministries and Road Safety Agencies within a Data Sharing Regional Participant Drafting of a memorandum of understanding An example MOU has been provided in (MOU) between all participants Appendix C

Page 17

Activity Actions Components The actual exchange of data will require the creation of a compatible computer linking Develop the technical capabilities to achieve a database or ‘HUB’ which will facilitate the secure linking capability for the non-sensitive exchange of data between each of the data Data Linking - local statistical crash data between the relevant sources. The central ‘HUB’ database will Ministries combine the data into records associated with each crash and display the information using advanced GIS technology To achieve this sharing will require:  The drafting of a memorandum of understanding (MOU) (Appendix C) between all the parties involved outlining the exchange mechanism, security Develop the technical capabilities to achieve a protocols and timelines. Data Linking - secure linking capability for the non-sensitive Regional statistical crash data between the WB6  Each recipient will require a compatible Regions and SEETO computer linking database or ‘HUB’ which will facilitate the exchange of data between each of the data sources.  An advanced analytical database to permit the end user to view and analyse the crash records provided. Undertaking this style of analysis will require the procurement of an ‘off the shelf’ Analytical Develop an ability to undertake advanced GIS capable crash database. Data Analysis analysis of the statistical crash data within a GIS environment. Many such programs also combine the ability to function as a ‘HUB’ to permit the linking of other data sources.

Page 18

1 Introduction and background

1.1 Background

1.1.1 TEN-T and Comprehensive and Core Network

The TEN-T Regulation 1315/2013 forms the current legal basis for the development of the Trans- European Networks (TEN-T)1. The European Commission has concluded that the TEN-T network would best be developed through a dual-layer approach consisting of a Comprehensive Network and a Core Network.  The Comprehensive Network constitutes the basic layer of the TEN-T. It consists of all existing and planned infrastructure meeting the requirements of the TEN-T Guidelines. The Comprehensive Network is to be in place by 31 December 2050.  The Core Network is a focused sub-set of the Comprehensive Network, overlaying it, to connect the strategically most important nodes, hubs, and links/routes of the Comprehensive Network. Figure 1-1 Comprehensive/Core Network in the Western Balkans.

Therefore, only parts of the Comprehensive network are selected for the Core Network, which are essentially the components of TEN-T with the highest European added value in terms of addressing cross border missing links, key bottlenecks, and multi-modal nodes. The Core Network is to be in place by 31 December 2030.

1 Recently amended (Commission Delegated Regulation (EU) 2016/758 of 04.02.2016)

Page 19

In a future EU enlargement, the transport networks of future Member States would be required to be integrated into the EU TEN-T Network at any given time. Coherence between network development and compliance with EU regulations would undeniably enhance the integration process.

The Western Balkans Comprehensive Network is strategically located with regard to the European Transport system. It constitutes a physical transport corridor that enables the continuity of different parts of the TEN-T Network, providing connections for the Central European Countries to the Black Sea and further beyond to Asia. In June 2015, the transport infrastructure related Ministries of the WB6 and the European Commission (DG NEAR and DG MOVE) indicatively identified the main transport axes that will be connected to the existing TEN-T Core Network Corridors2. This was carried out in accordance with the application of the “Planning methodology for the trans-European transport network (TEN-T)3)”, which sets out many of specific criteria to identify the network’s Core nodes and, subsequently, Core links in terms of connecting Core nodes.

The WB6 agreed on the alignment of their core transport networks, which will be developed in line with EU recommendations. Independent of their anticipated future membership of the EU, these Regional Participants are already moving towards improving their transport systems in terms of both infrastructure and operational measures.

Furthermore, in June 2015 during the TEN-T Days in Riga, three of the nine identified Core Network Corridors (CNC) were extended for the Western Balkans. The three identified CNCs are:  the Orient-East Mediterranean (OEM) Corridor which connects central Europe with the maritime interfaces of the North, Baltic, Black and Mediterranean Seas  the Mediterranean (MED) Corridor which links the Iberian Peninsula with the Hungarian- Ukrainian border  the Rhine/Danube Corridor which provides the main East–West link between continental European countries, connecting France and Germany, Austria, the Czech Republic, Slovakia, Hungary, Romania, and Bulgaria all along the Main and Danube rivers to the Black Sea.

1.1.2 Road safety in the region

The South East Europe Region has a high road crash rate compared to EU countries with the 6 SEETO Regional Participants having almost 84 road deaths per million population in 2016 compared to the EU28 at just over 50 road deaths per million of population. In 2016, more than 1,500 were killed and almost 55,000 were injured in the SEETO Regional Participants according to MAP20184. The road safety reform progress around the WB6 varies but is generally low. The EU Directive 2008/96/EC is not (or only partly) transposed in national legislations.

2 As considered by Article 8 of the Regulation (EU) 1315/2013. The indicative extension of the TEN-T Network to the Western Balkans Region is articulated in EC Regulation 2016/758 which amended the TEN-T Regulation. 3Building the Transport Core Network: Core Network Corridors and Connecting Europe Facility {COM (2013) 940 final}

4 SEETO Multi-Annual Development Plan, Multi-Annual Plan 2018, Common problems – Shared solutions

Page 20

Figure 1-2 Number of fatalities in the regional participants. (From MAP 20184)

Since many of the SEETO Regional Participants will implement and improve the road network in the coming years, it would make sense to implement road safety audits as soon as possible in the SEETO Regional Participants. Road safety audit is a systematic and independent assessment of a road or traffic project from a road safety point of view. Since it focuses on planning, design and around opening of the roads, it is a proactive management tool to identify safety deficiencies of a project and improve the project before the road is actually implemented and before the accidents occur.

An EU financed project was launched in 2008, “Support for Implementing Measures for South East Europe Core Regional Transport Network Multi Annual Plan (MAP) 2008-2012”. The project focused on railways and road safety auditing and was finalised in 2009. The project assisted the Regional Participants in implementing MAP soft measures in the Roads sub sector, with a focus on Road Safety Audit and Inspection and Regional Road Safety Strategy. The main results on road safety auditing were:  Proposal for law and regulations to implement mandatory RSA  A draft Road Safety Audit Manual and check lists  A draft RSA Agreement and an Action Plan for its implementation  Testing RSA procedures through pilot projects with reference to design documents and existing roads on the SEETO network  Draft short term Regional Road Safety Strategy.

Page 21

As part of the project, three workshops were carried out covering the above topics.

To initiate the implementation of road safety audits in the SEETO Regional Participants, in 2014 SEETO wanted to offer to help facilitate the process by providing training in road safety audits for key persons in SEETO Regional Participants who could then be spearheads to implement road safety auditing systems in their countries. The training was built on the basis of the deliverables of the project from 2008 and the TA assistance to SEETO in 2014 to help prepare curriculum and training material as well as a handbook in Road Safety Principles. A pilot training programme following the principles of the developed curriculum and training material was financed by TAIEX and brought in in 2014 and 2015.

In 2016, Road Safety Inspection Guidelines were developed along with its training curriculum and one training session has been delivered.

In a meeting in Ljubljana, on 25 April 2018 a Road Safety Declaration was endorsed by Transport Ministers / representatives of the Western Balkans. This declaration suggested actions for:

 strengthening road safety management  promote safer infrastructure  promote the protection of road users  promote the use of safer vehicles  enhance cooperation and exchange of experience.

1.2 Project Purpose and Objectives

The preparation of Road Safety Inspection (RSI) and Audit (RSA) Plans for core/comprehensive network in Western Balkans (WB6) and Pilots Project commenced with a Kick-off Meeting on the 13 June 2017 with an expected project duration of 12 months.

The purpose of this TA is to prepare short-term plans (2018-2020) for road safety inspection and audit for the whole Core and Comprehensive Road Network in the Western Balkans. This consultancy will - as RSI/RSA pilots - also deliver a part (10% and 6 projects, respectively) of these overall plans in 2018.

The objective is to provide direct support to the Western Balkans’ ministries responsible for transport and infrastructure and to road authorities for programming infrastructure maintenance and to assist the SEETO Secretariat in monitoring the implementation of relevant transport measures in the framework of Connectivity Agenda.

The specific objectives of this TA are to support the implementation of the 2nd, 3rd and 4th road safety measures under the CRMMP for 2016/2017: . Prepare three-year RSI plan for the core and comprehensive network and pilot RSIs on high accident sections . Help to ensure that road safety audits are carried out according to the Directive 2008/96/EC on all projects on the core and comprehensive network and undertake sample audits . Support Regional Participants in establishment of a national system for continuous road crash data collection (by 2018).

All project activities should be located on the core and comprehensive network. 1.3 Activities carried out

The activities and their completion period are indicated in the table below for each of the components.

Page 22

Activities Started Completed (date) (date) Mobilization – Inception period 05 June 2017 30 June 2017 1.1 Map existing core and comprehensive networks in WB6 13 June 2017 27 March 2018 List of undertaken RSI in 2014-2016 including recommendations 1.2 13 June 2017 28 February 2018 provided Preparation of 3-year RSI Plans (2018-2020) under two methods 1.3 01 November 2017 28 May 2018 and estimate of inputs-cost Undertake RSI on 10% (high risk) of core/comprehensive 1.4 30 July 2017 24 April 20/18 network List of current (rehabilitation-new construction) road projects in 2.1 13 June 2017 27 March 2018 preparation Preparation of RSA Plan (as per 2008/96/EC and SEETO RSA 2.2 15 October 2017 29 March 2018 Handbook) and estimate of inputs-cost Undertake RSA for sample of 6 projects, one per SEETO 2.3 01 October 2017 16 March 2018 member Undertake missions in WB6 to assess current road crash data 3.1 28 August 2017 28 February 2018 collection-analysis systems Set up a concept for common system in WB6 based on EU 3.2 10 December 2017 30 April 2018 practice Prepare road map for establishing national system for 3.3 15 January 2018 30 April 2018 continuous road crash data collection and analysis

The list of deliverables is presented in the following table.

Deliverables Submitted 1. Inception Report 30 June 2017 Interim Report for Component 1 (RSI): Inventory of network and conducted RSI and 2. 08 December 2017 RSI Plan 2018-2020 3. Interim Report Component 2 (RSA): Inventory of projects and RSA Plan 23 November 2017 4. Draft Report for Component 1 (RSI) 27 March 2018 5. Draft Report for Component 2 (RSA) 17 April 2018 6. Draft Report Component 3 (Accident data) 30 April 2018 7. Final Report for Component 1 (RSI) 31 May 2018 8. Final Report Component 2 (RSA) 15 May 2018 9. Final Report Component 3 (Accident data) 15 June 2018 10. Final consolidated report (all components) 29 June 2018

1.4 The Team of Non-Key Experts (NKE)

The Team of Non-Key Experts, established for the scope of this specific project, consists of eight (8) Senior Experts, presented in the following table, as per the AO. A replacement took place of position 7 (Road Safety Inspector 3).

Page 23

Table 1-1 Team of Experts

Position in ToR Name Category in Financial Proposal 1 Project Manager Jesper Mertner SNKE 2 Lead Road Safety Auditor Matt Chamberlain SNKE 3 Road Safety Auditor 2 Lárus Ágústsson SNKE 4 Road Safety Auditor 3 Darko Cvoric SNKE 5 Lead Road Safety Inspector Stelios Efstathiadis SNKE 6 Road Safety Inspector 2 Krsto Lipovac SNKE 7 Road Safety Inspector 3 Mostafa Naser replaced by Srbislav Gugleta SNKE 8 RS Data Specialist Mike Fell SNKE

This team was supplemented by four local SNKE which supported (horizontally) all three, on-going, connectivity reform sub-projects (ITS, Maintenance Plans and Road Safety), as follows:  Emiljano Zhuleku for Albania and Kosovo  Jovan Hristoski for the former Yugoslav Republic of Macedonia  Amna Redzepagic for Bosnia and Herzegovina  Dusan Savkovic for Serbia and Montenegro.

The project was organised as shown below.

Figure 1-3 Project Team organisation.

Page 24

2 Component 1 - Road Safety Inspections

This section presents the findings of the RSI component including the three-year plan (2018/19- 2020/21) for road safety inspection of the SEETO core and comprehensive road network and some findings from the pilot studies.

The scope is to prepare a three-year RSI plan for the core and comprehensive network and pilot RSIs on high accident sections. The four activities contained within this component are:

Activity 1 Map existing core and comprehensive road network in Western Balkans.

Activity 2 Compile a list of all Road Safety Inspections that have been implemented during the last 3 years (2014-2016), including those that followed the EuroRAP/iRAP inspection methodology.

Activity 3 Prepare a three-year Plan (2018-2020) for road safety inspection for the core and comprehensive road network. This plan recommends on the use of two methods: i) traditional road safety inspection using SEETO’s road safety inspection guidelines, and ii) EuroRAP road survey and star rating methodology. The Plan includes the required inspectors’ inputs and an indicative cost to deliver it by consulting firms.

Activity 4 Undertake road safety inspections using SEETO’s road safety inspection guidelines on 10% (about 550 km – in reality 580 km were undertaken) of the core and comprehensive road network that is considered highest risk portion of the network based on fatal crash data. SEETO members prepared a list of their high risk sections and the Consultant then compile a list of about 550 km (in reality 580 km) of these high risk roads by maintaining a reasonable distribution among all SEETO’s members. The decision on the specific sections to be inspected was taken jointly with SEETO and by also taking into account existing and foreseen TAs for RSI.

2.1 Existing Core and Comprehensive road network in Western Balkans

The total length of the SEETO comprehensive and core network is according to MAP20185 5462 km of which the Core road network is 3,522 km and of the Comprehensive road network is 1,940 km long. On the entire network, approx. 2,198 km are Corridors and approx. 3,264 km are Routes.

The SEETO road network consists of Corridors and Routes, as follows:

Corridor Vc (400 km): CRO border/Bosanski Samac (BIH) – Sarajevo (BIH) – Doljani/CRO border Corridor VIII (657 km): Tirane/ Durres/ Vlore (ALB) – Skopje (MKD) – Deve Bair/BG border Corridor X (726 km): CRO border /Batrovci –Belgrade (SRB) – Skopje (MKD) – Bogorodica/GR border Corridor X B (185 km): HU border/ Horgos—Novi Belgrade (SRB) Corridor X C (110 km): Nis (SRB) —Gradina/BG border Corridor X D (117 km): Veles (MKD) —Medzitlija/ GR border Route 1 (147 km): CRO border/Neum Northwest – Neum (BIH) –Bar (MNE) Route 2a (228 km): CRO border/Gradiska – Banja Luka (BIH) – Lasva (BIH) Route 2b (395 km): Sarajevo (BIH) – Podgorica (MNE) – Vore (ALB)

5 SEETO Multi-Annual Development Plan, Multi-Annual Plan 2018, Common problems – Shared solutions

Page 25

Route 2c (125 km): Fier (ALB) —Kakavija/GR border Route 3 (185 km): Sarajevo (BIH) —Uzice (SRB) Route 4 (601 km): Romanian border/Vatin – Belgrade (SRB) – Podgorica (MNE) – Bar (MNE) Route 5 (213 km): Cacak (SRB) – Krusevac (SRB) – Paracin (SRB) – Vrska Cuka/BG border Route 6a (259 km): Ribarevina (MNE) – Ribarice (SRB) – Pristina (KOS) – Skopje (MKD) Route 6b (205 km): Pristina (KOS) –Peje/Pec (KOS) – Kolasin (MNE) Route 7 (314 km): Lezhe (ALB) – Pristina (KOS) – Doljevac (SRB) Route 8 (78 km): Podmolje (MKD) – Bitola MKD) Novi Sad (SRB) – Ruma (SRB) – Loznica (SRB)/Zvornik (BIH) – Tuzla (BiH) – Doboj (BiH) – Route 9a (305 km): Banja Luka (BiH) Route 10 (142 km): Miladinovci (MKD) – Stip (MKD) – Novo Selo (MKD)

R4 X R9A Xb R4 R4 X X R9A R9A R4 R2A R9A Vc R4 X R2A R5 R3 R4 R3 R5 X Xc R2B R7 Vc R4 X Xc R6A R7 R2B X R6B R4 R1 R2B R6A R1 R7 VIII R4 R2B R10 R7 X R10 R1 VIII Xd X R8 VIII Xd VIII

VIII R2C

The following map shows the Core and Comprehensive network of the Western Balkans (WB6). 2.2 Road Safety Inspections implemented in the past

Information has been gathered regarding the Road Safety Inspections that have been implemented during the last 3 years, including those that followed the EuroRAP/iRAP inspection methodology.

Page 26

The following Table incorporates all information received from the Regional Participants. Thus, the road sections that do not have RSI will be the starting point for the 3-year plan proposal.

Corridors Past 3years All sections SEETO Network /Routes RSI iRAP Muriqan(MNE border) - Koplik Core Route 1 Yes Koplik - Skhoder Core Route 1 Skhoder - F. Kruje Core Route 1 Yes F. Kruje - Lezhe Core Route 1 Hani i Hotit (MNE border) - Fush Kruje Comprehensive Route 2b Fush Kruje - Vore Comprehensive Route 2b Yes

Rrogozhine - Fier Core Route 2c Yes Fier - Tepelene Core Route 2c ALB Tepelene - Kakavia (GR border) Core Route 2c Yes Morine Vermice (KOS border) - Lezhe Core Route 7 Yes Qaf Thane (MKD border) - Elbasan Core Corridor VIII Yes Elbasan - Tirane Core Corridor VIII Yes Tirane - Durres Core Corridor VIII Durres - Vlore Core Corridor VIII Fier - Vlore Core Corridor VIII Neum west - Neum South Core Route 1 Gradiska (CRO border) - Banja Luka - Core Route 2a Jajce Jug Jajce Jug - Donji Vakuf Core Route 2a Donji Vakuf - Lasva Core Route 2a Sarajevo - Hum (MNE border) Comprehensive Route 2b Sarajevo - Lapisnica Comprehensive Route 3 Lapisnica - Ljubogosta Comprehensive Route 3 Yes Ljubogosta - Podromanija Comprehensive Route 3

BIH Podromanija - Rogatica Comprehensive Route 3 Yes Rogatica - Vardiste (SRB border) Comprehensive Route 3 Banja Luka - Doboj - Karakaj (SRB border) Comprehensive Route 9a Bosanski Samac (CRO border) - Matuzici Core Corridor Vc Matuzici - Ozimica Core Corridor Vc Ozimica - Topcic Polje Core Corridor Vc Topcic Polje - Sarajevo - Jablanica Core Corridor Vc Jablanica - Potoci Core Corridor Vc Potoci - Doljani (CRO border) Core Corridor Vc

Page 27

past 3years Corridors All sections SEETO Network /Routes RSI iRAP Djeneral Jankovic (KOS border) - Skopje Core Route 6a Yes Podmolje - Bitola Comprehensive Route 8 Yes Miladinovci - Stip - Novo Selo (BG border) Comprehensive Route 10 Yes Kafasan (ALB border) - Skopje Core Corridor VIII Yes Skopje - Stracin Core Corridor VIII Yes

Stracin - Kriva Palanka Core Corridor VIII Yes

MKD Kriva Palanka - Deve Bair (BG border) Core Corridor VIII Yes Tabanovce (SRB border) - Skopje - Core Corridor X Yes Bogorodica (GR border) Veles - Prilep Comprehensive Corridor Xd Yes Prilep - Bitola Comprehensive Corridor Xd Yes Bitola - Medzitlija (GR border) Comprehensive Corridor Xd Yes Brnjak (SRB border) - Veternik Comprehensive Route 6a Veternik - Lipljan Core Route 6a Lipljan - Hani i Elezit ( MKD border) Core Route 6a Kuqishte (MNE border) - Kijeve/Kijevo Comprehensive Route 6b

KOS Kijeve/Kijevo - Gjurgjice/Djurdjice Comprehensive Route 6b Gjurgjice/Djurdjice - Fushe Kosove/Kosovo Comprehensive Route 6b Polje Vermice/Vrbnica (ALB border) - Merdare Core Route 7 (SRB border) Debeli Brijeg (CRO border) - Sukobin (ALB Core Route 1 border) Scepan Polje (BIH border) - Bozaj (ALB Comprehensive Route 2b border) Dobrakovo (SRB border) - Mioska Core Route 4

MNE Mioska - Podgorica Core Route 4 Podgorica - Bar Core Route 4 Ribarevine - Dracenovac (SRB border) Comprehensive Route 6a Kolasin - Kula (KOS border) Comprehensive Route 6b

Page 28

past 3years Corridors All sections SEETO Network /Routes RSI iRAP Uzice - Kotroman (BIH) Comprehensive Route 3 Vatin (RO border) - Belgrade - Orlovaca Core Route 4 Orlovaca-Stepojevac Core Route 4 Stepojevac-Celije Core Route 4 Celije - Knezevici Core Route 4 Knezevici - Kokin Brod Core Route 4 Yes Kokin Brod - Gostun (MNE border) Core Route 4 Cacak/Preljina - Mrcajevci Comprehensive Route 5 Mrcajevci - Vrnjci Comprehensive Route 5 Vrnjci - Kamidzora Comprehensive Route 5 Kamidzora- Paracin - Vrska Cuka (BG Comprehensive Route 5 border) Spiljani (MNE border) - Brnjak (KOS Comprehensive Route 6a border) Nis - Merosina Core Route 7 Yes

Merosina - Merdare (KOS border) Core Route 7 Novi Sad/Petrovaradin - Sremska Comprehensive Route 9a SRB Kamenica Sremska Kamenica - Irig Comprehensive Route 9a Yes Irig - Ruma Comprehensive Route 9a Ruma - Klenak Comprehensive Route 9a Yes Klenak - Loznica - Mali Zvornik (BIH Comprehensive Route 9a border) Batrovci (CRO border) - Kuzmin Core Corridor X Kuzmin - Sremska Mitrovica Core Corridor X Yes Sremska Mitrovica - Beograd/Dobanovci Core Corridor X Beograd/Dobanovci - Bubanj Potok Core Corridor X Bubanj Potok - Mali Pozarevac Core Corridor X Mali Pozarevac - Presevo (MKD border) Core Corridor X Horgos (HU border) - Feketic Core Corridor Xb Feketic-Sirig Core Corridor Xb Sirig - Beograd/Dobanovci Core Corridor Xb Nis - Gradina (BG border) Core Corridor Xc

2.3 Three-Year Plan for Road Safety Inspections

2.3.1 Methodology scheme

A 3-year plan for road safety inspection of the SEETO Core and Comprehensive Road Network has been prepared. This way, until 2021 the total SEETO Core and Comprehensive Road Network will have been inspected according to the SEETO manual for Road Safety Inspections, which actually is in line with the EC 2008/96 Directive.

As an initial stage of road network assessment, the EuroRAP Risk Mapping procedure is proposed where reliable crash data are easily available. This can then be repeated following e.g. 3-5 years to see what the results and benefits are. Risk mapping is a fast and low-cost process of evaluating road corridors based on reliable crash data, traffic volumes and mapping information. The EuroRAP Risk

Page 29

Mapping is proposed to be adopted only in the case that the required data are readily available and reliable.

In order to prepare the Plan for RSI, first the road safety inspections that have been implemented within the last 3 years will be taken into consideration and the relevant road sections will be excluded. Afterwards, the (pilot) Road Safety Inspections performed under Activity 4 of the Connecta project will be taken into account, as well. Thus, the three-year Plan will be established for the rest of road sections (that have not been inspected).

The approach for the Plan is recommended, as follows:

 The road network under consideration should be surveyed according to the iRAP methodology in order to classify the road sections under one of the five risk rating categories. This procedure could be completed by the end of 2018.

The ‘traditional’ Road Safety Inspection using SEETO’s road safety inspection guidelines should be performed as follows:

 The 1st year for all 1-star road sections  The 2nd year for all 2- and 3-star road sections  The 3rd year for all 4- and 5-star road sections.

This 2-step approach has as scope to quickly proceed with the assessment of the most dangerous road sections. Performing ‘traditional’ RSI requires substantial resources, effort, time and budget. Furthermore, the execution of a large scale (in terms of road length) ‘traditional’ RSI would present outcomes after a considerable time period. Thus, implementing gradually the ‘traditional’ RSI to the Core and Comprehensive road network, according to the available budget, seems the most logical process. Therefore, in order to categorize safety levels of the road network, and give more importance and substance to the most dangerous sections, the initial assessment with the iRAP methodology is the quickest, least expensive and most acceptable process.

2.3.1.1 EuroRAP Road Risk Mapping

Risk Maps are colour-coded maps showing the risk of death and/or serious injury on individual road sections across a road network. This could as mentioned above be used by Regional Participants that have reliable and readily available crash data to compare the before and after situation. If data are not reliable and readily available the Regional Participant should follow the suggested process with iRAP screening and then traditional road safety inspections and not wait for crash data to be available.

Risk Maps are statistically designed to support national road safety strategies and add an extra layer of information alongside existing approaches. Using an international and common basis of measurement that can be used to assess priorities, Risk Mapping identifies the safest and most dangerous road sections within a region or Regional Participant. Comparisons between countries enable benchmarking and progress to be tracked. Knowing where risk has been reduced and the measures that have worked are essential in building best practice and knowledge transfer.

Risk Mapping, by its very nature, relies on the use of historic crash and traffic flow data. As such, when published, some routes may already have had road safety improvements. Others may be more difficult to change and on these roads it is particularly important for road users to be aware that they face higher risks than they might expect. Risk Mapping should therefore be updated at regular intervals to ensure that they represent the most up-to-date picture.

Risk Maps based on crash rates show the combined influence of behaviour, road and vehicle. Rates per vehicle kilometre travelled can show the likelihood of a particular type of road-user (e.g. car driver, motorcyclist, lorry driver), on average, being involved in a road crash.

Page 30

2.3.1.1.1 Road length

To identity road sections which demonstrate differences in general road standard performance, it is essential to assess lengths that minimise the impact of year-on-year variability in crash numbers and present a stable longer term estimate of crash risk.

While it is typical for fatal and serious crash rates to differ from road type group averages, variability is considerably greater for short sections. Where a section comprises a short length between junctions, crash rates may be unrepresentative of average rates, since crashes at junctions will form a disproportionately large contribution to the total in that length. This may be due to a proportionately higher influence from junction crashes at the ends of the short sections.

Sections less than 5 km tends to show greater year-on-year variability in crash numbers, were more likely to change risk rating from one period to another, and were therefore less reliable when compared over time. For motorways and dual and single carriageways these differences were significant up to section lengths of 10 km.

When assessing whether individual sections have fatal and serious crash rates that are above or below average, minimum thresholds of 10 km for motorways and dual carriageways and 5 km for single carriageways should be used as a starting point in assessing crash numbers. Where it is not possible to aggregate short sections, care must be taken when interpreting risk ratings.

2.3.1.1.2 Crash data

EuroRAP protocols focus on fatal and serious crashes. In addition to reflecting the key policy targets across Europe, such crashes reflect the ability of the road design to ‘contain’ the event and are likely to be reported more consistently than those falling in the ‘slight injuries’ and ‘damage only’ categories. They also represent the severity levels generally used in national targets and those that can have life- changing consequences.

EuroRAP Road Risk Mapping presents crashes at the severe end of the crash spectrum and provides adjustment factors to allow for differences between countries in reporting, standard crash definitions and elements such as the quality of medical care that may influence these issues. These factors are based on standard ratios between the number of fatal and serious crashes using values that are reviewed at frequent intervals as new evidence is presented.

Analysis has shown that when crash numbers are compared over time, the general relationship is strong, but the variation in frequency can become large when the numbers fall below 20 crashes per road section over three years.

2.3.1.1.3 Traffic flow

Traffic flow data is used as an exposure measure in expressions of fatal and serious crash risk in EuroRAP Road Risk Mapping.

2.3.1.1.4 Assessment period

A widely misunderstood aspect of road safety relates to the way in which crash numbers vary from year to year and can appear to show a trend that requires urgent attention, only for the trend to reverse a year later. Regression-to-mean, also sometimes called ‘bias by selection’, can complicate evaluations at sites with high crash numbers. Locations chosen for treatment following a year with particularly high numbers, often in practice will tend to reduce in the next year even if no treatment is applied. It is believed that the regression-to-mean effect can over-state the effect of a treatment by 5-30%, depending on the assessment period.

Page 31

A simple way of assessing regression-to-mean and changes in the environment is by using control sites chosen in exactly the same way as the treated sites, identified as having similar problems, but left untreated. In practice, it is difficult to find matched control sites and, if investigated, to justify not treating them.

The effect does, however, tend to be diminished if longer periods of time are selected.

Where crash numbers are insufficient to meet the criteria suggested of 20 per road section over three years, the data period could be extended. However, it should be noted that extending the period beyond three years will increase the likelihood of significant network changes over the period of investigation and therefore a thorough review of planned or potential large scale changes should be assessed at an early stage. Some variance over time may therefore not show up. In some circumstances lack of change in the colour banding of the same road section over time may be a good indicator of the robustness of the technique.

2.3.1.2 The iRAP Method

The protocols are developed by the International Road Assessment Programme (iRAP). iRAP is a registered charity dedicated to saving lives through safer roads. The programme is the umbrella organisation for EuroRAP, AusRAP, usRAP, KiwiRAP and ChinaRAP. Road Assessment Programmes (RAP) are now active in many countries worldwide. iRAP provides tools for inspecting roads and developing Star Ratings, Safer Roads Investment Plans and Risk Maps.

The main objective of the iRAP method is the improvement of the road users’ safety by proposing cost- effective investment plans. The method indicates that the severity of a road crash can be reduced through intervention during the sequence of events happening during this crash. As it is known, an injury crash results from a chain of events, starting with an initial event, probably resulting from several factors, which leads to a dangerous situation. The basic idea is to intervene at any point in this chain, in order to reduce the kinetic energy of all road users involved in the crash to a tolerable level. Such an intervention may not only reduce the number of crashes but also the severity of injuries.

The initial step for the implementation of the RAP method is the inspection and record of the infrastructure elements of a road network, which relate to road safety. The record leads to the quantification of the safety that a road section provides to its users by awarding safety scores (Star Rating Scores). The Star Rating Scores express the safety capacity of a road section in a 5-Stars scale. This quantification aims to identify the most appropriate countermeasures, which will increase the infrastructure’s road safety score. The Safer Roads Investment Plan (SRIP) includes all the countermeasures proved able to provide the greater safety capacity and maximize the benefit over spent cost of the planned investments. Thus, the SRIPs are considered as a valuable tool for the authorities, stakeholders and investors in order to decide for the most cost-effective and efficient road infrastructure investments.

2.3.1.2.1 Measuring the road infrastructure safety

The assessment of road safety requires Road Safety Inspections of the road network sections and the assignment of a safety score to them. The inspection is conducted by visual observation and record of the road infrastructure elements which are related (directly or not) to road safety and have a proven influence on the likelihood of a crash or its severity. The RAP uses two types of inspection: the drive- through and the video-based inspection. During the first one, the record of the infrastructure’s elements is performed manually, with the help of specialized software, while during the second, a specially equipped vehicle is used to provide recorded video to be used for a virtual drive-through of the network and an automated identification of the infrastructure’s elements.

Page 32

Following the RSI, the Road Protection Score (RPS) is calculated. The RPS is a unit-less indicator, which depicts the infrastructure’s safety capacity for each road user type and it is calculated for 100 m road segments. Road user types considered are car occupants, motorcyclists, bicyclists and pedestrians, who may be involved in road crashes. For each road user type and for every 100 m road segmentation the respective RPS is calculated as follows:

RPSn,u = RPS n,u,c = L n,u,c * S n,u,c * OS n,u,c * EFI n,u,c * MT n,u,c cc where “n” is the number of 100 m road segment, “u” the type of road user and “c” the crash type that the road user type “u” may be involved in. The following variables are taken into consideration:

 L: the Likelihood that the “i” crash may be initiated  S: the Severity of the “i” crash  OS: the degree to which risk changes with the Operating Speed for the specific “i” crash type  EFL: the degree to which a person’s risk of being involved in the “i” type of crash is a function of another person’s use of the road (External Flow Influence)  MT: the potential that an errant vehicle will cross a median (Median Traversability).

2.3.1.2.2 The Star Rating process

The aim of the Star Rating process is the award of the “n” 100 m road segments with Stars, depicting the safety offered to each of the “u” road users’ types. The Star Rating system uses the typical international practice of recognising the best performing category as 5-star and the worst as 1-star (5 stars scale), so that a 5-star road means that the probability of a crash occurrence, which may lead to death or serious injury is very low. The Star Rate is determined by assigning each RPS calculated to the Star Rating bands. The thresholds of each band are different for each road user and were set following significant sensitivity testing to determine how RPS varies with changes in road infrastructure elements. The assignment procedure leads to the development of a risk-worm6 chart, which depicts the variation of the RPS score in relation to the position (distance from the beginning) on the road under consideration. The final output of the Star Rating is the Star Rating Maps, in which the “n” road sections are shown with different colour, depending on their Star award (5-star green and 1-star black).

6 Risk Worm is a function of iRAP to focus and identify risky spots. It analyses the risk per section.

Page 33

2.3.1.3 The ‘traditional’ RSI

The ‘traditional’ RSI is performed according to the SEETO guidelines for Road Safety Inspections. The inspection process is presented at the following figure.

START OF THE RSI Client awards RSI to the inspector

Ordering

Client hands over all documents/data to the inspector

Independent RSI report prepared by the inspector Undertaking

Client decides whether to implement RSI report

Client RSI considers: report shows no safety proposed problems measures not accepted Client considers: proposed measures accepted Completion

Client explains the reasons for rejecting proposed measures

END OF THE RSI

The following steps outline the procedure of the inspector’s work:

 STEP 1 Preparatory work in the office  STEP 2 On-site field study  STEP 3 RSI report  STEP 4 Remedial measures and follow-up.

Page 34

2.3.1.3.1 The partners in the RSI process and their roles

The client (usually the road authority or private road operating company) and the inspector (or team of Inspectors) participate in the inspection process.

Order the RSI: Usually, the decision to order an inspection is taken by the client (road authority). However, it may be regulated by a Ministry decision or by law as well as by the financing donor.

Selecting the team: The client commissions the inspector, who can either be an individual or a team. A list of potential inspectors compiled by the client can be helpful for the selection process. It is important to consider including members with experience regarding all aspects of facility maintenance including signage, traffic lighting control, vegetation, snow removal, etc. It may be useful to include a police officer who is experienced in road safety and crash investigation.

2.3.1.3.2 Preparatory work in the office

Background information about the road, the function of the road, the standard of the road and traffic volumes should be obtained as a first step. Information from local residents might prove useful and can be obtained through face-to-face discussions or a questionnaire. The list below provides information about the sort of questions that should be asked and the answers recorded during the preparatory work:  Road function  Traffic situation  Road standards.

The RSI team should have the necessary equipment to perform their tasks.

2.3.1.3.3 Field Study

When an intersection is included in the road to be inspected it is necessary to inspect part of the intersecting road as well (at least the approaches), both by vehicle and on foot. Site inspections should be undertaken over a range of traffic and environmental conditions likely to be encountered. Both night- time and daytime inspections are essential to appreciate the situation. It may also be necessary to view the location at other times of the day (e.g. when school finishes, during peak hours or the weekly market). The core part of the RSI is to identify deficiencies on the road that may cause crashes or could have an influence on the severity of crashes.

2.3.1.3.4 RSI findings and report

The RSI team or expert notes in the report the problems detected and provides recommendation on inventions and measures that will reduce or remove the road safety problem. The RSI team may also do an assessment of the risk, e.g. High, Medium or Low.

2.3.1.3.5 Completion of the RSI

Upon receipt of the report, the client must consider the indicated problems and proposals and make a decision on how or if the proposed measures will be implemented. A ‘Completion meeting’ should be organised in order to finalise the ‘Response to the inspection report’.

2.3.1.3.6 Follow-up and evaluation

For the foreseen periodic RSI of the core network it is very important that a sufficient and effective way for serious follow up is identified.

Page 35

In addition, it could be helpful to conduct some studies at a later time to evaluate the effects of the remedial measures. The road authority could organise such. Behavioural studies should be carried out in the same way and in the same locations as during the investigation. Traffic volumes and speeds should be checked, as well as the traffic environment.

2.3.2 Requirements - Qualifications

There are specific requirements for the realization of this 3-year plan, in terms of capacity/capability and resources.

2.3.2.1 EuroRAP road risk mapping requirements

Members may only use the RAP Road Risk Mapping protocol if they have a licence to do so. The licence gives access to the detailed specifications for data collation and analysis, ensuring consistency of output in form and style. The licence also appoints the Member as principal manager and communications outlet within a territory and gives access to the EuroRAP name and logo for the communication of results. The organisation has specific standards and procedures that should be followed. More detailed information may be found at the official EuroRAP site (www.eurorap.net).

Additionally, it is obligatory that risk maps produced be Quality Assured. This, again, is a procedure according to specific standards.

In terms of performance, it is envisaged that (on average) a Regional Participant’s road network may be risk mapped within one month by an accredited consultant.

2.3.2.2 iRAP requirements

An iRAP may be performed only by a certified supplier of iRAP. The organisation has specific standards and procedures that should be followed. Furthermore, there are specific procedures for someone to be accredited by iRAP. More detailed information may be found at the official site of iRAP (www.irap.org).

Additionally, it is obligatory that all road sections assessed according to iRAP methodology be Quality Assured. This, again, is a procedure according to specific standards.

Finally, there is the availability of uploading all iRAP assessed road section to a cloud-based software (named ViDA), where various analyses may be performed. In order to achieve this, specific training needs to be obtained.

In terms of performance, it is envisaged that an average of 30 km road length may be completed within one working day of the iRAP Team (based on international experience).

2.3.2.3 Traditional RSI

The team performing ‘traditional’ RSI (based on the SEETO Road Safety Inspection manual) should comprise, ideally, 3 persons (team leader and 2 members).

The RSI team should be comprised of experienced road safety engineers, with more than 10 years of experience on the field. The team leader should be a certified road safety inspector and/or auditor.

In terms of performance, it is envisaged that an average of 10 km road length may be completed within one working day of the RSI Team (based on international experience, but also experience that

Page 36

Connecta RSI team gained during pilots in the context of this sub-project). This comprises both field work and reporting.

2.3.3 The 3-year plan proposal

The following Table incorporates all information that has been received by the WB6 Regional Participants. It also contains the RSI conducted by Connecta as “pilots” (approximately 580 km of the network).

Past/ Connect SEETO Corridors / current a pilot Missing All sections Network Routes Past RSI iRAP RSI RSI Muriqan(MNE border) - Koplik Core Route 1 √ Koplik - Skhoder Core Route 1 √ Skhoder - F. Kruje Core Route 1 √ F. Kruje - Lezhe Core Route 1 √ Hani i Hotit (MNE border) - Fush Kruje Comprehensive Route 2b √ Fush Kruje - Vore Comprehensive Route 2b √

Rrogozhine - Fier Core Route 2c √ Fier - Tepelene Core Route 2c √ ALB Tepelene - Kakavia (GR border) Core Route 2c √ Morine Vermice (KOS border) - Lezhe Core Route 7 √ Qaf Thane (MKD border) - Elbasan Core Corridor VIII √ Elbasan - Tirane Core Corridor VIII √ Tirane - Durres Core Corridor VIII √ Durres - Vlore Core Corridor VIII √ Fier - Vlore Core Corridor VIII √ Neum west - Neum South Core Route 1 √ Gradiska (CRO border) - Banja Luka - Jajce Jug Core Route 2a √ Jajce Jug - Donji Vakuf Core Route 2a √ Donji Vakuf - Lasva Core Route 2a √ Sarajevo - Hum (MNE border) Comprehensive Route 2b √ Sarajevo - Lapisnica Comprehensive Route 3 √ Lapisnica - Ljubogosta Comprehensive Route 3 √

Ljubogosta - Podromanija Comprehensive Route 3 √ Podromanija - Rogatica Comprehensive Route 3 √ BIH Rogatica - Vardiste (SRB border) Comprehensive Route 3 √ Banja Luka - Doboj - Karakaj (SRB border) Comprehensive Route 9a √ Bosanski Samac (CRO border) - Matuzici Core Corridor Vc √ Matuzici - Ozimica Core Corridor Vc √ Ozimica - Topcic Polje Core Corridor Vc √ Topcic Polje - Sarajevo - Jablanica Core Corridor Vc √ Jablanica - Potoci Core Corridor Vc √ Potoci - Doljani (CRO border) Core Corridor Vc √ Djeneral Jankovic (KOS border) - Skopje Core Route 6a √ √ Podmolje - Bitola Comprehensive Route 8 √ √ Miladinovci - Stip - Novo Selo (BG border) Comprehensive Route 10 √ √ Kafasan (ALB border) - Skopje Core Corridor VIII √ √

Skopje - Stracin Core Corridor VIII √ √ Stracin - Kriva Palanka Core Corridor VIII √ √ MKD Kriva Palanka - Deve Bair (BG border) Core Corridor VIII √ √ Tabanovce (SRB border) - Skopje - Bogorodica (GR border) Core Corridor X √ √ Veles - Prilep Comprehensive Corridor Xd √ √ Prilep - Bitola Comprehensive Corridor Xd √ √ Bitola - Medzitlija (GR border) Comprehensive Corridor Xd √ √ Brnjak (SRB border) - Veternik Comprehensive Route 6a √ Veternik - Lipljan Core Route 6a √ Lipljan - Hani i Elezit ( MKD border) Core Route 6a √ Kuqishte (MNE border) - Kijeve/Kijevo Comprehensive Route 6b √

KOS* Kijeve/Kijevo - Gjurgjice/Djurdjice Comprehensive Route 6b √ Gjurgjice/Djurdjice - Fushe Kosove/Kosovo Polje Comprehensive Route 6b √ Vermice/Vrbnica (ALB border) - Merdare (SRB border) Core Route 7 √

Page 37

Past/ Connect SEETO Corridors / current a pilot Missing All sections Network Routes Past RSI iRAP RSI RSI Debeli Brijeg (CRO border) - Sukobin (ALB border) Core Route 1 √ Scepan Polje (BIH border) - Bozaj (ALB border) Comprehensive Route 2b √

Dobrakovo (SRB border) - Mioska Core Route 4 √ Mioska - Podgorica Core Route 4 √ MNE Podgorica - Bar Core Route 4 √ Ribarevine - Dracenovac (SRB border) Comprehensive Route 6a √ Kolasin - Kula (KOS border) Comprehensive Route 6b √ Uzice - Kotroman (BIH) Comprehensive Route 3 √ Vatin (RO border) - Belgrade - Orlovaca Core Route 4 √ Orlovaca-Stepojevac Core Route 4 √ Stepojevac-Celije Core Route 4 √ Celije - Knezevici Core Route 4 √ Knezevici - Kokin Brod Core Route 4 √ Kokin Brod - Gostun (MNE border) Core Route 4 √ Cacak/Preljina - Mrcajevci Comprehensive Route 5 √ Mrcajevci - Vrnjci Comprehensive Route 5 √ Vrnjci - Kamidzora Comprehensive Route 5 √ Kamidzora- Paracin - Vrska Cuka (BG border) Comprehensive Route 5 √ Spiljani (MNE border) - Brnjak (KOS border) Comprehensive Route 6a √ Nis - Merosina Core Route 7 √

Merosina - Merdare (KOS border) Core Route 7 √ Novi Sad/Petrovaradin - Sremska Kamenica Comprehensive Route 9a √ SRB Sremska Kamenica - Irig Comprehensive Route 9a √ Irig - Ruma Comprehensive Route 9a √ Ruma - Klenak Comprehensive Route 9a √ Klenak - Loznica - Mali Zvornik (BIH border) Comprehensive Route 9a √ Batrovci (CRO border) - Kuzmin Core Corridor X √ Kuzmin - Sremska Mitrovica Core Corridor X √ Sremska Mitrovica - Beograd/Dobanovci Core Corridor X √ Beograd/Dobanovci - Bubanj Potok Core Corridor X √ Bubanj Potok - Mali Pozarevac Core Corridor X √ Mali Pozarevac - Presevo (MKD border) Core Corridor X √ Horgos (HU border) - Feketic Core Corridor Xb √ Feketic-Sirig Core Corridor Xb √ Sirig - Beograd/Dobanovci Core Corridor Xb √ Nis - Gradina (BG border) Core Corridor Xc √

The previous table shows (last column) the road sections that have not been inspected during the last three years (either with ‘traditional’ RSI or with EuroRAP/iRAP methodology) and not included in the Connecta pilot RSI.

It is suggested that for the remaining road sections (road network that has not been inspected during the last three years and not included at Connecta pilot RSI – column ‘missing RSI’ of previous table) the following methodological approach be followed:

1. Prepare iRAP maps, according to the star rating methodology.

2. Perform detailed road safety inspections for the road sections that have the worst performance according to the iRAP star rating methodology as described in section 2.3.1.

It is estimated that if a Regional Participant performs RSI for the 20% of the Regional Participant’s core and comprehensive road network, within 5 years the whole network will have been inspected. If funds are short, the iRAP may be performed and the RSI will focus only on the less performed road sections.

Page 38

2.3.3.1 Cost estimations

In order to estimate the cost of the proposed assessments, the following assumptions have been taken into consideration.

For EuroRAP Road Risk Mapping implementation, an accredited consultant would require (if data is readily available) a team work of data and map processing.

Taking into account all cost items involved, it is estimated that the total cost per km of length included in the Risk Map is between 20€ and 30€. The total cost incorporates all associated required costs (i.e. QA, mapping according to standards, etc.).

For iRAP implementation, a licenced team leader along with two junior members are required. Furthermore, an accredited iRAP system needs to be involved. Lastly, according to the iRAP protocol a Quality Assurance procedure needs to be followed.

In terms of resources required, adapting the assumption that for iRAP implementation, overall and as an average, in terms of performance, the iRAP team may complete 30 km of road length within a working day, the average cost per km may be estimated. This assumption of performance, integrates the time and cost for survey planning, field tasks, elements coding, data uploading to ViDA, and reporting.

Taking into account all cost items involved, it is estimated that the total cost per km of length assessed is between 100€ and 120€. The total cost incorporates all associated required costs (i.e. iRAP system, labour, reporting, etc.).

For the ‘traditional’ RSI, three road safety experts (one team leader and two team members) need to be involved. Since considerable time is on field, travel and accommodation costs are also involved.

In terms of resources required, having in mind that approximately 10 km of road length may be inspected at a working day of the ‘traditional’ RSI team, as an average estimation and overall of a road network, the average required budget per road length in km may be estimated. This estimation includes the costs and time spent by the ‘traditional’ RSI team for meetings, planning, inspecting and reporting, as well as of travel and accommodation, etc.

Taking into account all cost items involved, it is estimated that the total cost per km of length inspected is at the range of 300€ to 350€. The total cost incorporates all associated necessary costs (i.e. equipment, labour, per diems, reporting, etc.).

The above indicative cost ranges are based on the hypothesis of inspections to be conducted by private contractor (as per ToR) and with a team of international (at least the leader) and local experts.

2.3.3.2 Albania

EuroRAP Road Risk Mapping for the whole core and comprehensive Albania network is anticipated to cost between 17,000€ and 26,000€.

The following table presents the road sections that either have not been inspected in the last 3 years or were not inspected as part of the Connecta pilot RSI.

Page 39

iRAP ‘traditional’ RSI Road sections SEETO Network Corridors/Routes min max min max Hani i Hotit (MNE Comprehensive Route 2b 32,100 37,450 border) - Fush Kruje Fier - Tepelene Core Route 2c 24,600 28,700 Durres - Vlore Core Corridor VIII 35,700 41,650

The total length of the above mentioned road sections is approximately 310 km. Since it is relatively small length, it is recommended to skip the initial phase of iRAP survey and rating, and proceed with ‘traditional’ road safety inspections.

Thus, it is proposed to implement RSI of approx. 100 km per year, for the next 3 years.

The anticipated cost is between 92,500€ and 108,000€.

2.3.3.3 Bosnia and Herzegovina

EuroRAP Road Risk Mapping for the whole core and comprehensive Bosnia and Herzegovina network is anticipated to cost between 21,000€ and 32,000€.

The following table presents the road sections that either have not been inspected in the last 3 years or were not inspected as part of the Connecta pilot RSI. iRAP ‘traditional’ RSI Road sections SEETO Network Corridors/Routes min max min max Neum west - Neum Core Route 1 500 600 1,500 1,750 South Gradiska (CRO border) - Banja Luka - Core Route 2a 12,700 15,240 38,100 44,450 Jajce Jug Donji Vakuf - Lasva Core Route 2a 6,800 8,160 20,400 23,800 Sarajevo - Hum (MNE Comprehensive Route 2b 9,700 11,640 29,100 33,950 border) Sarajevo - Lapisnica Comprehensive Route 3 250 300 750 875 Ljubogosta - Comprehensive Route 3 2,750 3,300 8,250 9,625 Podromanija Rogatica - Vardiste Comprehensive Route 3 6,560 7,872 19,680 22,960 (SRB border) Banja Luka - Doboj - Comprehensive Route 9a 19,800 23,760 59,400 69,300 Karakaj (SRB border) Bosanski Samac (CRO border) - Core Corridor Vc 8,000 9,600 24,000 28,000 Matuzici Topcic Polje - Core Corridor Vc 16,820 20,184 50,460 58,870 Sarajevo - Jablanica Potoci - Doljani (CRO Core Corridor Vc 6,200 7,440 18,600 21,700 border)

The total length of the above mentioned road sections is approximately 900 km.

According to the recommended approach, iRAP methodology is recommended for these road sections. This task could be fulfilled by the end of 2018.

Page 40

‘traditional’ RSI until April 2021. Finally, for the 4- and 5- star ranking road sections is recommended to implement ‘traditional’ RSI until end 2021. It is noted that, the exact percentage of 1-star (or 2-star and so on) ranking road length over the total network could be known before iRAP is implemented, thus assumptions are applied.

The total anticipated cost is 90,000€ - 108,000€ for iRAP and from 270,000€ to 315,000€ for ‘traditional’ RSI.

2.3.3.4 The former Yugoslav Republic of Macedonia

EuroRAP Road Risk Mapping for the whole core and comprehensive former Yugoslav Republic of Macedonia network is anticipated to cost between 16,500€ and 25,000€.

In the former Yugoslav Republic of Macedonia, a project that implements iRAP methodology for all Core and Comprehensive Road Network is ongoing. The results are expected soon. Therefore, there is no need for further iRAP implementation.

The following table presents the road sections that either have not been inspected in the last 3 years or were not inspected as part of the Connecta pilot RSI.

iRAP ‘traditional’ RSI Road sections SEETO Network Corridors/Routes min max min max Djeneral Jankovic (KOS Core Route 6a 1,600 1,920 4,800 5,600 border) - Skopje Podmolje - Bitola Comprehensive Route 8 7,800 9,360 23,400 27,300 Miladinovci - Stip - Comprehensive Route 10 14,200 17,040 42,600 49,700 Novo Selo (BG border) Kafasan (ALB border) - Core Corridor VIII 19,000 22,800 57,000 66,500 Skopje Skopje - Stracin Core Corridor VIII 6,200 7,440 18,600 21,700 Kriva Palanka - Deve Core Corridor VIII 1,600 1,920 4,800 5,600 Bair (BG border) Tabanovce (SRB border) - Skopje - Core Corridor X 18,100 21,720 54,300 63,350 Bogorodica (GR border) Veles - Prilep Comprehensive Corridor Xd 6,100 7,320 18,300 21,350 Bitola - Medzitlija (GR Comprehensive Corridor Xd 1,400 1,680 4,200 4,900 border)

The total length of the above mentioned road sections is approximately 760 km.

According to the iRAP output, the 1-star ranking road sections are recommended to be ‘traditionally’ road safety inspected by mid-2019. The 2- and 3- star ranking road sections should follow and implement ‘traditional’ RSI until end 2020. Finally, for the 4- and 5- star ranking road sections is recommended to implement ‘traditional’ RSI until end 2021. It is noted that, the exact portion of 1-star (or 2-star and so on) ranking road length over the total network is not known yet, therefore assumptions are applied.

The anticipated total cost is between 228,000€ and 266,000€.

Page 41

2.3.3.5 Kosovo

EuroRAP Road Risk Mapping for the whole core and comprehensive Kosovo network is anticipated to cost between 7,500€ and 11,000€.

The following table presents the road sections that either have not been inspected in the last 3 years or were not inspected as part of the Connecta pilot RSI. iRAP ‘traditional’ RSI Road sections SEETO Network Corridors/Routes min max min max Brnjak (SRB border) - Comprehensive Route 6a 21,600 25,200 Veternik Lipljan - Hani i Elezit ( MKD Core Route 6a 16,800 19,600 border) Kuqishte (MNE border) - Comprehensive Route 6b 20,100 23,450 Kijeve/Kijevo Vermice/Vrbnica (ALB border) - Merdare (SRB Core Route 7 37,500 43,750 border)

The total length of the above mentioned road sections is approximately 320 km. Since it is relatively small length, it is recommended to skip the initial phase of iRAP survey and rating, and proceed with ‘traditional’ road safety inspections.

Thus, it is proposed to implement ‘traditional’ RSI of around 110 km per year, for the next 3 years. The anticipated total cost is estimated from 96,000€ up to 112,000€.

2.3.3.6 Montenegro

EuroRAP Road Risk Mapping for the whole core and comprehensive Montenegro network is anticipated to cost between 13,000€ and 19,000€.

The following table presents the road sections that either have not been inspected in the last 3 years or were not inspected as part of the Connecta pilot RSI. iRAP ‘traditional’ RSI Road sections SEETO Network Corridors/Routes min max min max Debeli Brijeg (CRO border) Core Route 1 12,200 14,640 36,600 42,700 - Sukobin (ALB border) Scepan Polje (BIH border) Comprehensive Route 2b 15,400 18,480 46,200 53,900 - Bozaj (ALB border) Dobrakovo (SRB border) - Core Route 4 8,100 9,720 24,300 28,350 Mioska Podgorica - Bar Core Route 4 5,000 6,000 15,000 17,500 Ribarevine - Dracenovac Comprehensive Route 6a 7,940 9,528 23,820 27,790 (SRB border) Kolasin - Kula (KOS Comprehensive Route 6b 9,900 11,880 29,700 34,650 border)

The total length of the above mentioned road sections is approximately 590 km.

According to the recommended approach, iRAP methodology is recommended for these road sections. This task could be fulfilled by the end of 2018.

Page 42

According to the iRAP output, the 1-star ranking road sections are recommended to be ‘traditionally’ road safety inspected within 2019. The 2- and 3- star ranking road sections should follow and implement ‘traditional’ RSI until April 2021. Finally, for the 4- and 5- star ranking road sections is recommended to implement ‘traditional’ RSI until end 2021. For the yearly allocation of resources, assumptions are utilised, since it’s not known what the road length per star category would be.

The total estimated cost is approx. 58,500€ - 70,000€ for iRAP and 175,500€ - 205,000€ for ‘traditional’ RSI.

2.3.3.7 Serbia

EuroRAP Road Risk Mapping for the whole Serbia core and comprehensive road network is anticipated to cost between 34,500€ and 52,000€.

The following table presents the road sections that either have not been inspected in the last 3 years or were not inspected as part of the Connecta pilot RSI.

iRAP ‘traditional’ RSI Road sections SEETO Network Corridors/Routes min max min max Uzice - Kotroman (BIH) Comprehensive Route 3 5,400 6,480 16,200 18,900 Vatin (RO border) - Core Route 4 11,700 14,040 35,100 40,950 Belgrade - Orlovaca Celije - Knezevici Core Route 4 14,600 17,520 43,800 51,100 Kokin Brod - Gostun (MNE Core Route 4 7,000 8,400 21,000 24,500 border) Mrcajevci - Vrnjci Comprehensive Route 5 4,400 5,280 13,200 15,400 Kamidzora- Paracin - Comprehensive Route 5 13,600 16,320 40,800 47,600 Vrska Cuka (BG border) Spiljani (MNE border) - Comprehensive Route 6a 2,800 3,360 8,400 9,800 Brnjak (KOS border) Merosina - Merdare (KOS Core Route 7 7,100 8,520 21,300 24,850 border) Irig - Ruma Comprehensive Route 9a 900 1,080 2,700 3,150 Klenak - Loznica - Mali Comprehensive Route 9a 8,000 9,600 24,000 28,000 Zvornik (BIH border) Batrovci (CRO border) - Core Corridor X 3,410 4,092 10,230 11,935 Kuzmin Sremska Mitrovica - Core Corridor X 4,760 5,712 14,280 16,660 Beograd/Dobanovci Mali Pozarevac - Presevo Core Corridor X 35,000 42,000 105,000 122,500 (MKD border) Horgos (HU border) - Core Corridor Xb 7,200 8,640 21,600 25,200 Feketic Sirig - Beograd/Dobanovci Core Corridor Xb 9,090 10,908 27,270 31,815 Nis - Gradina (BG border) Core Corridor Xc 11,000 13,200 33,000 38,500

The total length of the above mentioned road sections is approximately 1,450 km.

According to the recommended approach, iRAP methodology is recommended for these road sections. This task could be fulfilled by the end of 2018.

According to the iRAP output, the 1-star ranking road sections are recommended to be ‘traditionally’ road safety inspected within 2019. The 2- and 3- star ranking road sections should follow and implement ‘traditional’ RSI until April 2021. Finally, for the 4- and 5- star ranking road sections is recommended to implement ‘traditional’ RSI until end 2021. The yearly allocation of ‘traditional’ RSI may only be estimated at best possible assumption.

Page 43

The total anticipated cost is approximately 146,000€ - 175,000€ for iRAP and 438,000€ - 511,000€ for ‘traditional’ RSI.

2.3.3.8 Yearly allocation of resources

Based on the previous sections, the following table summarises the required financial resources on a yearly basis, expressed in Euro currency for complementing inspections in SEETO core/comprehensive road network. The necessary budget is indicative according to average cost values per road length and per methodology followed.

Year 1 Year 2 Year 3 Year 4 Total EuroRAP 21,500 21,500 ALB iRAP 0 ‘traditional' RSI 33,000 33,000 34,000 100,000 EuroRAP 26,500 26,500 BiH iRAP 99,000 99,000 ‘traditional' RSI 135,000 100,000 58,000 293,000 EuroRAP 9,000 9,000 MKD iRAP 0 ‘traditional' RSI 34,000 35,000 35,000 104,000 EuroRAP 21,000 21,000 KOS1 iRAP 0 ‘traditional' RSI 110,000 85,000 52,000 247,000 EuroRAP 16,000 16,000 MNE iRAP 64,500 64,500 ‘traditional' RSI 95,000 65,000 30,000 190,000 EuroRAP 43,500 43,500 SRB iRAP 160,500 160,500

‘traditional' RSI 210 ,000 165 ,000 99 ,000 474,000 EuroRAP 137,500 0 0 0 137,500 Total iRAP 324,000 0 0 0 324,000 ‘traditional' RSI 177,000 593,000 451,000 187,000 1,408,000

This is in fact an up to 3.5 year plan (in contrast to ToR for a 3 year plan), taking into account the date of this report and the relevant sub-project completion date, as well as the time required for iRAP implementation.

Hence, it was necessary to extend it till end of 2021, for some Regional Participants with extensive network and inspection needs.

2.3.3.9 Allocated costs per route/corridor and section in the three-year Plan (2018-2020)

The following Table presents the allocated costs per route/corridor and section in the three-year Plan (2018-2020) for road safety inspection.

Page 44

EuroRAP iRAP traditional Corridors/ SEETO Regional EuroRAP traditional All sections iRAP Risk Map Total RSI Routes Network Participant Risk Map RSI Total (€) (€) Total (€) Bosanski Samac (CRO Core BIH 2,000 8,800 26,000 border) - Matuzici Core Matuzici - Ozimica BIH 738 - - Core Ozimica - Topcic Polje BIH 600 - - Corridor Topcic Polje - Sarajevo - 10,000 34,122 100,815 Vc Core BIH 4,205 18,502 54,665 Jablanica Core Jablanica - Potoci BIH 908 - - Potoci - Doljani (CRO Core BIH 1,550 6,820 20,150 border) Qaf Thane (MKD border) - Core ALB 2,025 - - Elbasan Core Elbasan - Tirane ALB 625 - - Core Tirane - Durres ALB 750 - - Core Durres - Vlore ALB 2,975 - 38,675 Corridor Core Fier - Vlore ALB 840 - - 14,590 - 125,775 VIII Kafasan (ALB border) - Core MKD 4,750 - 61,750 Skopje Core Skopje - Stracin MKD 1,550 - 20,150 Core Stracin - Kriva Palanka MKD 675 - - Kriva Palanka - Deve Bair Core MKD 400 - 5,200 (BG border) Tabanovce (SRB border) - Core Skopje - Bogorodica (GR MKD 4,525 - 58,825 border) Batrovci (CRO border) - Core SRB 853 3,751 11,083 Kuzmin Kuzmin - Sremska Core SRB 533 Mitrovica Corridor X Sremska Mitrovica - 17,103 47,487 199,128 Core SRB 1,190 5,236 15,470 Beograd/Dobanovci Beograd/Dobanovci - Core SRB 738 - - Bubanj Potok Bubanj Potok - Mali Core SRB 515 - - Pozarevac Mali Pozarevac - Presevo Core SRB 8,750 38,500 113,750 (MKD border) Horgos (HU border) - Core SRB 1,800 7,920 23,400 Corridor Feketic 4,625 17,919 52,943 Xb Core Feketic-Sirig SRB 553 - - Core Sirig - Beograd/Dobanovci SRB 2,273 9,999 29,543 Corridor Core Nis - Gradina (BG border) SRB 2,750 12,100 35,750 2,750 12,100 35,750 Xc Comprehensive Veles - Prilep MKD 1,525 - 19,825 Corridor Comprehensive Prilep - Bitola MKD 1,050 - - 2,925 - 24,375 Xd Bitola - Medzitlija (GR Comprehensive MKD 350 - 4,550 border) Muriqan (MNE border) - Core ALB 800 - - Koplik Core Koplik - Skhoder ALB 325 - - Core Skhoder - F. Kruje ALB 1,793 - - Route 1 Core F. Kruje - Lezhe ALB 898 - - Core Neum west - Neum South BIH 125 550 1,625 12,690 35,420 104,650 Debeli Brijeg (CRO border) Core MNE 3,050 13,420 39,650 - Sukobin (ALB border) Gradiska (CRO border) - Core BIH 3,175 13,970 41,275 Banja Luka - Jajce Jug Route 2a Core Jajce Jug - Donji Vakuf BIH 825 - - Core Donji Vakuf - Lasva BIH 1,700 7,480 22,100 Hani i Hotit (MNE border) - Comprehensive ALB 2,675 - 34,775 Fush Kruje Comprehensive Fush Kruje - Vore ALB 350 - - Route 2b Sarajevo - Hum (MNE 9,300 27,610 116,350 Comprehensive BIH 2,425 10,670 31,525 border) Scepan Polje (BIH border) - Comprehensive MNE 3,850 16,940 50,050 Bozaj (ALB border) Core Rrogozhine - Fier ALB 1,125 - - Core Fier - Tepelene ALB 2,050 - 26,650 Route 2c 4,625 0 26,650 Tepelene - Kakavia (GR Core ALB 1,450 - - border)

Page 45

EuroRAP iRAP traditional Corridors/ SEETO Regional EuroRAP traditional All sections iRAP Risk Map Total RSI Routes Network Participant Risk Map RSI Total (€) (€) Total (€) Comprehensive Sarajevo - Lapisnica BIH 63 275 813 Comprehensive Lapisnica - Ljubogosta BIH 170 - - Comprehensive Ljubogosta - Podromanija BIH 688 3,025 8,938 Route 3 Comprehensive Podromanija - Rogatica BIH 715 - - 4,625 16,456 48,620 Rogatica - Vardiste (SRB Comprehensive BIH 1,640 7,216 21,320 border) Comprehensive Uzice - Kotroman (BIH) SRB 1,350 5,940 17,550 Dobrakovo (SRB border) - Core MNE 2,025 8,910 26,325 Mioska Core Mioska - Podgorica MNE 1,350 - - Core Podgorica - Bar MNE 1,250 5,500 16,250 Vatin (RO border) - Core SRB 2,925 12,870 38,025 Belgrade - Orlovaca Route 4 15,030 51,040 150,800 Core Orlovaca-Stepojevac SRB 553 - - Core Stepojevac-Celije SRB 553 - - Core Celije - Knezevici SRB 3,650 16,060 47,450 Core Knezevici - Kokin Brod SRB 975 - - Kokin Brod - Gostun (MNE Core SRB 1,750 7,700 22,750 border) Comprehensive Cacak/Preljina - Mrcajevci SRB 308 - - Comprehensive Mrcajevci - Vrnjci SRB 1,100 4,840 14,300 Route 5 Comprehensive Vrnjci - Kamidzora SRB 520 - - 5,328 19,800 58,500 Kamidzora- Paracin - Vrska Comprehensive SRB 3,400 14,960 44,200 Cuka (BG border) Djeneral Jankovic (KOS Core MKD 400 - 5,200 border) - Skopje Brnjak (SRB border) - Comprehensive KOS 1,800 - 23,400 Veternik Core Veternik - Lipljan KOS 190 - - Route 6a Lipljan - Hani i Elezit ( MKD 6,475 11,814 81,705 Core KOS 1,400 - 18,200 border) Ribarevine - Dracenovac Comprehensive MNE 1,985 8,734 25,805 (SRB border) Spiljani (MNE border) - Comprehensive SRB 700 3,080 9,100 Brnjak (KOS border) Kuqishte (MNE border) - Comprehensive KOS 1,675 - 21,775 Kijeve/Kijevo Kijeve/Kijevo - Comprehensive KOS 275 - - Route 6b Gjurgjice/Djurdjice 5,125 10,890 53,950 Gjurgjice/Djurdjice - Fushe Comprehensive KOS 700 - - Kosove/Kosovo Polje Comprehensive Kolasin - Kula (KOS border) MNE 2,475 10,890 32,175 Morine Vermice (KOS Core ALB 2,775 - - border) - Lezhe Vermice/Vrbnica (ALB Core border) - Merdare (SRB KOS 3,125 - 40,625 Route 7 7,850 7,810 63,700 border) Core Nis - Merosina SRB 175 - - Merosina - Merdare (KOS Core SRB 1,775 7,810 23,075 border) Route 8 Comprehensive Podmolje - Bitola MKD 1,950 25,350 1,950 0 25,350 Banja Luka - Doboj - Comprehensive BIH 4,950 21,780 64,350 Karakaj (SRB border) Novi Sad/Petrovaradin - Comprehensive SRB 195 - - Sremska Kamenica Route 9a Comprehensive Sremska Kamenica - Irig SRB 353 - - 8,598 31,570 93,275 Comprehensive Irig - Ruma SRB 225 990 2,925 Comprehensive Ruma - Klenak SRB 875 - - Klenak - Loznica - Mali Comprehensive SRB 2,000 8,800 26,000 Zvornik (BIH border) Miladinovci - Stip - Novo Route 10 Comprehensive MKD 3,550 - 46,150 3,550 0 46,150 Selo (BG border) Total: 137,138 324,038 1,408,485 137,138 324,038 1,408,485

Page 46

2.4 Pilot Road Safety Inspections

2.4.1 The selection – planning process

Pilot road safety inspections have been carried out by the team (Activity 4, according to ToR) using SEETO’s road safety inspection guidelines on 10% (about 550 km, as per ToR, but finally actually 580 km) of the core and comprehensive road network that is considered highest risk portion of the network based on fatal crash data. SEETO members prepared a list of their high risk sections and the team compiled a list of about 550 km of these high risk roads by maintaining a reasonable distribution among all SEETO’s members.

Initially the High Risk Road Sections from all WB6 Regional Participants were assessed. The information received was not comparable and should not be, because all Regional Participants have different characteristics and different problems. The crashes occurring on these roads are only an indication. Normally, it is not possible to compare between different roads from various Regional Participants, if their traffic characteristics, volumes, design, etc. are not known. Furthermore, the total length either of the TEN-T road network or of the high risk road sections are not the same in all Regional Participants.

As some Regional Participants were not be able to deliver specific crash information for these road sections, and more importantly the exchange of information (either with meetings or information exchange) at the initial stage would create more problems (time delays and having a second round of comparing information between the WB6 Regional Participants), a proposal was developed by the team for SEETO and the Road Safety WG to consider.

The area coverage, the population, the total number of road crashes, the total number of road fatalities, the total number of road injuries, the relevant figures from WHO, the under-reporting of road safety statistics, etc. of each Regional Participant, were taken into consideration along with the team’s engineering judgement, and initially the total road section length of each Regional Participant to be inspected as part of Connecta project was formed. Apart from statistics, it was aimed to include a representative sample of each Regional Participant's road network. In order to introduce RSI capacity building within each Regional Participant and have sound information for the plan to be developed later, it was proposed to have more than one road section within each Regional Participant and not less than a total of 40 km for each SEETO Regional Participant. The only exception to this rationale was Montenegro, where it was proposed to inspect only one road ("Podgorica- Mioska") instead of two road sections (namely "Lipci - ljuta" and "Ribarevine - Berane"), because it was considered a much more important road. Furthermore, it was intended not to underestimate the smaller Regional Participants.

Furthermore, since it seems that according to statistics, Albania and Bosnia and Herzegovina face proportionally bigger road safety problems, and Serbia is the best performing Regional Participant (according to WHO) in terms of fatality rate per 100,000 population, we decreased the share of Serbia and increased a bit the road length inspection in Albania and Bosnia and Herzegovina. Since there was a need for a fair sample from each Regional Participant, the final proposed RSI road length per Regional Participant was the following:  Albania: 110 km  Bosnia and Herzegovina: 125 km  The former Yugoslav Republic of Macedonia: 60 km  Kosovo: 50 km  Montenegro: 50 km  Serbia: 155 km

The following Table presents the rationale for the determination of the length per Regional Participant.

Page 47

WHO WHO

)

estimated estimated total % ies

wiki)

idents (2013) idents

WHO

Country Area (2017 Population est. Pop acc. Length Acc (2014) Accidents (2013) Fatalities (2014) Fatalities (2014) Injured total % Injured (2014) Fatalities (2013) estimated Fatalities % (2013) estimated total fatalities Length WHO % total (2013) total % Accidents (2014) Fatalit (2014) total % Area Area acc. Length %total Pop (km) Average acc. Length Fatalities Injured acc. Length (km) Average per rate Fatality pop. 100.000 ( length Proposed (km) approx. ALB 28,750 2,877,000 88 2,075 1,914 295 264 2,353 5% 478 22.1% 111 2% 19% 13% 72 16% 81 104 27 66 15.1 110 BiH 51,197 3,531,000 107 37,725 36,225 334 297 10,067 21% 676 31,3% 156 37% 21% 23% 129 20% 130 117 117 117 17.7 125 MKD 25,710 2,069,000 63 11,000 3,853 198 130 6,056 13% 198 9,2% 46 4% 9% 12% 65 11% 56 51 70 61 9.4 60 KOS* 10,908 1,859,000 57 19,929 16,301 118 111 9,713 20% 16% 8% 5% 27 10% 78 44 113 78 50 MNE 13,810 679,000 21 5,264 5,531 74 56 1,278 3% 74 3,4% 17 6% 4% 6% 35 4% 18 22 15 18 11.9 50 SRB 88,360 7,058,000 215 37,140 35,152 650 536 17,953 38% 735 34,0% 170 36% 38% 40% 222 39% 197 211 208 210 7.7 155

Page 48

Additionally, all road sections delivered by WB6 were assessed. The selection included the most important road sections, as well as some less busy roads. An attempt was also made to be close to the proper road length per Regional Participant.

In the following Table, the proposed Road Sections to be included for pilot RSI of Connecta, are highlighted in green. Lastly, it should be mentioned that in order to be in line with the incoming information of High Risk roads, only full road length was proposed and not sections out of the information received (since it would not be possible to justify which segment to select).

Page 49

Regional Corridor/ Name of the Project SEETO Network/ Section RSI Proposed Section Start Node Section End Node Type of road Participant Route/Node (sections) Core/Comprehensive length (km) Length Secondary Intercity (Kat C, ALB E762 dhe SH1 Shkoder - Koplik Road Start Tuzit Start of By Pass Koplik Comprehensive 13 13,0 according the Albanian Code) Road Intersection Intersection of Nation Secondary Intercity (Kat C, ALB E 851 Milot - Rreshen Laç -Lezhe to Nation Road to road Burrel - Core 10,6 according the Albanian Code) Road Peshkopi Main Intercity (partially for Overpassing F. Roundabout in Lezha Thumane - Milot) and ALB E 762 F. Kruje - Lezhe Core 35,9 35,9 Kruje exit secondary (the remaining part) Interchange of By Pass ALB SH 2 Tirane - Durres Overpassing Kamez Main intercity Core 30 30,0 Shkozet Road Intersection to ALB SH 3 dhe E 86 Pogradec - Bilisht Entry in Bilisht city Secondary intercity Comprehensive 65 Reshit Çollaku Road Main intercity (Levan - Vlore) Intersection to Intersection to road ALB E 853 Fier - Vlore and secondary for ( Fier- Comprehensive 33,6 33,6 Aulona Road Sinan Ferhati Levan) ALB: Sub-total: 112,5 Е-661 (Gradiška - BiH Route 2a 0+000 32+000 Motorway Core 32 Banja Luka) M-18 (Dobro polje- BiH Route 2b 0+000 18+809 The Main Road Comprehensive 18,809 18,8 Miljevina) M-18 (Brod na Drini BiH Route 2b 3-granica RS 0+000 20+715 The Main Road Comprehensive 20,715 20,7 (Šćepan polje)) M-19 (Podromanija- BiH Route 3 0+000 21+559 The Main Road Comprehensive 21,559 21,6 Sumbulovac) М-4 (Donje BiH Route 9а 0+000 15+350 The Main Road Comprehensive 15,35 Caparde-Karakaj 1) Е-661 (M 5 : Jajce BiH Route 2a 0+000 33+000 The Main Road Core 33 Jug - Donji Vakuf) Е-73 (M 17: Karuše BiH Corridor Vc 0+000 29+485 The Main Road Core 29,5 - Ozimica) Е-73 (M 17: Ozimica BiH Corridor Vc 0+000 23+993 The Main Road Core 24 24,0 - Topčić Polje) Е-73 (M 17: BiH Corridor Vc 0+000 36+312 The Main Road Core 36,3 36,3 Jablanica - Potoci) BiH: Sub-total: 121,4 MKD Gostivar Kicevo 48 MKD Bitola Prilep 42 42,0 MKD Gradsko Prilep 53,5

Page 50

Regional Corridor/ Name of the Project SEETO Network/ Section RSI Proposed Section Start Node Section End Node Type of road Participant Route/Node (sections) Core/Comprehensive length (km) Length MKD Bitola Resen 32 MKD Stracin Kriva Palanka 27 27,0 MKD Strumica Border with Bulgaria Not in CCRN 21,5 MKD: Sub-total: 69,0 Prishtine/N-2 and N- National/ KOS N9 0 Fushe Kosove National road 6 9 Interchange the main road Gjurgjice/ R7-R6b National/ KOS R6b 0 Fushe Kosove Core 28 28,0 Interchange the main road Gjurgjice/ R7-R6b National/ KOS R6b 0 Kijeve Comprehensive 11 11,0 Interchange the main road National/ KOS R6a 0 Prishtine Mitrovice Comprehensive 35 the main road Veternik/ N‐2 and N- Lipjan/N-2 and N-25 National/ KOS R6a 0 25.2 Core 7,6 7,6 roundabout the main road interchange KOS: Sub-total: 46,6 Belgrade 8 SRB)- MNE Route 4 Podgorica- Mioska PG Mioska National 54 54,0 Podgorica(MNE) Neum ( Cro) - Debeli MNE Route1 Lipci - ljuta Lipci Ljuta National 21,3 Brijeg- Bar Sarajevo ( Bih) - MNE Route 2b Nikšić - Danilovgrad Nikšić Danilovgrad National 37,5 Podgorica ( MNE) Peje ( Kos) - Kolasin MNE Route 6b Ribarevine - Berane Ribarevine Berane National 28,7 ( MNE) MNE: Sub-total: 54,0 SRB Route 4 IB22 9 61 2way Comprehensive 22,1 22,1 SRB Route 9 IB21 97 98 motoput Comprehensive 7,8 7,8 SRB Corridor X A1 11 12 Autoput Core 20,6 20,6 SRB Corridor X A1 9 11 Autoput Core 29,5 29,5 SRB Route 4 IB22 61 62 2way Comprehensive 22,1 22,1 SRB Corridor Xb A1 37 38 Autoput Core 22,1 22,1 SRB Route 5 IB23 77 78 2way Comprehensive 20,8 20,8 SRB Route 5 IB22 75 76 2way Comprehensive 12,3 12,3 SRB: Sub-total: 157,1 TOTAL: 560,6

Page 51

The proposed road sections were a reasonable distribution among SEETO’s members, taking into consideration the individual characteristics, needs, statistics and road safety situation of the 6 Regional Participants, and were agreed by SEETO and RSWG members.

After SEETO approval, the RSI team drafted a mission plan and started to contact the RSWG members from each relevant Regional Participant and set meetings.

During the meetings, apart from the data / information request local experts were invited to participate in the field tasks. The latter (although not required by ToR) was part of the Connecta input to the Regional Participant’s capacity building of RSI.

The RSI field tasks were performed according to the following time schedule:  3-7 September, 8-9 November, 4-8 December Serbia  10-19 September, 11-14 December Bosnia and Herzegovina  25-29 September Albania  2-7 October Montenegro  5-8 November Kosovo  19-23 November The former Yugoslav Republic of Macedonia.

The final selected road sections as well as the initial time plan had to be modified.

This was due to issues regarding inspections to be carried out by unlicensed Inspectors raised by the Republic of Srpska of the Bosnia and Herzegovina, during the September 19th meeting at Banja Luka. These issues caused the planned RSI at Republic of Srpska to be rejected and the unperformed RSI length to be re-allocated to the Federation of Bosnia and Herzegovina (following SEETO SC’s decision and no objection by DG NEAR). Therefore, the following 2 road sections substituted the originally selected RSI road sections that were located at the Republic of Srpska.

RSI Corridor/ Name of the Project Section_ Section_ Type of SEETO Core/ Section Route (sections) Start Node End Node road Comprehensive length (km) Е-661 (M 5 : Jajce The Main Route 2a 0+000 33+000 Core 33 Jug - Donji Vakuf) Road Corridor Е-73 (M 17: Karuše The Main 0+000 29+485 Core 29,5 Vc - Ozimica) Road

Furthermore, some additional road sections (precisely two from the former Yugoslav Republic of Macedonia around 18 km long) were requested to be included in the pilot RSI. Connecta did this, despite this being additional to what was initially decided.

These two sections are presented at the next table. RSI Corridor/ Name of the Project Section_ Section_ Type of SEETO Core/ Section Route (sections) Start Node End Node road Comprehensive length (km) R-106 Prilep - Drenovo Prilep Drenovo 8,5 Drenovo - Rosoman - R-106 Drenovo Rosoman 9,9 connection A1

Page 52

After implementing the field tasks, the road length per section was precisely defined. So, small dissimilarities were identified between the declared lengths and the real ones.

Thus, the road sections that have been inspected in the field have a total length of 580 km (30 km more than the indicated in ToR).

2.4.2 General findings

Pilot road safety inspections have been carried out by the team using SEETO’s road safety inspection manual on approximately 580 km of the core and comprehensive road network that is considered highest risk portion of the network based on fatal crash data (based on information provided by each Regional Participant).

Below are presented, in brief, the most important or/and frequent RSI findings along the road sections inspected in WB6. Further details on findings and proposed remedial measures are included in the specific pilot RSI reports (separately for each road section).

2.4.2.1 Assessment of the deficits

The most critical deficits in respect of road safety are (of course, depending on each separate road section):  Co-existence of long distance travel and local trips - increased speed conflicts and demand for overtaking.  Many accesses (houses and commercial businesses next to the road). In particular, accesses where left turns are allowed (especially from the main carriageway) contribute to many conflicts and increased risk of crashes.  Conflicts between vehicles and pedestrians. Activities around the road generate demand for pedestrian movement by the road and crossing the road at places with a speed limit and operation speed of more than 50 kph.  Unpaved areas near the road which have unregulated / unchanneled exits and entries to the main carriageway. Vehicles need to decelerate before exiting or accelerate after entering the traffic on the road. Manoeuvres interrupt other drivers, create traffic conflicts and could result in road crashes.  The absence of adequate pedestrian facilities at urban segments, near bus stops, etc. could force pedestrians to use the carriageway for walking along the road. It could increase risk and cause crashes involving pedestrians.  Parked vehicles (legal or illegal) in the settlements reduce sight distance, reduce space for pedestrians, take driver’s attention away and could make confused traffic situations and cause road crashes.  Legal and illegal advertising signs (billboards), placed in the safety zone of the road, taking driver’s attention away. Some advertisment billboards reduce sight distance, especially near the intersections. This could increase risk of road crashes near advertisment billboards.  Inadequate pedestrian facilities on urban subsections, near bus stops, near houses and commercial plots, could force pedestrians to use carriageway for walking along the road. It could increase risk and cause road crashes with pedestrians.

Page 53

 Insufficient space and inappropriate design of bus stops, with missing information signs in advance, could create unexpected situations in the traffic and cause crashes. It could make bus deceleration and acceleration difficult and unsafe.  Damaged, not maintained or not safe guardrails with unsafe ends, gaps and unsafe connections.  Road safety barriers are missing in some places where guardrails are needed to protect vehicles from hitting hard objects or from running off the road at dangerous places (i.e. high slopes).  Guardrails installed to protect culverts or hard objects near the road with inadequate length. These short guardrails cannot protect hard objects and would not stop vehicles which run off the road.  Unsafe barrier ends (terminals).  Sharp curves without chevron signs to inform and advice drivers  Street lighting not adequate in some villages, at schools and bus stops.

2.4.2.2 Proposals and Recommendations

Among the deficiencies, there are some that could greatly be improved with some low cost implementation measures. Some general, initial remedies of the RSI team are described below.

2.4.2.2.1 Improve the guardrail system

Add guardrails where needed, repair guardrails, improve connection, extend guardrails or change the end of the guardrails.

The guardrail system should have proper length in order to be an appropriate restraint system for users. Guardrails will not work in the system if the elements do not have the required length. It could also create too many guardrail beginnings that may be problematic with regard to road safety.

Recommendation is based on the EN 1317-4 (or EN 1317-7 new end terminal treatment). In order to protect errant vehicles from exiting the road and to secure safe stops, end terminals should be constructed based on the mentioned EN standards.

The required length of the guardrail, relevant to the object to be protected

Page 54

Example of transition between concrete barrier and guardrail

2.4.2.2.2 Unsafe guardrail ends

Recommendation is based on the EN 1317-7 new end terminal treatment. In order to help errant vehicles to stop safely, end terminals should be constructed based on the mentioned EN standards.

The ends of the guardrails have to be rounded

Tangent energy-absorbing terminals

Page 55

Example of a proper energy absorbing terminals

Example of barrier ends at intersection/access

Page 56

2.4.2.2.3 Tunnel entrances

A recommended solution which could efficiently solve the problem with tunnel portals is installation of tunnel portal safety barriers or crash cushions. The barriers cover the beginning and end of the head front of tunnels portals. For bidirectional tunnels and tunnels regularly operated for bidirectional flows, transitions should be designed for safety in both traffic directions. Guardrails can finish on the portal so vehicles are protected from the head front of the portal. An example of possible implementation tunnel portal safety barriers could be like in this case (picture below).

Example of the placement of safety barriers before parallel tunnel portals

2.4.2.2.4 Pedestrian crossing with proper lighting

General recommendation for pedestrian crossing lighting should be enhanced with additional lighting poles at the intersection zones and in the areas where pedestrian movement is expected.

Proposed measures, in accordance with the standards, include placement of adequate lighting that allows the lighting both of the crossing itself and the accesses, to ensure that drivers on the main road notice pedestrians in time.

Good example of lighting of pedestrian crossing which makes positive silhouette of a pedestrian

Page 57

2.4.2.2.5 Agriculture access roads

In order to protect users on the main road, pavement on the access roads from agriculture fields should be made of asphalt, concrete or gravel. Satisfactory length for the pavement on the access road is min 15-20 m. Access from the agricultural roads should be gravel, because this material helps previously accumulated dirt from wheels of the agricultural vehicle to be stripped before they access the state road.

This measure reduces bringing mud to the road and therefore the risk of vehicle encountering mud and, consequently skidding and loss of control of the vehicle.

Agricultural access road from ribbed concrete (expensive but most efficient solution, example from Serbia)

2.4.2.2.6 Additional signage for curves

Adding chevron signs before and along the sharp curves is recommended. The number of chevron signs should be calculated based on the length and radii of the curve as well as a principle that at each point in time the driver should see at least three chevron signs. It is recommended that, depending on the radii of the curve, distances between two chevron signs should be 8-15 m for curve radii of 60 m, 15-25 m for curve radii of 150 m and 25 m for curves with larger radii.

Proper installation of chevrons in curves

It is also recommended to install chevron signs with fluorescent base.

Page 58

2.4.2.2.7 Built-up Area Gates and Speed Management

Considering that the speed limit outside of settlements is more than 50 kph (i.e. 80 kph) and in settlements 50 km/h, it is necessary for the infrastructure to be adapted so that the drivers are forced to reduce speed when entering in the settlement. The following figure shows one of the possible solutions to the problem.

Adapted infrastructure (gates) for entry into the settlement

The need to change driving behaviour when entering built-up areas should be emphasized by some infrastructure elements in order to indicate very clearly that driving conditions are changed. The typical signs, according to the current traffic code, are not enough according to the RSI Team. The recommendation of RSI Team is to have median islands (as gates) at all built-up areas’ limits.

At the entrance of built-up areas, an island accompanied with the relevant signage creates a ‘threshold’ effect to show drivers that they are ‘entering’ an urban area and that there is a change in road type and its function, which requires lower speeds, attention to vulnerable road users (pedestrians, etc.) and unanticipated movements.

Page 59

Example - entering/exiting island to/from built-up areas

Good example of “gate” in Hungary

Good example with additional signage for “gate” in Austria

The gates should where relevant and at least at schools and bus stops be supplemented with speed management schemes.

Page 60

2.4.2.2.8 Access control

One of the big problems seen on most of the inspected roads are many and often unregulated accesses. It is recommended to regulate accesses e.g. by:  Close illegal accesses if it is possible,  Connect more neighbouring accesses with a service road, if it is possible. The service road should be directed to the main road with deceleration and acceleration lanes, or  Do designs and improvements of regular individual accesses, in case it is possible.

A high number of accesses gives higher risk of crashes

As a recommendation for solving the problem of access control at the inspected section, the following solutions are recommended:  Solution 1. Reconsider closing accesses on the road and redirecting them to an existing service (collecting) road and connect them to the main road (or to secondary road first and then to the main road) with a road junction. This solution is ideal if feasible. Thus, a feasibility study is primarily required.

Solution 1- Close accesses and redirect vehicles to regular intersection

Page 61

Image 1a. Visualisation example of solution 1 (It is a planning problem and may be solved in close cooperation with local communities, planners and designers )

 Solution 2. Rearrange and improve regulation of the accesses, constructing acceleration and deceleration lanes in order to ensure safe entering and exiting to the road. This is feasible only when the relevant property has the available ’face’ on the main road.

Solution 2- Rearrange the accesses

Image 2a. Visualisation example of solution 2

Page 62

 Solution 3. Merge and/or construct a traffic island to allow entrance and exit to the road at controlled places. Available width from carriageway to the nearest property should be at least 4 m. This is the best solution when the Road Authority has the right-of-way for the available roadside area.

Solution 3- Merge and/or construction of a traffic island

Image 3a. Visualisation example of solution 3

2.4.2.3 Bus stops

It is recommended that bus stops (especially outside settlements) should incorporate an island separating (forcing) the stopped bus from through traffic, bus stop sign at the bus stop and at some meters ahead (according to speed limit) for informing drivers following a bus to expect its speed decrease, lighting, etc.

Example of safe bus stop – should be provided with safe facilities for crossing pedestrians

Page 63

2.5 RSI reporting

The approx. 580 km SEETO network inspected (as pilots), have been classified by Regional Participants and sections for reporting purposes. Target was to deliver one report for each separate road section inspected.

The result is 24 separate (and standalone) RSI pilot reports for the 6 Regional Participants of SEETO.

The pilot RSI reports have been submitted to each Regional Participant, according to the following table. The submission status below is at the end of March. Submission Corridor Name Section Start Node Section End Node Date (2018) ALB 1 E762 SH1 Shkoder-Koplik Road Start Tuzit Start of By Pass Koplik 23 April ALB 2 E 762 F. Kruje - Lezhe Overpassing F. Kruje r/a in Lezha exit 26 April ALB 3 SH 2 Tirane - Durres Overpassing Kamez I/C of By Pass Shkozet 24 April ALB 4 E 853 Fier - Vlore I/C to Aulona Road I/C to rd Sinan Ferhati 26 April BiH 1 Route 2a Е-661 (M5) Jajce Jug Donji Vakuf 31 March BiH 2 Corridor Vc Е-73 (M 17) Karuše Ozimica 28 February BiH 3 Corridor Vc Е-73 (M 17) Ozimica Topčić Polje 28 February BiH 4 Corridor Vc Е-73 (M 17) Jablanica Potoci 30 March MKD 1 А3 Bitola Prilep 29 January MKD 2 А2 Stracin Kriva Palanka 31 January MKD 3 R-106 Prilep Drenovo 16 April MKD 4 R-106 Drenovo Rosoman 16 April KOS 1 R6b Fushe Kosove Gjurgjice/ R7-R6b I/C 10 April KOS 2 R6b Gjurgjice/ R7-R6b I/C Kijeve 29 March KOS 3 R6a Veternik/N‐2 N-25.2 I/C Lipjan/N-2 & N-25 r/a 10 April MNE 1 Route 4 Podgorica Mioska 31 January SRB 1 Route 4 IB22 Orlovaca Stepojevac 10 April SRB 2 Route 4 IB22 Stepojevac Celije 26 March SRB 3 Route 5 IB23 Vrnjci(Ugljarevo) Kamidzora 26 March SRB 4 Route 5 IB22 Preljina Mrcajevci 14 February SRB 5 Route 9 IB21 Petrovaradin Sremska Kamenica 26 March SRB 6 Corridor X A1 Bubanj Potok Mali Pozarevac 30 March SRB 7 Corridor X A1 Beograd (Dobanovci) Bubanj Potok 13 April SRB 8 Corridor Xb A1 Feketic Sirig 23 March

All RSI pilot reports were submitted by middle of April. 2.6 Conclusion

Following the completion of the missions to undertake pilot RSIs in each of the Regional Participants, the opinion from each participant is that the missions were very valuable in raising awareness of the inspection process and its benefits.

Page 64

The RSI reports produced as part of this assignment will also be a useful tool for participants to give some guidance on methodology, types of safety hazards, etc.

The Road Safety Inspections (580 km in total, 24 individual pilot reports) carried out in the WB6 Regional Participants showed overall lack of maintenance leading to many of the road safety problems.

In the RSIs carried out, there were many common issues shared within all Regional Participants. This includes the overall lack of maintenance which requires urgent action. Crash barriers (missing, inadequate, damaged, etc.), property accesses and high operating speed vehicles passing through villages are the most common hazards identified in all Regional Participants.

Consistent use of unsafe barrier terminals, short barrier lengths, missing barriers, gaps in barrier and outdated bridge parapet implementation were common issues identified in all inspections. Commercial and residence accesses, especially at single carriageway roads (non-motorway), are safety hazards that need urgent confrontation. Junctions and interchanges are usually with many problems, outdated design and missing elements. Roads passing through built-up areas or areas with concentrated pedestrian movements do not have adequate infrastructure elements for vulnerable road users.

The 3-year RSI plan incorporated EuroRAP-IRAP and traditional RSI (as per SEETO Guidelines), as appropriate.

Page 65

3 Component 2 - Road Safety Audit

This component helps to ensure that road safety audits are carried out as per Directive 2008/96/EC on all projects on the core and comprehensive road network. The three activities contained within this component are:

Activity 1 Compile a list of all expected rehabilitation and new construction road projects on the core and comprehensive road network that are currently at the concept or preliminary design stage in the six Western Balkans Regional Participants.

Activity 2 Prepare a plan to undertake road safety audits, at various stages as per Directive 2008/96/EC and SEETO’s Road Safety Audit Guidelines for the list of projects prepared under Activity 1. Such plan should include the Audit stages that are required, as well as the required Auditors inputs and indicative costs.

Activity 3 Undertake Road Safety Audits for a sample of six projects, one in every SEETO Regional Participant.

3.1 List of Rehabilitation and Construction Projects

Activity 1, involves compiling a list of all rehabilitation and new construction road projects on the core and comprehensive road network that are currently under preparation. The list should include the following details for every project:  Engineering scope and estimated cost  Current preparation stage of every project (e.g. concept, preliminary design, detailed design or bidding stage)  Planned implementation schedule  Financing sources.

Initial requests were sent by SEETO in June 2017 at project commencement to all SEETO Regional Participants for a list of all expected rehabilitation and new construction road projects on the core and comprehensive road network that are currently at the concept, preliminary or main design stage. Information was received from Serbia, Montenegro and Bosnia and Herzegovina at this stage.

This information has been used to feed into Activity Two and according to the Work Plan the deadline for information was September 2017. However, due to slow responses from SEETO participants this information remains incomplete.

In September and October, the RSA Project Team requested information from those remaining Regional Participants from which nothing had previously been received. The RSA Project Team also anticipate filling any information gaps during missions for Activity 3 Pilot RSA's. At the beginning of 2018 SEETO sent a reminder to Regional Participants on outstanding data supplemented with a priority on data needed and data that would be nice to have.

This information has been provided in the following sections and any current information gaps highlighted. The tables below represent the information in the raw format as received by the SEETO Participants. The legend below explains the information contained in each column: Type of Road: Whether main or motorway

Page 66

Design Stage: Concept, Preliminary. Detailed, Bidding Works Type: Rehabilitation, upgrade or new construction Start/End: Construction timeline Total Cost: Total Construction Costs Budget: Cost of works met by state WB: Cost met by the World Bank or other IFIs Funding Source: Funded by state or IFIs

Page 67

3.1.1 Albania - List of Rehabilitation and Construction Projects identified

Section Section Lengt SEETO Type of Design Total cost Budget funding Route Works type: Start: End: WB (M€) Start End h (km) Network Road stage (M€) (M€) source SH2 Tirana Vore 16 Core Main road Detailed Resurfacing 2017 2021 0.945 0.4723 0.473 WB Corridor VIII SH2 Vore Durres 22 Core Main Road Detailed Resurfacing 2017 2021 3.257 1.628 1.628 WB Corridor VIII Resurfacing SH4-N Rrogozhin Durres 37 Core Main road New 2017 2021 3.982 1.991 1.991 WB Corridor VIII e construction SH4-N Rrogozhine Lushnje 20 Core Main Road Detailed Resurfacing 2017 2021 0.07 0.035 0.035 WB Corridor VIII Resurfacing SH4-N Lushnje Fier 22 Core Main road New 2017 2021 9.552 4.776 4.776 WB Corridor VIII construction Resurfacing SH8 Comprehensive Fier Vlore 39 Main Road Detailed New 2017 2021 1.155 0.577 0.577 WB Corridor VIII Core construction SH7 Elbasan Librazhd 25 Core Main road Detailed Resurfacing 2017 2021 1.455 0.727 0.727 WB Corridor VIII SH7 Librazhd Perrenjas 34 Core Main Road Detailed Resurfacing 2017 2021 0.846 0.423 0.423 WB Corridor VIII Resurfacing A3 Tirana Elbasan 32 Core Main road Detailed New 2017 2021 1.055 0.527 0.527 WB Corridor VIII construction SH1-N Comprehensive Shkoder Lezhe 42 Main Road Detailed Resurfacing 2017 2021 1.420 0.710 0.710 WB Route 1 Core SH1-N Lezhe Milot 13 Core Main road Detailed Resurfacing 2017 2021 1.615 0.807 0.807 WB Route 1 SH1-N Fushe Mamurras 14 Core Main road Detailed Resurfacing 2017 2021 1.222 0.611 0.611 WB Route 1 Kruje SH1-N Fushe Kruje Vore 13 Core Main Road Detailed Resurfacing 2017 2021 2.494 1.247 1.247 WB Route 1 Resurfacing SH1-N Conceptua Hani i Hotit Shkoder 35 Comprehensive Main road New 2017 2021 0.859 0.4295 0.4295 WB Route 2b l construction SH4-N Fier Tepelene 71 Core Main road Detailed Resurfacing 2017 2021 4.438 2.219 2.219 WB Route 2c SH4-O Tepelene Gjirokaster 23 Core Main Road Detailed Resurfacing 2017 2021 0.786 0.393 0.393 WB Route 2c SH4-O Gjirokaster Kakavija 31 Core Main road Detailed Resurfacing 2017 2021 4.438 2.219 2.219 WB Route 2c SH1-N Milot Mamurras 14 Core Main Road Resurfacing 2017 2021 0.752 0.376 0.376 WB Route 7 A1 Concess Milot Rreshen 25.9 Core Main Road Detailed Resurfacing 2017 2021 Route 7 64.0

Page 68

Section Section Lengt SEETO Type of Design Total cost Budget funding Route Works type: Start: End: WB (M€) Start End h (km) Network Road stage (M€) (M€) source A1 Rreshen Kalimash 60.2 Core Main road Resurfacing 2017 2021 Concess Route 7 A1 Kalimash Morine 28.3 Core Main Road Resurfacing 2017 2021 Concess Route 7 SH3 Perrenjas Pogradec 27 Core Main Road Detailed Resurfacing 2017 2021 2.696 1.348 1.348 WB Resurfacing SH3 Pogradec Korce 40 Comprehensive Main road Partially new 2017 2021 1.400 0.700 0.700 WB construction Kapeshtic SH3 Korce 32 Comprehensive Main Road Detailed Resurfacing 2017 2021 1.058 0.529 0.529 WB e Qafe SH3 Perrenjas 13 Core Main road Detailed Resurfacing 2017 2021 0.295 0.147.5 0.147.5 WB Thane

Missing information on Design Stage for 6 projects is needed. Sources in the table: Black letters and numbers – Regional Participant Blue letters and numbers – Consultant’s information Red letters and numbers – SEETO or SEETO MAP 2018 Final

Page 69

3.1.2 Bosnia and Herzegovina (Federation of Bosnia and Herzegovina and Republic of Srpska) - List of Rehabilitation and Construction Projects identified

Route Name Start End Length Type SEETO Network Design Works: Start: End: Total cost Funding (km) stage (M€) source Route 2a M-16 (Klašnice 2-Šargovac) 33+010,00 42+463,00 9.453 Main Core Rehabilitation 2017 Route 2a M-16 (Šargovac - Banja Luka 1) 42+463,00 46+374,00 3.911 Main Core Rehabilitation 2017 M-16 (Banja Luka 1 - Banja Luka Route 2a 46+374,00 47+430,00 1.056 Main Core Rehabilitation 2017 (Čajavec)) M-16 (Banja Luka (Čajavec)-Banja Luka Route 2a 47+430,00 48+386,00 0.956 Main Core Rehabilitation 2017 (Rudarska)) M-16 (Banja Luka (Rudarska)-Srpske Route 2a 48+386,00 52+704,00 4.318 Main Core Rehabilitation 2017 - Toplice) Route 2a M-16 (Srpske Toplice-Karanovac) 52+704,00 59+396,00 6.692 Main Core Rehabilitation 2017 - Main 2017 National Route 2a M-16 (Karanovac-Crna Rijeka) 59+396,00 92+182,00 32.786 Main Core Rehabilitation 2018 3 design 2018 budget Route 2a M-16 (Crna Rijeka-granica RS (Ugar)) 92+182,00 98,717,00 6.535 Main Core Rehabilitation 2017 - - Partially Main June, Aug, Route 2b Brod na Drini - Hum Brod na Drini Hum Main Comprehensive reconstruction 3.5 EBRD design 2016 2018 and partially new EBRD (?) Main 2017 Route 9a "9. Januar" Laktaši - Prnjavor 0+000,00 35+912,51 35.913 Motorway Comprehensive New 2018 47+ and EIB design 2015 (47)

Missing Design Stage and End date that is needed. Nice to have: Total cost (M€), Budget cost (state) (M€), IFI cost (M€) and Funding source. Sources in the table: Black letters and numbers – Regional Participant Blue letters and numbers – Consultant’s information Green letters and numbers – EBRD and EIB information Brown letters and numbers (road section added by the Consultant) – WBIF IPF Bimonthly Report 1801-1802 Orange letters and numbers – PE “Republic of Srpska Roads” Procurement plan for 2018 (http://www.putevirs.com/korisnik/dokumenti/Plan_javnih_nabavki_za_objavu_2018.pdf)

Page 70

3.1.3 The former Yugoslav Republic of Macedonia - List of Rehabilitation and Construction Projects identified

Route Name Start End Length Type SEETO Design stage Works: Start: End: Total cost Funding (km) Network (M€) source Corridor VIII (Kicevo – Podmolje) and Route 8 Kicevo – Ohrid motorway New (Podmolje – Ohrid) Rankovce – Kriva Corridor VIII express New Palanka EIB(100),EBR Demir Kapija – Demir Corridor X Smokvica 28,2 motorway Detailed New 2015 2018 286 D (107) and Smokvica Kapija EC(grant 45) Corridor X Gradsko – Veles motorway Rehabilitation Corridor X Miladinovci – Petrovec motorway Rehabilitation Corridor X Negotino - Demir Kapija motorway Rehabilitation Corridor X Kumanovo- Miladinovci motorway Rehabilitation Corridor X Smokvica – Gevgelija motorway Rehabilitation Corridor Xd Gradsko – Prilep express New Construction Corridor Xd Prilep – Raec Prilep Raec 8.8 Core and 5.0 reconstruction Apr Mar Route 6a Skopje - Blace Skopje Blace motorway Core Preliminary New 1.0 EBRD 2015 2018 Interchange express Route 10 Stip – Radovis Bucim 37 New 2015 78.4 EBRD (64) Tri Cesmi Route 10 Miladinovci – Stip motorway New express It is not part of Stip – Kocani SEETO road New network

Brief project list provided in June 2017 and further information was requested in Oct 2017 and March 2018. Additional information collected by consultant in May 2018. Nothing further received to date. Most data is missing including the essential ones such as Route, Start and end year, and design stage. Sources in the table: Black letters and numbers – Regional Participant Green letters and numbers – EBRD, EIB or CFCD (Central Financing and Contracting Department- http://cfcd.finance.gov.mk/) information Red letters and numbers – SEETO or SEETO MAP 2018 Final Brown letters and numbers (road section added by the Consultant) – WBIF IPF Bimonthly Report 1801-1802 (Skopje – Blace) and SEETO MAP 2018 Final (Prilep – Raec)

Page 71

3.1.4 Kosovo - List of Rehabilitation and Construction Projects identified

Length SEETO Design Total cost Funding Route Name Start End Type Works: Start: End: (km) Network stage (M€) source Route 6a Smrekonicë Mitrovivë 5 Main Comprehensive Preliminary Rehabilitation 2018 2020 WBIF, Detailed Kosovo Route 6b Kijevë Zahaq Kijevë Zahaq 30 Main Comprehensive New construction 2018 2021 143.2 Preliminary budget, EBRD, EIB Route 6b Pristina-Peje Pristina Peje 40 Main Comprehensive Detailed New 2015 83.0 R6 Merdare/ Concept Sep Oct Besi Merdare/Serbia Border Besi 26.5 Main Core New construction 150 EBRD Route 7 Serbia border Preliminary 2017 2018

End date missing for one projects which is needed. Nice to have would be start/end chainages, costs and financing sources that have not been provided at the time of writing. Sources in the table: Black letters and numbers – Regional Participant Green letters and numbers – EBRD information Red letters and numbers – SEETO or SEETO MAP 2018 Final Brown letters and numbers (road section added by the Consultant) – WBIF IPF Bimonthly Report 1801-1802 and SEETO MAP 2018 Final

Page 72

3.1.5 Montenegro - List of Rehabilitation and Construction Projects identified

Route Name of the Project Start End Length Type of SEETO Technical Design Works: Start: End: Total cost EBRD EIB (sections) (km) road Network description stage: (M€) (M€) (M€) Route 1 Debeli brijeg - H novi 854+500 859+500 5 National Core Rehabilitation Rehabilitation 2018 2018 Route 1 Debeli brijeg - H novi 859+500 861+900 2.4 National Core Rehabilitation Rehabilitation 2018 2018 Route 1 Petrovac - Bar 1+200 18+000 16.8 National Core Rehabilitation Rehabilitation 2017 2018 EIB 74.2 (WBIF Route 2b Šćepan polje – Plužine 0+000 25+000 25 National Core Reconstruction New 70 0 grant 4.2) Route 2b Zaborje - Jasenovo polje 46+000 65+000 19 National Core Rehabilitation Rehabilitation 2019 2020 Route 2b Jasenovo polje - Vidrovan 65+000 70+500 5.5 National Core Rehabilitation Rehabilitation 2018 2018 Route 2b Danilovgrad - Podgorica 114+500 129+500 15 National Core Reconstruction New 2017 2018 EBRD Route 3 Lipci - Ljuta 15 National Core Rehabilitation Rehabilitation 2018 2018 Route 4 Tivat -Budva 885+500 901+500 16 National Core Reconstruction New 2017 2019 EBRD Route 4 Barski most - Dobrakovo 131+200 135+300 4.1 National Core Rehabilitation Rehabilitation 2017 2018 EIB Route 5 Kamenovo - Petrovac 911+700 923+000 10.3 National Core Rehabilitation Rehabilitation 2017 2018 EIB Route 5 Dobrakovo - Bijelo polje 135+300 145+600 10.3 National Core Rehabilitation Rehabilitation 2017 2018 EIB Obilaznica Bijelo polje - Route 6a 150+000 153+300 3.3 National Core Rehabilitation Rehabilitation 2017 2018 Ribarevine Route 6a Ribarevine - Poda 1089+872 1104+300 14.5 National Core Rehabilitation Rehabilitation 2017 2018 EIB Route 6a Poda -Berane 1104+559 1117+300 12.8 National Core Rehabilitation Rehabilitation 2017 2018 EIB Route 6a Berane - Tunel Lokve 1117+300 1132+507 15.2 National Core Rehabilitation Rehabilitation 2017 2018 EIB Route 6a Tunel Lokve - Rožaje 1133+507 1148+200 14.7 National Core Rehabilitation Rehabilitation 2017 2018 EIB Route 6a Obilaznica Rožaje II Phase 0 5+000 5 National Core New New 2017 2019 EIB Route 6a Rožaje - Špiljani 1148+200 1168+200 20 National Core Rehabilitation Rehabilitation 2017 2019 EBRD Route 6b Andrijevica - Berane 0 16+000 16 National Core Rehabilitation Rehabilitation 2018 2019 Route 7 Bar - Ulcinj 23+900 45+300 21.4 National Core Rehabilitation Rehabilitation 2018 2019 Route 7 Ribarevine - Lepenac 1075+200 1089+850 14.6 National Core Rehabilitation Detailed Rehabilitation 2017 2018 EIB Route 8 Ulcinj - Krute 45+300 60+000 14.7 National Core Rehabilitation Rehabilitation 2017 2018 EIB Route 8 Mojkovac - Kolašin 1067+500 1046+150 21.35 National Core Rehabilitation Preliminary Rehabilitation 2018 2019 Route 9a Kolašin - Crkvine 1037+200 1046+150 8.95 National Core Rehabilitation Preliminary Rehabilitation 2018 2019 Route 10 Tuneli 7,8,9 1010+530 1011+200 0.67 National Core Reconstruction New 2018 2019 Obilaznica golubovci - Route 11 946+300 962+800 16.5 National Core Rehabilitation Rehabilitation 2018 2019 Virpazar Virpazar - Paštrovačka Route 12 946+300 931+500 14.8 National Core Rehabilitation Rehabilitation 2018 2020 gora Paštrovačka gora - Route 13 931+500 924+000 7.5 National Core Rehabilitation Rehabilitation 2018 2020 Petrovac

Design Stage (Preliminary, Conceptual, Detailed, etc.) and partly Start and End dates that is needed was not provided at time of writing. Would be nice to have: Total cost (M€), Budget cost (state) (M€), IFI cost (M€) and Funding source has been provided at time of writing. Sources in the table: Black letters and numbers – Regional Participant Green letters and numbers – EBRD information

Page 73

3.1.6 Serbia - List of Rehabilitation and Construction Projects identified

Length Type of Design Project Total cost Budget Funding Route Start End SEETO Network Works: Project end: WB (M€) (km) road stage start: (M€) (M€) source Corridor X Ražanj 1 Paraćin 1 24.781 Highway Core Detailed Rehabilitation 2018 Jan-19 0.597 EBRD Corridor X Aleksinac Ražanj 1 23.645 Highway Core Detailed Rehabilitation 2018 Jan-19 Corridor X Ruma 1 Pećinci 1 12.99 Highway Core Detailed Rehabilitation 2018 Dec-18 Corridor X Sremska Mitrovica Ruma 1 13.75 Highway Core Detailed Rehabilitation 2018 Jan-19 Corridor X Niš 1 (Trupale) Niš 3 (Batušinac) 9 Highway Core Detailed Rehabilitation 2018 Jan-19 0.735 EBRD Corridor X Niš 3 (Batušinac) Niš 1 (Trupale) 8.998 Highway Core Detailed Rehabilitation 2018 Jan-19 Corridor X Ražanj 1 Aleksinac 23.655 Highway Core Detailed Rehabilitation 2018 Jun-19 0.198 EBRD Corridor X Brestovac 1 Doljevac 1 6.155 Highway Core Detailed Rehabilitation 2018 Apr-19 0.051 EBRD Border CRO/SER Corridor X Kuzmin 21.805 Highway Core Detailed Rehabilitation 2017 Feb-18 0.0055 0.191 0.005462126 EBRD (Batrovci) Route 3 Požega Kotroman 60 Highway Comprehensive Prelim construction 2006 830 Border of APV WB/EIB/PER Route 4 Pančevo (Kovin) 2.766 Motorway Core Detailed Rehabilitation 2017 Jul-18 0.261 (Pančevo) S Border of APV Route 4 Pančevo (Kovin) 2.754 Motorway Core Detailed Rehabilitation 2017 Jul-18 (Pančevo) WB/EIB/PER Route 4 Kneževići Užice 3 15.234 Motorway Core Detailed Rehabilitation 2017 Jun-18 0.00381 0.153 0.003816098 S Route 4 Požega Boljare 107 Highway Core Prelim construction 2006 1830.9 Route 5 Pojate Preljina 110 Highway Comprehensive Prelim construction 2015 640 WB/EIB/PER Route 7 Merošina Prokuplje (Orljane) 11.503 Motorway Core Detailed Rehabilitation 2017 Nov-18 0.243 S Route 7 Beloljin Rudare 24.171 Motorway Core Detailed Rehabilitation 2017 Nov-18 Feasibility 2014 EIB (106), Route 7 Niš Pločnik 40 Highway Core construction Dec 2018 Prelim Feb 2017 WBIF (9.595), EIB, WBIF, 855 Serbia budget Serbia Feasibility 2014 212.595 (10.15), other budget, other Route 7 Pločnik Merdare 37 Highway Core construction Dec 2018 Prelim Feb 2017 sources sources (86.85) 0.0038188 WB/EIB/PER Route 9a Irig 2 Ruma 1 (motorway) 15.245 Expressway Comprehensive Detailed Rehabilitation 2017 Feb-18 0.136 0.003818854 54 S Route 9a Novi Sad Ruma Expressway Comprehensive Prelim construction 2008 Šabac Route 9a Ruma 112 Expressway Comprehensive Prelim construction 2008 360

Route 9a Šabac Loznica Expressway Comprehensive Prelim construction 2008

Missing End date (6 sections) that is needed. Nice to have: Budget cost (state) (M€), IFI cost (M€), Funding source and Total cost (M€). Sources in the table: Black letters and numbers – Regional Participant Red letters and numbers – WBIF IPF Bimonthly Report 1801-1802 and SEETO MAP 2018 Final

Page 74

3.2 Road Safety Audit Plan for identified projects

Under Activity 2, the RSA Project Team has prepared a three-year plan (2018/19-2020/21) to undertake road safety audits, at various stages as per Directive 2008/96/EC and SEETO’s Road Safety Audit Guidelines for the list of projects identified under Activity 1. The plan includes the Audit stages that are required for each project, as well as the required Auditors inputs and indicative costs.

All roads in the plan are on the core and comprehensive network and if they become TEN-T then RSA is mandatory according to EU Directive 2008/96/EC. RSA should be done on both upgrading projects and new projects.

Currently, only Republic of Srpska within Bosnia and Herzegovina (and very recently also Albania) have trained and certified Road Safety Auditors within the Region. Therefore, an assumption has been made that Road Safety Audits carried out on the projects identified within Activity 1 will be carried out by appropriately qualified and experienced consultants.

In those projects that are IFI funded, their policies often insist on at least Design Stage Audits being carried out on those projects. Therefore, the cost of these Audits will be met by the IFIs as part of the project.

For this purpose, the following estimated costs have been provided in the table below for each stage of Audit specified in the SEETO RSA Guidelines. In addition, time estimates have been provided in person days and assuming an Audit Team of two, an Audit Team Leader (TL) and Audit Team Member (TM). These costs are based on previous Audit assignments carried out in the Region by international consultants but as more Auditors become certified within the region these costs should be revised. A cost of 600 Euros per auditor per day has been provided as an estimate.

It should be noted that some larger projects may have many design drawings for review so may take more person days and some projects have already commenced so only completion of later stages of audit are possible. This is therefore an estimate only. The costings could vary depending on scheme complexity so a typical figure has been provided as a guide. Estimated Cost Time Estimate (Euros) Audit Stage (person days) may vary due to may vary due to scheme size scheme size Stage One – Prelim Design 10,200 17 (7 TL, 10 TM) Stage Two – Detailed Design 12,000 20 (8 TL, 12 TM) Stage Three – Pre-Opening 9,000 15 (5 TL, 10 TM) Stage Four – Early Operation 9,000 15 (5 TL, 10 TM)

Page 75

3.2.1 Albania – Plan for Road Safety Audits

Total Cost Design Stage One Stage Stage Three Stage Four Total Person Route Section Start Section End Length (km) START: END: RSA Inputs Stage RSA Two RSA RSA RSA Days RSAs (Euros) Corridor VIII Tirana Vore Detailed 16 2017 2021 Not Req 2018 2021 2021 30,000 50 Corridor VIII Vore Durres Detailed 22 2017 2021 Not Req 2018 2021 2021 30,000 50 Corridor VIII Durres Rrogozhine 37 2017 2021 Not Req TBC 2021 2021 18,000 30 Corridor VIII Rrogozhine Lushnje Detailed 20 2017 2021 Not Req 2018 2021 2021 30,000 50 Corridor VIII Lushnje Fier 22 2017 2021 Not Req TBC 2021 2021 18,000 30 Corridor VIII Fier Vlore Detailed 39 2017 2021 Not Req 2018 2021 2021 30,000 50 Corridor VIII Elbasan Librazhd Detailed 25 2017 2021 Not Req 2018 2021 2021 30,000 50 Corridor VIII Librazhd Perrenjas Detailed 34 2017 2021 Not Req 2018 2021 2021 30,000 50 Corridor VIII Tirana Elbasan Detailed 32 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 1 Shkoder Lezhe Detailed 42 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 1 Lezhe Milot Detailed 13 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 1 Mamurras Fushe Kruje Detailed 14 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 1 Fushe Kruje Vore Detailed 13 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 2b Hani i Hotit Shkoder 35 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 2c Fier Tepelene Detailed 71 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 2c Tepelene Gjirokaster Detailed 23 2017 2021 Not Req 2018* 2021 2021 18,000 50 Route 2c Gjirokaster Kakavija Detailed 31 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 7 Milot Mamurras 14 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 7 Milot Rreshen 25.9 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 7 Rreshen Kalimash 60.2 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 7 Kalimash Morine 28.3 2017 2021 Not Req 2018 2021 2021 30,000 50 SH3 Perrenjas Pogradec Detailed 27 2017 2021 Not Req 2018 2021 2021 30,000 50 SH3 Pogradec Korce 40 2017 2021 Not Req 2018 2021 2021 30,000 50 SH3 Korce Kapeshtice Detailed 32 2017 2021 Not Req 2018 2021 2021 30,000 50 SH3 Perrenjas Qafe Thane Detailed 13 2017 2021 Not Req 2018 2021 2021 30,000 50

Page 76

3.2.2 Bosnia and Herzegovina – Plan for Road Safety Audits

Design Length Stage Stage Stage Three Stage Four Total Cost RSA Total Person Route Name Start: End stage (km) One RSA Two RSA RSA RSA Inputs (Euros) Days RSAs Main Route 2a M-16 (Klašnice 2-Šargovac) 9.453 2017 Not Req 2018 TBC TBC 12,000 20 design Route 2a M-16 (Šargovac - Banja Luka 1) 3.911 2017 Not Req 2018 TBC TBC 12,000 20

Route 2a M-16 (Banja Luka 1 - Banja Luka (Čajavec)) 1.056 2017 Not Req 2018 TBC TBC 12,000 20 M-16 (Banja Luka (Čajavec)-Banja Luka Route 2a 0.956 2017 Not Req 2018 TBC TBC 12,000 20 (Rudarska)) M-16 (Banja Luka (Rudarska)-Srpske Route 2a 4.318 2017 Not Req 2018 TBC TBC 12,000 20 Toplice) Route 2a M-16 (Srpske Toplice-Karanovac) 6.692 2017 Not Req 2018 TBC TBC 12,000 20 Main 2017/ Route 2a M-16 (Karanovac-Crna Rijeka) 32.786 2018 Not Req 2018 2018 2018 30,000 50 design 2018 Route 2a M-16 (Crna Rijeka-granica RS (Ugar)) 6.535 2017 Not Req 2018 TBC TBC 12,000 20 Main Route 2b Brod na Drini – Hum 2016 2018 Not Req Not Req 2018 2019 18,000 30 design Main Route 9a "9. Januar" Laktaši - Prnjavor 35.913 2018 2018 Not Req 2018 2018 2019 30,000 50 design

Page 77

3.2.3 The former Yugoslav Republic of Macedonia– Plan for Road Safety Audits

Length Stage Stage Two Stage Three Stage Four Total Cost RSA Total Person Route Name Design stage (km) Start: End: One RSA RSA RSA RSA Inputs (Euro) Days RSAs Corridor VIII (Kicevo – Podmolje) and Route 8 Kicevo – Ohrid (Podmolje – Ohrid) Corridor VIII Rankovce – Kriva Palanka

Corridor X Demir Kapija – Smokvica Detailed 28,2 2015 2018 Not Req Not Req 2018 2019 18,000 30

Corridor X Gradsko – Veles

Corridor X Miladinovci – Petrovec

Corridor X Negotino - Demir Kapija

Corridor X Kumanovo- Miladinovci

Corridor X Smokvica – Gevgelija

Corridor Xd Gradsko – Prilep

Corridor Xd Prilep – Raec 8.8

Route 6a Skopje - Blace Preliminary Apr, 2015 Mar 2018 Not Req Not Req 2018 2018 18,000 30

Route 10 Stip – Radovis 37 2015

Route 10 Miladinovci – Stip

Page 78

3.2.4 Kosovo – Plan for Road Safety Audits

Length Stage Stage Stage Stage Total Cost RSA Total Person Route Name DESIGN STAGE Start: End: (km) One RSA Two RSA Three RSA Four RSA Inputs (Euros) Days RSAs Preliminary/Detaile Route 6a Smrekonicë Mitrovivë 5 2018 2020 Not Req 2018* 2020 2020 18,000 30 d Route 6b Kijevë Zahaq Detailed 30 2018 2021 Not Req 2018 2021 2021 30,000 50

Route 6b Pristina-Peje Detailed 40 2015 Not Req TBC TBC TBC Sep, Oct, Route 7 Besi Merdare/Serbia Border Concept 26.5 Not Req Not Req 2018 2019 18,000 30 2017 2018

*Stage Two Audit completed as pilot RSA as part of this project

Page 79

3.2.5 Montenegro – Plan for Road Safety Audits

Stage Stage Stage Three Stage Total Cost RSA Total Person Route Name of the Project (sections) Design Stage Length (km) Start: End: One RSA Two RSA RSA Four RSA Inputs (Euros) Days RSAs Route 1 Debeli brijeg - H novi 5 2018 2018 Not Req Not Req 2018 2019 18,000 30 Route 1 Debeli brijeg - H novi 2.4 2018 2018 Not Req Not Req 2018 2019 18,000 30 Route 1 Petrovac - Bar 16.8 2017 2018 Not Req Not Req 2018 2019 18,000 30 Route 2b Šćepan polje – Plužine 25 Route 2b Zaborje - Jasenovo polje 19 2019 2020 2019 2019 2020 2020 40,200 67 Route 2b Jasenovo polje - Vidrovan 5.5 2018 2018 Not Req 2018 2018 2019 30,000 50 Route 2b Danilovgrad - Podgorica 15 2017 2018 Not Req Not Req 2018 2019 18,000 30 Route 3 Lipci - Ljuta 15 2018 2018 Not Req Not Req 2018 2019 18,000 30 Route 4 Tivat -Budva 16 2017 2019 Not Req Not Req 2019 2020 18,000 30 Route 4 Barski most - Dobrakovo 4.1 2017 2018 Not Req Not Req 2018 2019 18,000 30 Route 5 Kamenovo - Petrovac 10.3 2017 2018 Not Req Not Req 2018 2019 18,000 30 Route 5 Dobrakovo - Bijelo polje 10.3 2017 2018 Not Req Not Req 2018 2019 18,000 30 Route 6a Obilaznica Bijelo polje - Ribarevine 3.3 2017 2018 Not Req 2018 2018 2019 30,000 50 Route 6a Ribarevine - Poda 14.5 2017 2018 Not Req 2018 2018 2019 30,000 50 Route 6a Poda -Berane 12.8 2017 2018 Not Req 2018 2018 2019 30,000 50 Route 6a Berane - Tunel Lokve 15.2 2017 2018 Not Req 2018 2018 2019 30,000 50 Route 6a Tunel Lokve - Rožaje 14.7 2017 2018 Not Req 2018 2018 2019 30,000 50 Route 6a Obilaznica Rožaje II Phase 5 2017 2019 Not Req Not Req 2019 2020 18,000 30 Route 6a Rožaje - Špiljani 20 2017 2019 Not Req Not Req 2019 2020 18,000 30 Route 6b Andrijevica - Berane 16 2018 2019 Not Req 2018 2019 2020 50,000 50 Route 7 Bar - Ulcinj 21.4 2018 2019 Not Req 2018 2019 2020 50,000 50 Route 7 Ribarevine - Lepenac Detailed 14.6 2017 2018 Not Req 2018* 2018 2018 18,000 30 Route 8 Ulcinj - Krute 14.7 2017 2018 Not Req Not Req 2018 2019 18,000 30 Route 8 Mojkovac - Kolašin Preliminary 21.35 2018 2019 2018* 2018 2019 2019 30,000 50 Route 9a Kolašin - Crkvine Preliminary 8.95 2018 2019 2018* 2018 2019 2019 30,000 50 Route 10 Tuneli 7,8,9 0.67 2018 2019 Not Req 2018 2019 2020 50,000 50 Route 11 Obilaznica golubovci - Virpazar 16.5 2018 2019 Not Req 2018 2019 2020 50,000 50 Route 12 Virpazar - Paštrovačka gora 14.8 2018 2020 Not Req 2018 2020 2021 30,000 50 Route 13 Paštrovačka gora - Petrovac 7.5 2018 2020 Not Req 2018 2020 2021 30,000 50

*Stage Two and Stage One Audits completed as pilot RSA as part of this project

Page 80

3.2.6 Serbia – Plan for Road Safety Audits

Total Cost Total Design Length Project Project Stage Stage Stage Three Stage Route Name of the Project (sections) RSA Inputs Person stage (km) start: end: One RSA Two RSA RSA Four RSA (Euros) Days RSAs Corridor X Ražanj 1 Paraćin 1 Detailed 24.781 2018 Jan-19 Not Req 2018 2019 2019 30,000 50 Corridor X Aleksinac Ražanj 1 Detailed 23.645 2018 Jan-19 Not Req 2018 2019 2019 30,000 50 Corridor X Ruma 1 Pećinci 1 Detailed 12.99 2018 Dec-18 Not Req 2018 2018 2019 30,000 50 Corridor X Sremska Mitrovica Ruma 1 Detailed 13.75 2018 Jan-19 Not Req 2018 2019 2019 30,000 50 Corridor X Niš 1 (Trupale) Niš 3 (Batušinac) Detailed 9 2018 Jan-19 Not Req 2018 2019 2019 30,000 50 Corridor X Niš 3 (Batušinac) Niš 1 (Trupale) Detailed 8.998 2018 Jan-19 Not Req 2018 2019 2019 30,000 50 Corridor X Ražanj 1 Aleksinac Detailed 23.655 2018 Jun-19 Not Req 2018 2019 2019 30,000 50 Corridor X Brestovac 1 Doljevac 1 Detailed 6.155 2018 Apr-19 Not Req 2018 2019 2019 30,000 50 Corridor X Border CRO/SER (Batrovci) Kuzmin Detailed 21.805 2017 Feb-18 Not Req 2018 2018 2018 30,000 50 Route 3 Požega Kotroman Prelim 60 2006 Not Req TBC TBC TBC Route 4 Border of APV (Pančevo) Pančevo (Kovin) Detailed 2.766 2017 Jul-18 Not Req 2018 2018 2019 30,000 50 Route 4 Pančevo (Kovin) Border of APV (Pančevo) Detailed 2.754 2017 Jul-18 Not Req 2018 2018 2019 30,000 50 Route 4 Kneževići Užice 3 Detailed 15.234 2017 Jun-18 Not Req 2018 2018 2018 30,000 50 Route 4 Požega Boljare Prelim 107 2006 Not Req Not Req TBC TBC Route 5 Pojate Preljina Prelim 110 2015 Not Req Not Req TBC TBC Route 7 Merošina Prokuplje (Orljane) Detailed 11.503 2017 Nov-18 Not Req 2018 2018 2019 30,000 50 Route 7 Beloljin Rudare Detailed 24.171 2017 Nov-18 Not Req 2018 2018 2019 30,000 50 Route 7 Niš Pločnik Feas 40 2014 2018 2018* 2018 TBC TBC 12,000 20 Route 7 Pločnik Merdare Feas 37 2014 2018 Not Req Not Req 2018 2019 18,000 30 Route 9a Irig 2 Ruma 1 (motorway) Detailed 15.245 2017 Feb-18 Not Req 2018 2018 2018 30,000 50 Route 9a Novi Sad Ruma Prelim 2008 Not Req Not Req TBC TBC Ruma Šabac Route 9a Prelim 2008 Not Req Not Req TBC TBC

Route 9a Šabac Loznica Prelim 2008 Not Req Not Req TBC TBC 112

*Stage One Audit completed as pilot RSA as part of this project

Page 81

3.2.7 Three-year plan by Corridor and Route

The following Table incorporates all information that has been received by the WB6 Regional Participants or collected during investigation process presented by Corridor and Route. Length Stage Stage Stage Stage Four Total Cost RSA Total Person Route Section Start Section End Design Stage (km) Start: End: One RSA Two RSA Three RSA RSA Inputs (Euros) Days RSAs Corridor VIII Tirana Vore Detailed 16 2017 2021 Not Req 2018 2021 2021 30,000 50 Corridor VIII Vore Durres Detailed 22 2017 2021 Not Req 2018 2021 2021 30,000 50 Corridor VIII Durres Rrogozhine 37 2017 2021 Not Req TBC 2021 2021 18,000 30 Corridor VIII Rrogozhine Lushnje Detailed 20 2017 2021 Not Req 2018 2021 2021 30,000 50 Corridor VIII Lushnje Fier 22 2017 2021 Not Req Not Req 2021 2021 18,000 30 Corridor VIII Fier Vlore Detailed 39 2017 2021 Not Req 2018 2021 2021 30,000 50 Corridor VIII Elbasan Librazhd Detailed 25 2017 2021 Not Req 2018 2021 2021 30,000 50 Corridor VIII Librazhd Perrenjas Detailed 34 2017 2021 Not Req 2018 2021 2021 30,000 50 Corridor VIII Tirana Elbasan Detailed 32 2017 2021 Not Req 2018 2021 2021 30,000 50 Corridor VIII (Kicevo – Podmolje) and Kicevo Ohrid Route 8 (Podmolje – Ohrid) Corridor VIII Rankovce Kriva Palanka Corridor X Ražanj 1 Paraćin 1 Detailed 24.781 2018 Jan-19 Not Req 2018 2019 2019 30,000 50 Corridor X Aleksinac Ražanj 1 Detailed 23.645 2018 Jan-19 Not Req 2018 2019 2019 30,000 50 Corridor X Ruma 1 Pećinci 1 Detailed 12.99 2018 Dec-18 Not Req 2018 2018 2019 30,000 50 Corridor X Sremska Mitrovica Ruma 1 Detailed 13.75 2018 Jan-19 Not Req 2018 2019 2019 30,000 50 Corridor X Niš 1 (Trupale) Niš 3 (Batušinac) Detailed 9 2018 Jan-19 Not Req 2018 2019 2019 30,000 50 Corridor X Niš 3 (Batušinac) Niš 1 (Trupale) Detailed 8.998 2018 Jan-19 Not Req 2018 2019 2019 30,000 50 Corridor X Ražanj 1 Aleksinac Detailed 23.655 2018 Jun-19 Not Req 2018 2019 2019 30,000 50 Corridor X Brestovac 1 Doljevac 1 Detailed 6.155 2018 Apr-19 Not Req 2018 2019 2019 30,000 50 Corridor X Border CRO/SER (Batrovci) Kuzmin Detailed 21.805 2017 Feb-18 Not Req 2018 2018 2018 30,000 50 Corridor X Demir Kapija Smokvica Detailed 28,2 2015 2018 Not Req Not Req 2018 2019 18,000 30 Corridor X Gradsko Veles Corridor X Miladinovci Petrovec Corridor X Negotino Demir Kapija Corridor X Kumanovo Miladinovci Corridor X Smokvica Gevgelija Corridor Xd Gradsko Prilep Corridor Xd Prilep Raec 8.8 Route 1 Debeli brijeg H novi 5 2018 2018 Not Req Not Req 2018 2019 18,000 30 Route 1 Debeli brijeg H novi 2.4 2018 2018 Not Req Not Req 2018 2019 18,000 30 Route 1 Petrovac Bar 16.8 2017 2018 Not Req Not Req 2018 2019 18,000 30 Route 1 Shkoder Lezhe Detailed 42 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 1 Lezhe Milot Detailed 13 2017 2021 Not Req 2018 2021 2021 30,000 50

Page 82

Length Stage Stage Stage Stage Four Total Cost RSA Total Person Route Section Start Section End Design Stage (km) Start: End: One RSA Two RSA Three RSA RSA Inputs (Euros) Days RSAs Route 1 Mamurras Fushe Kruje Detailed 14 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 1 Fushe Kruje Vore Detailed 13 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 2a M-16 (Klašnice 2 Šargovac) Main design 9.453 2017 Not Req 2018 TBC TBC 12,000 20 Route 2a M-16 (Šargovac Banja Luka 1) 3.911 2017 Not Req 2018 TBC TBC 12,000 20 Route 2a M-16 (Banja Luka 1 Banja Luka (Čajavec) 1.056 2017 Not Req 2018 TBC TBC 12,000 20 Banja Luka Route 2a M-16 (Banja Luka (Čajavec) 0.956 2017 Not Req 2018 TBC TBC 12,000 20 (Rudarska) M-16 (Banja Luka Route 2a Srpske Toplice) 4.318 2017 Not Req 2018 TBC TBC 12,000 20 (Rudarska) Route 2a M-16 (Srpske Toplice Karanovac) 6.692 2017 Not Req 2018 TBC TBC 12,000 20 2017/ Route 2a M-16 (Karanovac Crna Rijeka) Main design 32.786 2018 Not Req 2018 2018 2018 30,000 50 2018 Route 2a M-16 (Crna Rijeka Granica RS (Ugar) 6.535 2017 Not Req 2018 TBC TBC 12,000 20 Route 2b Brod na Drini Hum Main design 2016 2018 Not Req Not Req 2018 2019 18,000 30 Route 2b Šćepan polje Plužine Detailed 25 Route 2b Zaborje Jasenovo polje 19 2019 2020 2019 2019 2020 2020 40,200 67 Route 2b Jasenovo polje Vidrovan 5.5 2018 2018 Not Req 2018 2018 2019 30,000 50 Route 2b Danilovgrad Podgorica 15 2017 2018 Not Req Not Req 2018 2019 18,000 30 Route 2b Hani i Hotit Shkoder 35 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 2c Fier Tepelene Detailed 71 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 2c Tepelene Gjirokaster Detailed 23 2017 2021 Not Req 2018 2021 2021 18,000 50 Route 2c Gjirokaster Kakavija Detailed 31 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 3 Lipci Ljuta 15 2018 2018 Not Req Not Req 2018 2019 18,000 30 Route 3 Požega Kotroman Prelim 60 2006 Not Req TBC TBC TBC Route 4 Border of APV (Pančevo) Pančevo (Kovin) Detailed 2.766 2017 Jul-18 Not Req 2018 2018 2019 30,000 50 Border of APV Route 4 Pančevo (Kovin) Detailed 2.754 2017 Jul-18 Not Req 2018 2018 2019 30,000 50 (Pančevo) Route 4 Kneževići Užice 3 Detailed 15.234 2017 Jun-18 Not Req 2018 2018 2018 30,000 50 Route 4 Požega Boljare Prelim 107 2006 Not Req Not Req TBC TBC Route 4 Tivat Budva 16 2017 2019 Not Req Not Req 2019 2020 18,000 30 Route 4 Barski most Dobrakovo 4.1 2017 2018 Not Req Not Req 2018 2019 18,000 30 Route 5 Kamenovo Petrovac 10.3 2017 2018 TBC TBC TBC TBC TBC TBC Route 5 Dobrakovo Bijelo polje 10.3 2017 2018 TBC TBC TBC TBC TBC TBC Route 5 Pojate Preljina Prelim 110 2015 Not Req Not Req TBC TBC Apr, Mar Route 6a Skopje Blace Prelim Not Req Not Req 2018 2018 18,000 30 2015 2018 Prelim/Detaile Route 6a Smrekonicë Mitrovivë 5 2018 2020 Not Req 2018 2020 2020 18,000 30 d Route 6a Obilaznica Bijelo polje Ribarevine 3.3 2017 2018 Not Req 2018 2018 2019 30,000 50 Route 6a Ribarevine Poda 14.5 2017 2018 Not Req 2018 2018 2019 30,000 50 Route 6a Poda Berane 12.8 2017 2018 Not Req 2018 2018 2019 30,000 50 Route 6a Berane Tunel Lokve 15.2 2017 2018 Not Req 2018 2018 2019 30,000 50 Route 6a Tunel Lokve Rožaje 14.7 2017 2018 Not Req 2018 2018 2019 30,000 50 Route 6a Obilaznica Rožaje II Phase 5 2017 2019 Not Req Not Req 2019 2020 18,000 30 Route 6a Rožaje Špiljani 20 2017 2019 Not Req Not Req 2019 2020 18,000 30

Page 83

Length Stage Stage Stage Stage Four Total Cost RSA Total Person Route Section Start Section End Design Stage (km) Start: End: One RSA Two RSA Three RSA RSA Inputs (Euros) Days RSAs Route 6b Kijevë Zahaq Detailed 30 2018 2021 Not Req 2018 2021 2021 30,000 50 Route 6b Pristina Peje Detailed 40 2015 Not Req TBC TBC TBC Route 6b Andrijevica Berane 16 2018 2019 Not Req 2018 2019 2020 50,000 50 Route 7 Merošina Prokuplje (Orljane) Detailed 11.503 2017 Nov-18 Not Req 2018 2018 2019 30,000 50 Route 7 Beloljin Rudare Detailed 24.171 2017 Nov-18 Not Req 2018 2018 2019 30,000 50 Route 7 Niš Pločnik Feas 40 2014 2018 2018 2018 TBC TBC 12,000 20 Route 7 Pločnik Merdare Feas 37 2014 2018 Not Req Not Req 2018 2019 18,000 30 Route 7 Bar Ulcinj 21.4 2018 2019 Not Req 2018 2019 2020 50,000 50 Route 7 Ribarevine Lepenac Detailed 14.6 2017 2018 Not Req 2018 2018 2018 18,000 30 Merdare/Serbia Sep, Oct, Route 7 Besi Concept 26.5 Not Req Not Req 2018 2019 18,000 30 Border 2017 2018 Route 7 Milot Mamurras 14 2017 2021 Not Req 2018 2021 2021 30,000 50 Route 7 Milot Rreshen 25.9 2017 2021 Not Req 2018 2021 2021 30,000 Route 7 Rreshen Kalimash 21.4 2018 2019 Not Req 2018 2019 2020 50,000 50 Route 7 Kalimash Morine 14.6 2017 2018 Not Req 2018* 2018 2018 18,000 30 Route 8 Ulcinj Krute 14.7 2017 2018 Not Req Not Req 2018 2019 18,000 30 Route 8 Mojkovac Kolašin Prelim 21.35 2018 2019 2018 2018 2019 2019 30,000 50 Route 9a Kolašin Crkvine Prelim 8.95 2018 2019 2018 2018 2019 2019 30,000 50 Route 9a Irig 2 Ruma 1 (motorway) Detailed 15.245 2017 Feb-18 Not Req 2018 2018 2018 30,000 50

Route 9a Novi Sad Ruma Prelim 112 2008 Not Req TBC TBC Not Req Route 9a Ruma Šabac Prelim 2008 Not Req Not Req TBC TBC

Route 9a Šabac Loznica Prelim 2008 Not Req TBC TBC Not Req Route 9a Laktaši Prnjavor Detailed 35.913 2018 2018 Not Req 2018 2018 2019 30,000 50 Route 10 Tuneli 7,8,9 0.67 2018 2019 Not Req 2018 2019 2020 50,000 50 Route 10 Stip Radovis 37 2015 Route 10 Miladinovci Stip Route 11 Obilaznica Golubovci Virpazar 16.5 2018 2019 Not Req 2018 2019 2020 50,000 50 Route 12 Virpazar Paštrovačka gora 14.8 2018 2020 Not Req 2018 2020 2021 30,000 50 Route 13 Paštrovačka gora Petrovac 7.5 2018 2020 Not Req 2018 2020 2021 30,000 50 SH3 Perrenjas Pogradec Detailed 27 2017 2021 Not Req 2018 2021 2021 30,000 50 SH3 Pogradec Korce 40 2017 2021 Not Req 2018 2021 2021 30,000 50 SH3 Korce Kapeshtice Detailed 32 2017 2021 Not Req 2018 2021 2021 30,000 50 SH3 Perrenjas Qafe Thane Detailed 13 2017 2021 Not Req 2018 2021 2021 30,000 50

Page 84

3.2.8 Yearly allocation of resources

Based on the tables in the previous section, the following table summarises the required financial resources on a yearly basis for implementing RSAs in each Regional Participant in Euros. Year 1 2018 Year 2 2019 Year 3 2020 Year 4 2021 Total ALB 300,000 - - 450,000 750,000 BiH 144,000 18,000 - - 162,000 MKD 27,000 9,000 - - 36,000 KOS 33,000 9,000 18,000 18,000 78,000 MNE 332,400 256,200 90,000 18,000 696,600 SRB 291,000 180,000 - - 471,000 Total 1,127,400 472,200 108,000 486,000 2,193,600

3.3 Pilot Road Safety Audits per WB6 Regional Participant

In this Activity, pilot RSAs have been carried out in each of the WB6 Regional Participants. A total of six RSAs - one in each Regional Participant. These were carried out in accordance with the SEETO RSA Manual (updated 2014) and Directive 2008/96/EC.

Requests were made in August 2017 to all Regional Participants to identify suitable projects for the pilot road safety audits. There have been subsequent changes to the original projects with Serbia and Albania offering an alternative project design, Bosnia and Herzegovina had originally proposed a project within the Republic of Srpska but issues with certification of the team to operate in the Republic of Srpska resulted in another project in the Federation being proposed. Kosovo originally proposed a preliminary design project funded by EBRD but problems with approval have resulted in another project being put forward. The projects identified to be audited are shown in the table below.

Projects in the design stage were considered preferable as early engagement could potentially deliver the largest road safety benefits. Regional Participant missions have been completed to deliver the road safety audits and provide the opportunity for capacity building and knowledge transfer for participants. However, it is important that the audit process is completed, and designers respond to audit recommendations. There is a risk that recommendations may lead to cost increases within the project but the potential crash savings (economic benefits of saving crashes) offered in most cases far outweigh these initial construction costs. The audit team, as part of their missions, ensured all the benefits of audit were outlined and the appropriate staff in Regional Participant received the latest training on the RSA process.

List of completed RSA Pilot Projects:

Participant Proposed Project Design Stage RSA Mission Date Report Submitted Albania SH4-O Gjirokastra By-Pass Detailed 4-8 December 2017 16 January 2018 Bosnia and M5 Jajce - Donji Vakuf TBC Preliminary 23-25 January 2018 17 March 2018 Herzegovina (Federation) The former Yugoslav Republic A1 Lenishka – Belovodica Detailed 9-12 October 2017 26 October 2017 of Macedonia Kosovo R6a Lumadh to Smrekonicë Preliminary 15-19 January 2018 12 February 2018 M-2 Kolašin Crkvine - Preliminary/Detail Montenegro 12-15 February 2018 26 March 2018 Mojkovac ed Serbia NIŠ-Pločnik Preliminary 28-30 November 2017 21 December 2017

Page 85

In preparation for the RSA Missions, the project team prepared the following guidance note to outline the roles and responsibilities, resources required for the audit mission and a proposed agenda for the visits.

The Road Safety Audit procedure is:  A formal process – similar to quality control  An independent process - the audit team should be independent of the project and not involved in the design (they can however be from the same organisation)  Should be carried out by someone with appropriate road safety and design experience and training.

The parties involved in a RSA:

Stakeholders in the Audit Process

Three parties are involved in the audit process: 1. The Client (usually the road authority, the Ministry relevant to Transport or private road operating company) 2. The designer (consultant or e.g. road authority) 3. The auditor.

The Client initiates the safety audits and commissions the auditor or audit team. All information and reports are distributed via the Client. Sometimes the Client may ask the designer to commission an auditor or audit team.

The Road Safety Audit Workflow

Page 86

3.3.1 Format for Missions to WB6 Regional Participants

Day and Timing Purpose of the Meeting Attendees Introductions of all Parties. Audit Team Audit Team Presentation on Project and Audit Process and Day One Road Authorities format for rest of mission AM approx. 3 Hours Designer Design Team to introduce the Project and hand over design Any other Stakeholders information to Audit Team in hard and soft copy formats Audit Team to study submitted design information (with Day One nominated attendees from road

PM authority if requested) to identify points of interest for site visit on Day Two Site Visit to road project (high visibility jackets to worn by all Audit Team Day Two parties). Road Authority All day if required Audit Team to identify, discuss, clarify and any road safety Design Team (not essential) problems with the design Any other interested parties Audit Team to present findings of the Audit and discuss with all parties Audit Team Day Three Road Authority PM approx. 2-3 Hours Opportunity to agree on Audit findings and recommendations Design Team and discuss practical implementation of any proposed Any other interested parties design changes

3.3.2 RSA Team Mission Reports

3.3.2.1 Albania

The RSA Mission to Albania took place from the 04-08 December 2017 in Tirana. The audit team consisted of:  Matt Chamberlain – Audit Team Leader  Darko Cvoric – Audit Team Member.

The team were also joined by Emiljano Zhuleku, the local appointed traffic engineer for Albania on the project.

The first day was spent meeting Mr Nikolin Berxhiku, representatives from ARA and the design team. The Audit Team explained the process and the format for the rest of the mission and what the aims and objectives of the mission were. The designer presented the project for the RSA the E853 Gjirokaster By-Pass and hard and soft copies of the design drawings were supplied to the Audit Team.

Proposed Alignment of Gjirokaster By-Pass.

Page 87

Due to extensive flooding in the area the Audit Team were not able to visit the site, so a desktop study of the drawings was undertaken on 5/6 December 2017.

The main issues resulting from the desktop Audit concerned:  Signing design and location  Crash barrier design.

3.3.2.1.1 Signing Design and Location

Example from western end of scheme, sign spacing too small and merge signs potentially masked by larger signs.

Example from Nyja Veriore Interchange – no chevron signing provided on compound radius.

Page 88

3.3.2.1.2 Crash Barrier Design

Example from Nyja Veriore interchange, hazardous potential gap in crash barrier at diverge point could be hazardous to an errant vehicle.

Example from Nyja Veriore interchange, no detail of connection between barrier types, the steel crash barrier is not connected to the bridge parapet. If steel barrier is struck by an errant vehicle, they will impact end of bridge parapet.

Example of a safe connection between different barrier types from the United Kingdom.

The audit team presented their findings to Ariana Hasani from ARA, Mr Nikolin Berxhiku from MoT and other representatives from ARA and the design team.

The completed Audit Report was delivered to ARA and the MoT on 16 January 2018 and at the time of writing, the audit team are awaiting the designer’s response.

3.3.2.2 Bosnia and Herzegovina

The RSA Mission to Bosnia and Herzegovina took place from 23-25 January 2018 in Sarajevo. The audit team consisted of:

Page 89

 Matt Chamberlain – Audit Team Leader  Lárus Ágústsson – Audit Team Member  Darko Cvoric – Audit Team Member.

The team were also joined by Amna Redžepagić-Čolić, the local appointed traffic engineer for Bosnia and Herzegovina on the project.

The first day was spent meeting representatives from PC Roads of Federation of Bosnia and Herzegovina (JP Ceste Federation Bosnia and Herzegovina) and representatives from the MoT. The Audit Team explained the process and the format for the rest of the mission and what the aims and objectives of the mission were. The design was presented for the RSA which is the section of the M5 Jajce - Donji Vakuf and hard and soft copies of the design drawings were supplied to the Audit Team.

Proposed section of M5

The audit team, accompanied by representatives from PC Roads of Federation Bosnia and Herzegovina (JP Ceste Federation Bosnia and Herzegovina) and visited the site on 24 January and identified any road safety problems.

The main issues resulting from the desktop Audit concerned:  Junction Design  Crash Barrier Design  Tunnel Design.

Page 90

3.3.2.2.1 Junction Design

Example from planned project – junction has no sheltered left turning lane, poor potential to and from the junction and tight radii unsuitable for large vehicles who will be required to use the proposed junction.

3.3.2.2.2 Crash Barrier Design

Example of existing crash barrier – unsafe terminal end and also too short and does not fully protect the hazard. The design should incorporate energy absorbent terminals.

An example of an energy absorbent end terminal from the United Kingdom.

Page 91

3.3.2.2.3 Tunnel Design

Existing tunnel just north of project section – tunnel portal forms a roadside hazard and lane widths very tight for two large vehicles to safely pass. New design should ensure adequate clearance for two large vehicles to pass.

Example from another new road project in Bosnia and Herzegovina – crash cushion fitted on tunnel headwall.

The audit team presented to representatives from PC Roads of Federation Bosnia and Herzegovina (JP Ceste Federation Bosnia and Herzegovina). All recommendations from the audit were verbally agreed.

The completed Audit Report was delivered to PC Roads of Federation of Bosnia and Herzegovina (JP Ceste Federation Bosnia and Herzegovina) on 17 March 2018 and at the time of writing, the audit team are awaiting the designer’s response.

3.3.2.3 The former Yugoslav Republic of Macedonia

The RSA Mission to the Former Yugoslav Republic of Macedonia took place from 9 to 12 October 2017 in Skopje. The audit team consisted of:  Matt Chamberlain – Audit Team Leader  Larus Agustsson – Audit Team Member.

Page 92

The team were also joined by Jani Hristoski, the local appointed traffic engineer for the Former Yugoslav Republic of Macedonia on the project.

The first day was spent meeting Darko Miceski and Darko Spiroski from the Ministry of Transport (MoT), Kiril Karkalesev from PESR and the Riste Ristov from the Designers, Prostor DOO Kumanovo. The Audit Team explained the process and the format for the rest of the mission and what the aims and objectives of the mission were. The designer introduced the project for RSA, A1 Lenishka – Belovodica and hard and soft copies of the design drawings were supplied to the Audit Team.

The Project consists of two parts:

Section One Leniska

Section Two Belovdica

The main issues resulting from the Audit concerned:  Crash barrier design  Junction layouts  Gore lengths and entry and exit radii on grade separated junctions  Bridge parapet design and interface with crash barrier.

Page 93

3.3.2.3.1 Crash Barrier Design

Examples of existing ramped and unsafe crash barrier terminals from the site visit. The barrier could be hazardous to errant vehicles and in many cases is too short and does not adequately protect the hazard.

An example of an energy absorbent end terminal from the United Kingdom

3.3.2.3.2 Junction Layouts

An example junction from the Audit within Section One. Two lanes continued through the junction with poor merge arrangement and no protection for left turning vehicles

Page 94

3.3.2.3.3 Gore Lengths and Radii

An extract from the proposed junction at Prilep. Potential conflict point between opposing lanes with no separation. Extra wide exit lane encouraging two vehicles to exit or late exit manoeuvres

3.3.2.3.4 Bridge Parapet Design

Existing Bridge in Section Two – poor design and condition of existing bridge parapet and no connection between barrier types, the steel crash barrier is not connected to the bridge parapet. If steel barrier is struck by an errant vehicle they will impact end of bridge parapet or potentially enter the river

Page 95

Example of a safe connection between different barrier types from the United Kingdom

The audit team presented their findings to the MoT, PESR and the designers on 12 October and all recommendations were verbally accepted.

The completed Audit Report was delivered to the MoT on 26 October 2017 and at the time of writing, the audit team are awaiting the designer’s response.

Page 96

3.3.2.4 Kosovo

The RSA Mission to Kosovo took place from 15-19 January 2018 in Pristina. The audit team consisted of:  Matt Chamberlain – Audit Team Leader  Lárus Ágústsson – Audit Team Member  Darko Cvoric – Audit Team Member.

The team were also joined by Emiljano Zhuleku, the local appointed traffic engineer for Kosovo on the project.

The first day was spent meeting Emir Morina, Gjynejt Mustafa and other representatives from the Infrastructure Directorate. The Audit Team explained the process and the format for the rest of the mission and what the aims and objectives of the mission were. The design was presented for the RSA which is the section of the R6a Lumadh to Smrekonicë and hard and soft copies of the design drawings were supplied to the Audit Team.

Proposed section of R6a

The audit team, accompanied by representatives from the Infrastructure Directorate and Kosovo Police, visited the site on 17 January and identified any road safety problems.

The main issues resulting from the Audit concerned:  Crash barrier design  Junction Design  Accesses  Roundabout Design.

Page 97

3.3.2.4.1 Crash Barrier Design

Example from existing road in a settlement with frequent gaps in the crash barrier and hazardous end terminals. Crash barrier should not be used in urban areas and this should be avoided in the new design.

Example from existing road – bridge parapet and barrier not connected. The new design should have a secure connection between barrier types.

Example of a safe connection between different barrier types from the United Kingdom.

Page 98

3.3.2.4.2 Junction Design

Example of planned junction from the project – very short acceleration and deceleration lanes.

Example of planned junction from the project – no opportunity for vehicles to turn left from the south or right from the north – an at grade roundabout would be more appropriate.

Example of planned junction from the project – Y-type interchanges are unsafe and should be avoided in the new design. The short section to by-pass roundabout should be removed and all vehicles will use the roundabout.

Page 99

3.3.2.4.3 Accesses

Example from project – many accesses directly onto the highway with risk of slowing vehicles and vehicles trying to access onto the mainline. The cross section should be adjusted with extra wide shoulder to act as sheltered lane for drivers entering/exiting accesses. Where many accesses exist, the speed limit should be reduced.

3.3.2.4.4 Roundabout Design

Example roundabout from project – single lane roundabout entries do not require two circulating lanes. This can lead to side swipe crashes. The deflection north to south is also very poor and no deflection splitter islands provided or signing.

Another example roundabout from project – good roundabout design with good deflection, single circulatory lane, splitter islands and correct signing.

The audit team presented their findings to Mr Emir Morina, Mr Gjynejt Mustafa and other representatives from the Infrastructure Directorate. All recommendations from the audit were verbally agreed.

The completed Audit Report was delivered to the Infrastructure Directorate on 12 February 2018 and at the time of writing, the audit team are awaiting the designer’s response.

Page 100

3.3.2.5 Montenegro

The RSA Mission to Montenegro took place from the 12-15 February 2018 in Podgorica. The audit team consisted of:  Lárus Ágústsson – Audit Team Member  Darko Cvoric – Audit Team Member.

The team were also joined by Dusan Savkovic, the local appointed traffic engineer for Montenegro on the project.

The first day was spent meeting representatives from MoT and the Design Team. The Audit Team explained the process and the format for the rest of the mission and what the aims and objectives of the mission were. The designs were presented for three sections of the M-2 Kolašin Crkvine - Mojkovac and hard and soft copies of the design drawings were supplied to the Audit Team.

The three sections audited were:

 Section One - Crkvine to Kolasin – carried out as a Stage One RSA  Section Two - Kolasin to Mojkovac – carried out as a Stage One RSA  Section Three - Lepenac – Ribarevine – carried out as a Stage Two RSA.

Plan of Section One Plan of Section Two Plan of Section Three

The audit team, accompanied by representatives from MoT and Designers visited the site on 13 February and identified any road safety problems.

The main issues resulting from the Audits concerned:  Road Alignment  Crash Barrier Design  Junction Design.

Page 101

3.3.2.5.1 Road Alignment

Example from Section Two of planned project – existing bends are poorly signed at sections where it is proposed to raise the speed limit to 60 kph. Chevron signs are poorly located and not conspicuous.

Example of good design for chevron signs and good conspicuous signs with colour contrast.

3.3.2.5.2 Crash Barrier Design

Example from Section Two – existing crash barrier is too short and with an unsafe terminal end.

Page 102

Example from Section One – gap in crash barrier between steel barrier and bridge parapet.

Existing underpass in Section Two – underpass headwall forms a roadside hazard and unsafe barrier terminal is also unsafe.

An example of an energy absorbent end terminal from the United Kingdom.

Page 103

Example from another new road project in the Region – crash cushion fitted on tunnel headwall.

3.3.2.5.3 Junction Design

Example from Section Three – very relaxed junction radii encourage high entry and exit speeds.

The audit team presented the findings from all three sections to representatives from MoT and the Design Team. All recommendations from the audit were verbally agreed.

The completed Audit Report was delivered to MoT on 26 March 2018 and at the time of writing, the audit team are awaiting the designer’s response.

Page 104

3.3.2.6 Serbia

The RSA Mission to Serbia took place from 28-30 November 2017 in Belgrade. The audit team consisted of:  Matt Chamberlain – Audit Team Leader  Larus Agustsson – Audit Team Member  Darko Cvoric – Audit Team Member.

The team were also joined by Dusan Savkovic, the local appointed traffic engineer for Serbia on the project.

The first day was spent meeting representatives from the Ministry of Construction, Transport and Infrastructure, Public Roads Serbia and CeS.TRA, the designers on the project. The Audit Team explained the process and the format for the rest of the mission and what the aims and objectives of the mission were. The designer presented the project for the RSA, E80 section Niš-Pločnik and hard and soft copies of the design drawings were supplied to the Audit Team.

Proposed Alignment of E80 Niš-Pločnik

The audit team, accompanied by representatives from Public Roads Serbia and CeS.TRA visited the site on 29 November and identified any road safety problems.

The main issues resulting from the Audit concerned:  Interchange Geometry  Roundabout Location.

Page 105

3.3.2.6.1 Interchange Geometry

Example from Interchange at Prokuplje East, entry radius is very acute and directs vehicle onto mainline.

1.1.1.1 Roundabout Location

Example from Roundabout at Prokuplje East, roundabout will be located on a crest. Drivers will have very limited forward visibility to new roundabout.

View looking west – new roundabout will be located just over crest.

Page 106

The audit team presented their findings to Ministry of Construction, Transport and Infrastructure, Public Roads Serbia and CeS.TRA on the 30th November and all recommendations were verbally accepted.

The completed Audit Report was delivered to Public Roads Serbia on 21 December 2017 and at the time of writing, the audit team are awaiting the designer’s response.

3.4 Conclusions

As many roads are being upgraded and constructed on the SEETO core and comprehensive network, among other reasons to ensure improved road safety, it is important to include road safety audits in the design and construction process. In addition, as all roads are part of the SEETO core and comprehensive network and if they become TEN-T, then RSA is mandatory according to EU Directive 2008/96/EC. RSA should be done on both upgrading projects and new projects.

The audits carried out at preliminary design phase also demonstrated that potentially unsafe design options can be modified at an early stage and will deliver the largest potential safety benefits. Therefore, the audit team were able to show participants that early engagement of auditors in the design process is the most effective.

The audit plans produced as part of this assignment will also will be a useful tool for participants to give some guidance on timing and appropriate costs for the different stages.

Page 107

4 Component 3 - Road Map for establishing system for continuous road crash data collection

This component supports the regional participants in implementing a harmonised crash data collection and analysis system. The activity according to ToR simply states the development of a ‘Road Map’ for establishing a national system for continuous road crash data collection. To achieve this task, missions were undertaken to each of the WB6 Regional Participants to assess current road crash data collection and analysis capabilities. The three activities contained within this component are:

Activity 1 Assess current road crash data collection-analysis systems.

Activity 2 Set up a concept for a common system in WB6 based on EU practice.

Activity 3 Prepare road map for establishing national system for continuous road crash data collection and analysis.

The purpose is to provide SEETO with the findings of the review and recommendations for Component 3 of the TA, Road Map for establishing system for continuous road crash date collection.

There is a need to collect more statistical data for the most serious injury and fatal collisions. This is based on the requirement to clearly understand the pre-impact approach paths, and actions of the involved. Understanding is essential in developing a data led approach to reducing such incidents.

The collection of all this additional data is not seen as the sole requirement of the investigating police officer but more of a shared obligation amongst the responsible Ministries; the detailed engineering data could be provided by Ministry of Transport for example.

To achieve this shared approach will require:  Training for both police and road safety engineers in crash scene analysis  A more advanced database capable of sharing ‘non-sensitive’ data amongst the Ministries  A more advanced database capable of providing the advanced analytical programmes to undertake the task. A description of such a database is provided later in this report.

The ability to provide this enhanced data will enable detailed analysis to be undertaken to fully understand why the collisions are occurring, this will lead to a more scientific data led approach to developing actions to reduce their reoccurrence.

4.1 Background and Terminology of Crash Data Base Systems

The phrase “Crash Database Systems” covers all the elements which constitute the methods and arrangements to collect, store and analyse any systematic report or information collected on road collisions and those injured in them (WHO 2010). This definition therefore includes the stakeholders, which are any persons involved with the system in any capacity. Generally, when Crash Data Systems are considered the focus tends to be on the IT systems primarily (associated computer hardware and software). It is however important to remember that all the elements are important.

The main elements and components of the crash data system are set out here:  Data collection fundamentals o Reporting Form . Comprehensiveness . Quality . Ease of filling o Reporting levels of incidents

Page 108

o Quality of data collection . Training/skill/commitment levels of staff . Resources available . Collection of location coordinates and/or location description o Links to other data sources . To improve data quality . To enable more advanced analyses  Data Capture factors o Paper based collection o Manual data entry o Electronic collection direct into database (on mobile devices) . Quality checks on data (Validation)  Data sharing/dissemination o Availability of data to all stakeholders . Filtering of sensitive information as appropriate . Direct access via internet link . Indirect access through manual import/export to other local systems . Regular summary reports available through various media  Analysis for management functions o Summary reports o Crash information management o Safe Systems Analysis . Strategy development . Progress against casualty or other reduction targets . Monitoring of Key Performance Indicators (KPIs) o Road safety analysis . Identification of trends . Identification of blackspots/spatial analysis . Economic appraisal . Management of site treatments or enforcement efforts . Analysis of problem locations . Evaluation and statistical analysis functions.

4.1.1 Data Collection

The initial data quality is a very important factor in maintaining the credibility of the Database System. The accuracy and consistency of the information recorded in the system is the fundamental and biggest issue in maintaining a credible analysis capability.

Clearly, inconsistencies in the data can lead to misleading results when performing analyses.

Data quality has three main properties:  Quality and comprehensiveness of the form  Quality and consistency of the data reporting on incidents  Levels of reporting or under-reporting.

4.1.2 Form Content and Quality

It is important to understand this report only deals with the form designed to collect statistical data associated with the collision. It is not intended to address the form used by the police for the legal recording and presentation of evidence in a court of law.

Within the WB6 Region the review identified some Regional Participants have separated the statistical form from the official report form while others have a combined form. The review identified where the report is combined the form is always presented as evidence in a court of law thus the statistical data

Page 109

is used as evidence which is not what it was ever intended to be. An example being where the statistical data shows a road surface defect it has resulted in the roads authority being implicated in causation.

The statistical reporting form needs to be comprehensive enough to collect the range of information required by road safety practitioners. Collecting the data is relatively difficult and can be time consuming, so it is important that the process is made as straightforward as possible.

Having a form that is shorter with predominately multiple-choice questions rather than requesting open information (e.g. written statements) will improve data quality and also encourage reporting. Logical and intuitive layout of paper-based data collection forms or the interfaces of mobile electronic data collection systems will also promote data quality and help to reduce under-reporting.

Generally, there should be an officially recognised working group to oversee and control the development of the statistical reporting form. The organisation should coordinate aspects such as the layout/design, data collection practices and field changes and version control. This body should include all the main end user stakeholders in addition to the key data collection stakeholder and should ensure that all the users’ needs are taken into account. It is important that any changes to the form fields are carefully considered, are minimal and can be harmonized with previously archived data as far as is possible; it is not advisable to make constant small changes to the data that is collected.

Having developed a form, the next most important element is to ensure all the police officers tasked with entering the data fully understand the meaning of every field and the options available. The need for a comprehensive training program is fundamental to guaranteeing the quality of the data within the database; failure to train the police officer correctly will invalidate all the data contained within the system.

4.1.3 International Recommendations

An EU project initiative has sought to define the basic set of parameters that should be collected on a crash and casualty reporting form (see CADaS version 3.6 20177). The CADaS report states that “the variables and values of CADaS may be considered as recommendations for national police road crash data collection reports”. The full CADaS list of data elements is considerably longer and more complex than the minimum set and is aimed at capturing the information that academics might like to be available for high level analysis and research purposes.

The Model Minimum Uniform Crash Criteria (2012)8 (MMUAC) is a US Government initiative that has identified the basic set of fields which should be included in a crash report form which will provide the information required for safety management purposes. MMUAC’s key aim is to promote greater uniformity of the data on crashes which is collected in the different US States. The current recommended list gives a total of 110 elements which it recommends should be captured; 77 of these being collected by police directly with a further 10 derived from the general scene information fields. The final 23 elements should be obtained from data linkages to external official data sources such as asset databases; these could be collected manually if data linkages are not in place.

Whilst the full CADaS list sets out an ideal and extensive set of questions, many countries are decreasing their crash database collection efforts by having shortened and briefer forms. The relative weakness of the data collected is compensated by linking the data to other sources such as hospital systems in Sweden or by conducting surveys to estimate underreporting rates in the Netherlands. Researchers and practitioners in these countries would prefer that police collected better primary data.

7 https://ec.europa.eu/transport/road_safety/sites/roadsafety/files/cadas_glossary_v_3_6.pdff

8 https://www.transportation.gov/sites/dot.gov/files/docs/MMUCC_4th_Ed.pdf

Page 110

4.1.4 Data Capture

Data capture is the general term for the processes to actually gather the information. For crash data systems it generally refers to the methods to get the information collected by the initial reporter into a computer database.

4.1.5 Data Collection Processes

The process has frequently taken the form of the reporting police officer taking notes at the crash scene or shortly after, then a structured form is completed from the notes when back in the office. The instruction to the reporting police officers is often that the pro forma should be completed directly at the crash location. It is however generally accepted that the often chaotic and dangerous nature of the scene immediately after the incident and the low priority for filling in the crash report form means that this seldom happens in practice. The quality of the collected information is likely to improve if the officer uses the specific form rather than doing this later from notes. The completed forms should be checked by a more experienced officer for accuracy and completeness before they are signed-off and filed.

In many instances the form, or a photocopy of it, is passed to support staff for entry into the computer database system. At this point some automated validation of the answers should occur to identify logical errors (where one answer is incompatible with another field’s recorded value). Automated checks for blank fields should also be programmed into the system.

The major problems with paper-based approaches are that:  Data entry is a significant waste of resources when there are more efficient electronic methods available  Validation is often performed by a person who did not attend the incident: o They may incorrectly interpret the information; not communicating with the original collector. o They can simply select any option.

There are variations on this model, in some locations (e.g. South Africa), crash report forms are read by optical scanners to populate the database automatically. In the UK some police authorities speak via a phone link to an operator who takes the officer through the questions and enters the answers into the database.

Current best practice makes use of latest advances in communications and mobile technology. These systems can enable the reporting officer to enter the data directly into a mobile electronic device (e.g. a Tablet/mobile computer/larger mobile phone). The device then synchronises with the central database either at the time the information is entered via a mobile data link or later when a Wi-Fi or physical link can be established.

This kind of system has the advantages that:  It removes the need for data entry staff to perform the additional task of inputting the information from paper forms o both time consuming and a potential source of additional errors in the data  It forces the person who attended the incident to validate the answers o saving them time and irritation later  It speeds up the process so that intelligence can be available in nearly real time  Bundled GPS equipment can be used to get accurate locations very easily  Bundled cameras can be used to obtain photographic records of the scene and vehicle damage o These are associated with the crash record and are stored in the central database.

Ideally mobile devices can save time and effort by accessing a variety of databases to populate data automatically. For example, inputting the national ID number for a person involved in a crash can potentially link to the national database and populate the age, gender, nationality and so on; inputting

Page 111

the vehicle registration can access vehicle make, age, colour etc. This should save significant effort and ensure the information is correct.

The Road Accident Data in the Enlarged European Union, Learning from Each Other. Brussels 20069 stated:

“Electronic systems have great potential to reduce the burden on the Police and speed up the data collection process. The equipment and software are expensive, however, and data could easily be lost if police officers do not use the new systems correctly. Also, there is a depressing history of large-scale IT systems that have failed to live up to expectations. It is necessary to see the results of major trials before deciding whether to recommend the adoption of these systems. Care should be taken to ensure that existing data quality checks are not discontinued if electronic data collection systems are adopted.”

4.1.6 Under Reporting

Under-reporting is a significant problem in most countries. In High Income countries it is widely considered that reporting rates of fatal crashes are generally close to 100% but this is often far from the case for many Low and Middle-Income countries.

High under-reporting rates undermine the confidence that can be put in the results of analysis and also the value of the dataset as the basis for any safety evaluation and also monitoring casualty reduction. What is important is that the sample is consistently representative of the real situation; the basic data collection process is consistent over time.

It is known that some crash types are more consistently under-reported; these being single motorcycle and bicycle crashes, in some countries pedestrian crashes are also significantly underrepresented in national databases.

It is important to have good estimates of under-reporting rates since this issue can hide the true costs of road injuries and crashes nationally. Some account of under reporting can be incorporated into economic evaluation since crash costs used in appraisal can be inflated to take account of the missing crashes.

A solution to poor reporting levels has been to link police crash and casualty data to hospital records in an attempt to identify road injuries and incidents reported at medical facilities but not to the police to gain estimates of the overall national problem.

In many countries in the world, there is a clear legal requirement to report all car crashes; from damage to fatal incidents to the police. In these countries, damage sustained by a vehicle cannot generally be officially repaired at garages without proof that the crash has been registered with the police. Whilst this seems to lead to high reporting rates of crashes, it is possible that not all casualties in a crash always get reported which is also a possible source of under-reporting.

4.1.7 Data Linking

International Traffic Safety Data and Analysis Group (IRTAD) has recently published a document which aims to summarise how data (targeted particularly for those incidents where casualties are seriously injured but not killed) should be linked to other data sources (Reporting on Serious Road Traffic Casualties, IRTAD 2012)10. It sets out all the issues very clearly and comprehensively.

The main issues and conclusions in summary are:

9 http://ec.europa.eu/transport/roadsafety_library/publications/sau_guide_best_practices_brochure.pdf

10http://www.itf-oecd.org/sites/default/files/docs/road-casualties-web.pdf

Page 112

 Knowing the true numbers of killed and also those seriously injured is important for: o Building investment cases and estimating economic losses o Making international comparisons o Targeting measures appropriately  Police do not always collect data on less severe crashes, but generally report most if not almost all incidents involving fatalities.  Medical data sources can be used to gain an understanding of under reporting levels and the true numbers injured  Police derived crash report data must be the basic source for analysis to support safety through engineering, enforcement and any other approach that requires spatial investigation and also for statistical purposes  Medical data has some strengths with greater detail about the individual casualty, but poor location and general details are usually collected, if it at all  Medical data collection can be highly variable nationally and even within a single hospital  Severe and slight injury classifications are generally very poorly defined internationally  Police are not sufficiently skilled (or generally receive training) to accurately assign severity  Established medical scales should be used to assign severity but this requires significant skill and effort (e.g. ICD, AIS, MAIS, ISS systems) o The definition of MAIS11 3+ has been broadly accepted as the best definition internationally for serious injury  Linking records to different data sets is best achieved through shared unique keys/fields rather than through matching a range of broader fields such as incident date/time etc.

The linking of crash data to other sources of information can have major benefits, in line with “big data” approaches; this can permit new insights into crash causation and factors that affect safety. It is relatively easy to link data sets where there is a well-defined and common unique key which is present in all databases; however, it is seldom the case that datasets which were not originally designed with linking in mind have good common anonymised keys. Linking data by other means such as statistical matching of multiple fields can be difficult with a relatively low success rate.

Crash Data will benefit significantly by linking to medical data to get improved accuracy of the severity assessments of those injured. This linking can also help to estimate under reporting rates from the proportions of casualties in hospital not known to police and vice versa.

4.1.8 Geographical Information Systems (GIS) Data

Plotting crashes on maps and linking incidents to spatial information held in other GIS layers is also technically “data linking”. This relies on linking the crash to spatial elements through the location where the crash occurred as defined by its map coordinate.

11 http://www.surgicalcriticalcare.net/Resources/injury_severity_scoring.pdf

Page 113

How well crashes can be linked to GIS data depends on how accurate the crash locations have been assigned but applications are also limited by how sophisticated the digital maps are and how intelligent the elements they contain are. Some digital GIS maps are simply files which contain definitions for the start and end points of lines which form roads etc. and are in all practicality backdrops only. In contrast in true GIS data sets the road line has the road name/number associated with it and possibly even traffic count data for the stretch.

Current best practise is to use GIS data as widely as possible and making this more possible there is a corresponding recent increase in the number of GIS based datasets that are available. There have also been significantly improving performance of GIS software served over the internet. There are also increasing numbers of useable quality “free” map sets available on the internet such as Open Street and Google Maps.

If the police details recorded for crashes are of poor quality, GIS data sets of road inventory or assets can be used to check that details are correct (number of lanes/speed limit etc.) or to add additional information about the location.

GIS layers can be used to check the location information for a crash in a coarse way, if the form claims the crash occurred in a certain Regional Participant yet the crash plots outside the boundaries of a layer defining that area then the problem could be automatically flagged for manual checking.

Crash patterns and occurrence can also be analysed in relation to GIS spatial objects that might affect safety levels. It may be possible to very quickly derive the crash occurrence rates on different road standards for example if these specific characteristics are available associated with road sections in the data sets. Another example is that crash patterns near schools could be reviewed very quickly if a layer with education establishments identified exists and if a “buffer” to select crashes in the vicinity around such objects.

4.1.9 Additional Data for Road Safety Analyses

In addition to crash data reported by police, collection of a range of information which influences road safety can be highly beneficial to those working to reduce road casualties. The management of additional data can be very advantageous since if this is stored in a central database it can also be accessed easily by the safety stakeholders and linked much more readily with the crash report information.

The new Safe Systems methods advocated by international road safety stakeholders takes a view that in addition to the “primary” safety indicators such as the numbers killed and seriously injured additional data that is known to be highly correlated (ideally causal) with safety risk and levels should be collected and used to complement the collection and analysis of reported crashes and casualties.

4.1.10 In Depth Crash Data

The crash reporting system collects fairly general information about a large number of the collisions occurring. This system should be complemented by in depth data collection programmes. These allow more detailed analysis of a sample of the more serious crashes to be undertaken that go beyond the scope of police data. Research programmes of this nature have been established successfully throughout the world. These programmes require considerable resource and expertise but once established they offer the opportunity to enhance the skills of road safety practitioners to a great extent by developing their understanding of crash causation factors. These studies are generally targeted at the more serious incidents and can be used to check the accuracy of the police reporting system. This approach also provides valuable information on issues such as general vehicle design and safety standards that are not adequately addressed by the police reporting processes.

Page 114

The current datasets being used within the WB6 Region in the main only provide a simplistic overview, there is therefore a requirement to separate each entity involved and provide a detailed description of its activities pre-impact, at impact and post impact.

For crashes involving vehicles it should be enhanced to include:  Pre-event direction of travel  Number of lanes in direction of travel; of the vehicle  Number of turn lanes in direction of travel of vehicle  Pre-crash vehicle position  Pre-crash movement  Critical pre-crash event category  Critical pre-crash event  Driver manoeuvre to avoid impact  Attempted avoidance manoeuvre  Pre-impact vehicle stability  Pre-impact location  Driver vision obscured by  Driver distracted by  Most harmful event  Most harmful object struck  Rollover initiation type  Rollover initiation location  Surface type roll initiated  Surface condition roll initiated  Pre-rollover event  Pre-rollover vehicle manoeuvre  Safety barrier impact event o Barrier impact type o Barrier impact secondary event o Object struck secondary event o Impact angle o Kerb in front or inline o Barrier type o Barrier passive safety standards o Barrier containment level  Narrow object impact event o Narrow object type o Narrow object passive safety standard o Narrow object secondary event o Narrow object test level  Detailed crash event sequence o Crash event o Object struck o Post event vehicle manoeuvre o Post event vehicle stability For passenger casualties it should be enhanced to include:  Student on journey to or from school  Detailed injury description  Location prior to the impact  Action prior to the impact  Safety features present and used  Ejection and path  Airbag deployment  Entrapment  Injury diagram For pedestrian casualties it should be enhanced to include:  Student traveling to or from school

Page 115

 Using mobile phone  Injury location  Casualty type to include: o Cyclist o Wheelchair o Rideable toys (skateboards etc.) o Baby carriage / stroller o Unknown type  Location at time of crash  Direction of travel  Action at time of crash  Contributing circumstances  Safety equipment  Condition / impairment at time of crash  Suspected alcohol / drugs. 4.2 Findings on crash databases systems from previous studies

The SAFEGE study12 provided a preliminary assessment of the crash database systems within the Regional Participants. The study used the CADaS basic datasets to assess the current status of the crash data collection. The main findings and conclusions are summarised in the following sections.

4.2.1 Albania

Crash statistical data in Albania is only collected and recorded by the Police using a paper-based system which is encoded onto a standalone database. The review undertaken in 2016 identified the following areas:

Accident Scene related data This contains thirteen fields of which two were not recorded; these being the European classifications associated with regional and administrative areas.

Road related data This contains twenty-five fields of which thirteen were not recorded; these being associated with road classification, junction controls, carriageway types and road markings.

Traffic Unit information This contains eighteen fields of which thirteen were not recorded; these being associated with vehicle types, makes and models, pre-impact manoeuvre, first point of impact, objects hit and hit and run.

Person related data This contains twenty-one fields of which seven were not recorded; these being associated with alcohol impairment results, drug impairment testing, driver distraction, medical / physical impairment, journey purpose and MAIS injury scale.

4.2.2 Bosnia and Herzegovina

Crash statistical data in the Federation of Bosnia and Herzegovina (Federation Bosnia and Herzegovina) is only collected and recorded by the Police using a paper-based system. Crash data in the Republic Srpska (Bosnia and Herzegovina) is only collected by the police using a paper-based form which is subsequently encoded onto a computer program.

12 Monitoring of the Road Safety Strategies in SEETO Members and Draft a Regional Short-term Action Plan – September 2015 - SEETO

Page 116

The review undertaken in 2016 identified the following areas:

Accident Scene related data This contains thirteen fields all were either completely or partially recorded

Road related data This contains twenty-five fields of which seventeen were not recorded; these being associated with GPS locations, road classification, road speed limits, junction controls, carriageway types, road markings and work zones.

Traffic unit information This contains eighteen fields of which ten were not recorded; these being associated vehicle make and model, first point of impact and objects hit.

Person related data This contains twenty-one fields of which eight were not recorded; these being associated with alcohol and drug testing, driver distraction, journey purpose and MAIS injury scale.

4.2.3 The former Yugoslav Republic of Macedonia

Crash statistical data in the former Yugoslav Republic of Macedonia is only collected and recorded by the Police using a paper-based system which is subsequently encoded onto a computer program.

The review undertaken in 2016 identified the following areas:

Accident Scene related data This contains thirteen fields of which all were either completely or partially recorded.

Road related data This contains twenty-five fields of which thirteen were not recorded; these being associated with road classification, junction controls, carriageway types and road markings.

Traffic unit information This contains eighteen fields of which thirteen were not recorded; these being associated with vehicle types, makes and models, pre-impact manoeuvre, first point of impact, objects hit and hit and run.

4.2.4 Kosovo

Crash statistical data in Kosovo is only collected and recorded by the Police using a paper-based system which is subsequently encoded onto a computer program.

The review undertaken in 2016 identified the following areas:

Accident Scene related data This contains thirteen fields of which all were either completely or partially recorded.

Road related data This contains twenty-five fields of which twelve were not recorded; these being associated with GPS coordinates, road classification, carriageway types and road markings.

Page 117

Traffic unit information This contains eighteen fields of which six were not recorded; these being associated with vehicle specifications and vehicle insurance.

Person related data This contains twenty-one fields of which eight were not recorded; these being associated with drug impairment testing, safety equipment, seating position, driver distraction, physical impairment, journey purpose and MAIS injury scale.

4.2.5 Montenegro

Crash statistical data in Montenegro is only collected and recorded by the Police using a paper-based system which is subsequently encoded onto a computer program.

The review undertaken in 2016 identified the following areas:

Accident Scene related data This contains thirteen fields of which all were either completely or partially recorded.

Road related data This contains twenty-five fields of which seventeen were not recorded; these being associated with GPS coordinates, road classification, speed limits, junction types, junction controls, carriageway types, road markings and work zones.

Traffic unit information This contains eighteen fields of which ten were not recorded; these being associated with vehicle types, makes and models, pre-impact manoeuvre, first point of impact, objects hit and hit and run.

Person related data This contains twenty-one fields of which ten were not recorded; these being associated with alcohol impairment results, drug impairment testing, seating position, driver distraction, journey purpose and MAIS injury scale.

4.2.6 Serbia

Crash statistical data in Serbia is only collected and recorded by the Police using a paper-based system which is subsequently encoded onto a computer program.

The review undertaken in 2016 identified the following areas:

Accident Scene related data This contains thirteen fields of which all were either completely or partially recorded.

Road related data This contains twenty-five fields of which only one was not recorded; this being associated with E road kilometre location of the crash.

Traffic unit information This contains eighteen fields of which only one was not recorded; this being associated with vehicle insurance.

Person related data

This contains twenty-one fields of which only one was not recorded; this being associated with MAIS injury scale.

Page 118

4.3 Current road crash data collection-analysis systems

The current road crash data collection system utilised in the WB6 Region is predominantly achieved using a paper-based form. The forms used are primarily created to gather the data associated with the police requirements for recording the crash. This report only concentrates on the statistical data associated with describing the elements involved in the incident.

The following section maps the agencies responsible for investigating road crashes from both a Police and Engineering perspective. The data and information provided by each of the Regional Participants are reviewed in line with international best practice and more specifically with compatibility with CADaS principles and methodologies. A major component of the review is identification of what processes are in place to ensure the quality of the data recorded is to the highest level and accurate recoding of the crash location.

The review identifies the data storage and exporting capabilities of each authority and Regional Participant to identify the potential for developing a national data linking capability within the WB6 Regional Participants based on EU practices. This analysis forms the bases of recommendations outlining the actions that is required to enable the exchange of data between systems.

4.3.1 Albania

The police in Albania have their own ‘Accident Information System’ (AIS), this is a standalone, Microsoft Access database that was created following a SweRoad project in 2005. The system incorporates some thirty data entry fields which can be used within a simple cross tab analysis to generate numerical table output. The program lacks any integral GIS capability. The statistical report form used by the police was created at the time of the SweRoad project and specifically designed for the AIS database.

The police disseminate, on a monthly basis, crash data to the Ministry of Infrastructure and Energy (MIE). The Ministry, on a yearly basis, undertake a blackspot analysis assessment using the crash data provided by the police. The current assessment dated 2017 identifies 118 blackspot locations on the National Road Network.

The requirement for a fully functioning advanced analytical and GIS capable database will be essential to achieve the proposals outlined in this report as the current database is very limited in its capability to expand. The database also lacks a data linking capability appropriate to the objectives of this report.

The Police also provide the Albanian Institute of Statistics13 (INSTAT) with crash data which they use in their online interactive database. The current Road crash statistics on the INSTAT web site indicates in 2016 there were 269 fatalities and 2,509 injured people from 2,033 accidents. The site did not provide the number of damage only collisions.

The crash data causation fields produced by INSTAT are:  Careless driving  Sudden stoppage  Noncompliance with traffic sign  Driving on the wrong side  Exceeding speed limit  Unexpected change of direction  Fail to give priority  Careless overtaking  Alcohol/drugs  Other.

13 http://www.instat.gov.al/en/

Page 119

The current crash report form has been developed beyond the basic CADaS datasets and aligns with the existing database. Further development, to align with the advanced CADaS datasets, is not possible within the capabilities of the current database. The institutions within Albania are aware of the problem and are in the process of developing the procurement of a new system capable of aligning with the advanced CADaS principles.

4.3.2 Bosnia and Herzegovina

The Police in Bosnia and Herzegovina are responsible for attending and investigating road crashes. They all use a paper-based form to record their findings. They however have a fragmented approach to statistical data collection:

 The police in FIB Republic of Srpska produce a statistical form which they encode into their database.  The police in Federation of Bosnia and Herzegovina are disseminated into each Canton which have created their own statistical data collection form. The paper forms are submitted to Federation of Bosnia and Herzegovina MOI who collate the information without the use of a computer database.

Given the variation in the current status of crash data collection within Bosnia and Herzegovina it is not possible to make a simple and meaningful assessment of their current situation beyond what was assessed in the SAFEGE study.

The initial requirement will be to agree on a standardised set of statistical data elements that must be collected. A fully functioning advanced analytical and GIS capable database will then be required to achieve the proposals outlined in this report.

The current crash data fields being recorded by FIBH MOI align with the basic CADaS datasets. They have developed an updated statistical crash report form that enhances their data collection closer to the advanced CADaS datasets (copy of the report is included in Appendix A).

4.3.3 The former Yugoslav Republic of Macedonia

The Police in the former Yugoslav Republic of Macedonia are responsible for attending and investigating road crashes. They use a paper-based form to record their findings. There are currently two forms; one for the crash report and the second for the Statistical data. It has been identified that the statistical data form is not being completed so it has been proposed the forms will be combined into a single report. The report is still in development but reportedly it complies with the CADaS dataset. A new crash database is also being developed to replace the current, very old system.

Currently the paper forms are quality checked prior to being submitted to the statistical department at the MOI. This department disseminates the statistical crash data using excel spreadsheets every three months which become the official crash data records for the Former Yugoslav Republic of Macedonia.

Due to the fact the revised report form had not official received authorisation at the time of the review the document was not available for review.

The requirement for a fully functioning advanced analytical and GIS capable database will be essential to achieve the proposals outlined in this report as the current database is very limited in its capability to be expanded. The database also lacks a data linking capability appropriate to the objectives of this report.

The current crash data fields being recorded are in line with the basic CADaS datasets. Institutions are aware of the need to enhance the collection process and reportedly have achieved this in line with the advanced CADaS datasets.

Page 120

4.3.4 Kosovo

The Police in Kosovo are responsible for attending and investigating road crashes. They use a paper- based form to record their findings. The data collected is encoded into the Kosovo Police Information System (KPIS). The initial report of the incident is required to be encoded into the database within 24 hours. The final encoding occurring when the investigation and report is completed.

The crash report form has a requirement for the GPS coordinates of the crash scene. However, the police state they are not currently collecting this as a matter of course. They indicated they were undertaking trails at the time of this review to identify the best method of achieving this requirement.

The current database is reportedly being upgraded and will have the capability to upload photographs and scanned documents. The proposals for the upgrade of the system will include a GIS and analytical capability.

The police have developed a mobile application to allow the public to report being involved in an accident. The application provides the location of the mobile phone at the time of the call and potentially the location of the incident.

The requirement for a fully functioning analytical and GIS capable database will be essential to achieve the proposals outlined in this report. In the short term the current database has the potential to be enhanced to achieve the primary goal. However, in the long term achieving the level of data linking and advanced analytical capability will require the procurement of a bespoke system.

The current crash data fields being recorded in the Kosovo Police Information System (KPIS) exceeds the basic CADaS requirements but has not, as yet, reached the full advanced dataset requirement. The limitations of the database would seem to be a factor in the degree of enhancement possible.

4.3.5 Montenegro

The Police in Montenegro are responsible for attending and investigating road crashes. They use a paper-based form to record their findings. The content of the forms is quality checked by a senior officer prior to it being entered into the police database. The crash database is part of the main police system which is developed using an SQL format.

Within each police station is an occurrence register (diary) in which all incidents reported to the police station are recorded. Where an officer attends a report of a road collision there is a brief description of the incident together with the date, time, location, involved parties and severity.

The Initial report as provided by the occurrence register is uploaded onto the police database within 24 hours of the incident being reported. The database, at this time, generates a unique Accident ID Number for the record.

In 2017 the police commenced trialling a system for collecting GPS coordinates for the crash scene locations. The Police patrol vehicles have been fitted with Automatic Vehicle Location (AVL) system which is connected through the police radio network to the police control room. The collection of GPS locations by this system went live throughout Montenegro in January 2018. The GIS mapping system used to display the GPS data uses Vector and satellite maps.

The police database is able to export crash data using an Excel spreadsheet format. The Ministry of Transport and Communications do not currently have a GIS capability. Proposals to develop a system during 2018 have been agreed. In 2011 a combined police and Ministry blackspot analysis project was undertaken where some 90 locations were identified.

The national and local ambulances of the Department for the Emergency Medical Aid, provided by the Ministry of Health have also been fitted with the same AVL system. Each ambulance is crewed by a

Page 121

registered doctor who completes a paper-based form for each patient they attend. The doctor uses the C10 injury coding system to describe the severity of injury, they also try and locate the patient to a vehicle and seat they occupied. When the patient arrives at the hospital a copy of the incident report from the ambulance remains with the patient. The duty doctor at the hospital also examines the patient and completes a report. The report includes an injury assessment (C10) for the patient which is provided to the Police to enable them to classify the injury severity.

The police provide crash data to the statistical office of Montenegro (MONSTST) who published a yearly statistical book in which it reproduces the current crash data records. The year book for 2017 provides details for 2016. Section 19-12 states there were 65 persons killed with 2,358 injured from 5,229 collisions. No other analysis or breakdowns are provided.

The current crash data fields being recorded are in line with the basic CADaS datasets. Institutions are aware of the need to enhance the collection process and are putting into place the components to achieve this especially associated with obtaining the GPS coordinates for the crash scene. The most significant limitation to enhancing the data collection process is the existing database being used.

4.3.6 Serbia

The Police in Serbia are responsible for attending and investigating road crashes. They use a paper-based form to record their findings. The content of the forms is quality checked by a senior officer prior to it being entered into the police database.

A project named “New Road Safety Database”, funded by the World Bank completed in 2016 provided the capability of connecting all relevant road safety data holders to a unique database. The greatest improvement to the datasets was in the accuracy of injury statistics and data from insurance companies. The crash database has a quality audit capability to check the accuracy of the information recorded. The location of the crashes is achieved by using Garmin GPS units and also from the police radio system.

A pilot study in 2017 trialled the inclusion of photographs and scanned sketch plans into the crash reports on the database. The limited storage capacity of the system was found to be an issue and the proposal has been dropped until additional storage capacity can be identified.

PE Roads of Serbia receive a monthly statistical data download from the police which they encode into their GIS mapping database ARCMAP. They undertake analysis of the data and publish their findings on the internet (www.bazabs.abs.gov.rs).

An example of the data produced by PE Roads that is available on the internet created from the police crash data. The map provides a colour coded location of fatal, serious and slight injury crashes.

An ability to plot selective data onto the map from the crash data fields is provided. The search and analytical capabilities of this system is currently limited and requires enhancement.

Page 122

The current crash data fields being recorded comply with mandatory fields in the advanced CADaS dataset database that is being used. The police database has limitations on the fields it records and on the exporting capability of the system. 4.4 Concept for a common system in WB6 based on EU practice

4.4.1 Data collection

Creating a common approach to the collection, recording and analysis of crash data within the WB6 Regional Participants will have a significant effect on its ability to undertake a holistic approach to reducing the road casualty statistics.

To achieve such an approach will require the linking of the various databases within each individual member of the WB6 region to permit the free flow of data between the Ministries and road safety agencies. The ultimate goal will then be to share statistical data with all of the other WB6 Regional Participants. The linking of databases has the major benefit associated with a “big data” approach. This style of approach has been found to generate new insights into identifying common factors that regionally affect road safety.

This common approach must start with realising a uniform set of information that must be collected by the Police, Medical and Road engineers in each Regional Participant. This information will only include the statistical data; which has no personal identification values, as defined by the CADaS methodology.

Within the WB6 Region all the data currently being collected by the various police departments are entered into their own specially developed databases. These databases vary in content and sophistication from the stand-alone Access database in Albania to the computer system in Kosovo. The majority of the police databases contain a significant amount of data associated with the work of the Police. Statistical cash data forms only a very small part of the information held within these systems. The majority of the current Police databases have not been designed, nor are they suitable to be used as a map-based crash data analytical system.

4.4.2 Data linking and sharing

Internationally, it has been seen that requiring the police alone to collect all the statistical data is unrealistic. An approach where other agencies provide relevant data associated with their area of expertise provides significant benefits. Such an approach will allow the police to concentrate on collecting the core transient scene data while the other responsible agencies provide more detailed information associated with their roles and responsibilities. Such a system will provide a significant enhancement in the quality of data than currently achievable.

The medical and engineering data within the WB6 region is, as a norm, predominately recorded using paper-based systems and as yet not being encoded onto any computer database, albeit there are proposals to achieve this in many Regional Participants. The inclusion of this data is essential to achieving a holistic and meaningful analysis. However, achieving this will create a real challenge even though there are a number of simple solutions to the problem.

Page 123

Internationally the best solution found has been to use a separate database, which is designed to link to all the Police, Medical and engineering databases as well as providing advanced crash data analysis and GIS capabilities. The linking of data across databases requires a pre-determined methodology that will enable data from one source to be matched with another; a person’s injury profile provided by the ambulance can be matched to the same person in the police accident report. A number of approaches to achieving this will be explored in the following sections.

4.4.3 Data Analysis

The primary reason for collecting crash data is often lost in the drive to develop methodologies to achieve the collection process. This report and the others before have all concentrated on what data should be collected without much emphasis on what should be done with it once it has been encoded into a database.

In simple terms the statistical information that is collected should describe the following components of the crash:  The scene, time, date, location, weather and description of the road etc.  The vehicles and objects involved  The casualties, their injury profiles  The pre-impact movements and actions of the vehicles / people involved  How the vehicles / people interacted (impacted) with each other  The post impact to rest movements of the vehicles / people involved.

The crash data, once collected, should be used to identify patterns and trends that can provide insights into why the crashes are happening. Such analysis must go beyond simply stating excess speed, driver behaviour and road defect.

Internationally it has been identified that adopting the Safe System approach, which represents a fundamental shift in road safety policy, has achieved significant benefits. At its core the Safe Systems approach accepts, as road users, we will make errors of judgment that will result in an incident. The goal however is to ensure that any road user that is involved in a collision will not sustain trauma that results in a fatality or life changing injury. Modern vehicles are built with this Safe System approach concept central to their design, provided they are driven within the limits of the law and environment (speed limit, road, weather, traffic conditions etc.). Passenger cars are also designed to ensure the passenger compartment maintains its integrity during a crash. The safety features within the vehicle will protect the occupants as long as they are seated correctly and wear their seatbelts.

The basic strategy of a Safe System approach is to ensure that in the event of a crash, the impact energies remain below the threshold likely to produce either death or life changing injury. It sees the road user as the weakest link in the transport chain, unpredictable and capable of error, education and information efforts notwithstanding. Although this is a very long-term objective, it transforms the level of ambition. The safe system approach opens up new potential for improving performance by addressing all elements of the road transport system together and finding synergies for trauma reduction when safer road and vehicle design, speed limits and compliance with road rules, are pursued in concert.

The Safe System approach recognises that humans as road users are fallible and will make mistakes. There are also limits to the kinetic energy exchange which humans can tolerate (e.g. during the rapid deceleration associated with a crash) before serious injury or death occurs. A key part of the Safe System approach requires that the road system be designed to take account of these errors and vulnerabilities so that road users are able to avoid serious injury or death on the road. Where a road user receives fatal or life changing injuries the analysis process should identify why this occurred.

Page 124

4.5 Road Map

The diagram below provides the outline of activities that must be adopted to achieve a common approach to creating quality crash data records.

•Conforms to Advanced CADaS datasets principles •Datasets are identical albeit translated to the local language Standardised Data

•Paper based records checked prior to encoding •Audit methodology of database records – quality, errors and omissions. Quality Check Data

•Real time sharing with Transport and Health Ministies •Online sharing with Road Safety Agencies Internal Sharing of •Online sharing with Official Statistical Institutions data

•Share data with neighbouring Regional Participants •Share data with SEETO Sharing data within WB6 Region

•CARE database •World Health Organisation Yearly reports to external agencies

4.5.1 Standardised Data

To achieve a common approach as stated earlier the first aspect that needs to be considered is the data that is actually available / collected. The current collection practices within the WB6 Regional Participants, with respect to crash data, is predominately achieved using a paper-based collection medium. The information collected by this process is currently undertaken by the police and can be divided into two components:

1. Data specifically relating to the legal requirement of the Police to investigate and record the facts of the incident for a criminal court process. This data invariably contains sensitive information that identifies the persons involved.

Page 125

2. Data describing the road, vehicles and casualties; this is known collectively as statistical information.

Only statistical data is the subject of this report; that which is designed to provide a description of the crash to enable analysis to determine what factors were present that effected / caused the crash.

In every road crash there are always multiple factors that preceded the actual collision. Any analysis of a collision will require information associated with what occurred prior to the collision, how the objects involved collided and finally how and where they came to rest.

The European Union has developed the CADaS data sets that it considers essential to allow any meaningful analytical analysis to be undertaken. These parameters will therefore be taken as the base factors that should be collected by each WB6 Regional Participant and are reproduced below:

4.5.2 CADaS Dataset

The SweRoad’s report of 2011, presented earlier in this report, set out the status of data collected by each Regional Participant against the basic CADaS datasets. Most Regional Participants today achieve these basic datasets as a norm. The need now is to enhance the datasets in line with the advanced CADaS datasets methodology.

Internationally, the interpretation of the elements from one language to another has been found to be one of the main problems. The manuals produced by the EU on CADaS are primarily produced in English however other languages are available. Within the WB6 region an agreed translation by each Regional Participant for each element should be made and a formal document produced.

The elements of the Advanced CADaS datasets V 3.6 2017 are reproduced below, those elements highlighted in blue are the basic dataset alternatives. It is recommended all Regional Participants develop these as a minimum but should strive to achieve the goals of this report by adopting the more detailed CADaS options. The elements highlighted in green are for European countries and will not form part of this report, as they are seen as long-term requirements for WB6.

The proposal is that all the below datasets will be adopted within WB6 regional as the de-facto unified approach to crash data recording convention.

Page 126

Code Description Importance Sub code Elements Accident A-1 Accident ID High Country code/Region/year/Number A1b Year code 4 digits A1c Accident number 6-digit number A-2 Accident date High DD/MM/2018 A-3 Accident time High 00:00 A-4 NUTS A-4 Nomenclature of Territorial Units A-5 LAU Local Administrative units A-6 Weather conditions High A-6.01 Dry / Clear A-6.02 Rain A-6.03 Snow A-6.04 Fog, Mist, Smoke A-6.05 Sleet, Hail A-6.06 Severe winds A-6.07 Other A-6.99 Unknown A-7 Light conditions High A-7.01 Daylight A-7.02 Twilight A-7.03 Darkness Streetlights lit A-7.04 Darkness Streetlights unlit A-7.05 Darkness no streetlights A-7.06 Darkness streetlights unknown Darkness no streetlights or A-7.07 streetlights unlit A-7.99 unknown A-8 Accident with Pedestrians Low A-8.00 Not applicable Pedestrian crossing street - no A-8.01 turning of vehicle - outside of junction Pedestrian crossing street - no A-8.02 turning of vehicle - at a junction Pedestrian crossing street - no A-8.52 turning of vehicle - not specified Pedestrian crossing street - no A-8.03 turning vehicle - not specified Pedestrian crossing - turning of A-8.04 vehicle turning right (left) Pedestrian walking along the road A-8.53 or stationary in the road Pedestrian crossing - turning of A-8.05 vehicle turning left (right) Pedestrian crossing - turning of A-8.06 vehicle turning- not specified A-8.07 Pedestrian stationary in the road A-8.08 Pedestrian walking along the road Pedestrian on pavement or bicycle A-8.09 lane Pedestrian walking along the road A-8.10 or stationary in the road A-8.11 Pedestrian accident - other A-8.99 Pedestrian accident - unknown A-9 Accident with parked car Low A-9.00 Not Applicable Hitting parked vehicles right (left) A-9.01 side of road Hitting parked vehicles left (right) A-9.02 side of road Hitting parked vehicles - side of the A-9.51 road - not specified Hitting parked vehicles - side of A-9.03 road - not specified Accident with parked vehicles - A-9.04 opening doors Other accident with parked A-9.05 vehicles Accident with parked vehicles - A-9.99 unknown A-10 Single vehicle collision Low A-10.00 Not applicable Single vehicle accident with A-10.01 animals

Page 127

Code Description Importance Sub code Elements Accident Single vehicle accidents obstacles A-10.02 on or above the road Single vehicle accidents with road A-10.03 work materials A-10.04 Accident between train and vehicle Single vehicle accidents with A-10.05 obstacles - other Single vehicle accident with AA-10.51 obstacles on the road - not specified Single vehicle accidents- leaving A-10.06 straight road - either side of the road Single vehicle accidents in a bend A-10.07 - going either side of the road A-10.08 Single vehicle accident on the road Single vehicle accident including A-10.09 rollover Single vehicle accident in junctions A-10.10 or entrances Single vehicle accidents without A-10.11 obstacles - other Single vehicle accidents without AA-10.52 obstacles - not specified A-10.99 Single vehicle accident - unknown A-11 At least 2 vehicles not turning Low A-11.00 Not applicable At least two vehicles - same A-11.01 direction - overtaking At least two vehicles - same A-11.02 direction - rear end collision At least two vehicles - same A-11.03 direction - entering traffic At least two vehicles - same A-11.04 direction - side collision At least two vehicles - same A-11.05 direction - others At least two vehicles - same A-11.51 direction no turning - not specified At least two vehicles - head on A-11.06 collision in general At least two vehicles - opposite A-11.07 direction no turning - reversing At least two vehicles - opposite A-11.08 direction no turning - others At least two vehicles - opposite AA-11.52 direction no turning not specified At least two vehicles - others no A-11.09 turning At least two vehicles - no turning - A-11.99 unknown At least 2 vehicles turning or A-12 Low crossing A-12.00 Not applicable At least two vehicles - turning A-12.01 same road - same direction - rear end collision At least two vehicles - turning A-12.02 same road - same direction - U- turn in front of the other vehicle At least two vehicles - turning A-12.03 same road - same direction - turning right (left) At least two vehicles - turning A-12.04 same road - same direction - turning left (right) At least two vehicles - turning A-12.05 same road - same direction - others At least two vehicles - turning AA-12.51 same road - same direction - not specified

Page 128

Code Description Importance Sub code Elements Accident At least two vehicles - same road - A-12.06 opposite direction - turning left (right) in front of another vehicle At least two vehicles - same road - A-12.07 opposite direction - U-turn in front of another vehicle At least two vehicles - same road - A-12.08 opposite direction - turning into same road At least two vehicles - same road - A-12.09 opposite direction - turning into opposite roads At least two vehicles - same road - A-12.10 opposite direction - turning right (left) in front of another vehicle At least two vehicles - same road - A-12.11 opposite direction - turning others At least two vehicles - turning or AA-12.52 crossing - same road- opposite direction - not specified At least two vehicles crossing (no A-12.12 turning g)- different roads At least two vehicles - different A-12.13 roads- turning right (left) in front of vehicle from the left (right) At least two vehicles - different A-12.14 roads- turning right (left) - head on collision At least two vehicles - different A-12.15 roads- both vehicles turning At least two vehicles - different A-12.16 roads- turning left (right) into traffic from the right (left) side At least two vehicles - different A-12.17 roads- turning left (right) into traffic from the left (right) side At least two vehicles - different A-12.18 roads- turning into traffic - others At least two vehicles - different AA-12.53 roads- not specified At least two vehicles - crossing or A-12.19 turning - others At least two vehicles - crossing or A-12.99 turning - unknown A-13 Hit & run Accident High A-13.00 Not applicable A-13.01 Not Hit & Run A-13.02 Hit & Run A-13.99 Unknown Road A-1 Accident ID High same as A-1 R-1 Latitude High R-1 Latitude R-1.9999999 Unknown R-2 Longitude High R-2 Longitude R-2.9999999 Unknown R-3 E Road Low R-3.0000 Not applicable R-3 E-road code R-3.9999 Unknown R-4 E Road Kilometre Low R-4.0000 Not applicable R-4.0000 E-road Kilometre R-4.9999 Unknown R-5 Function class 1st road High R-5.01 Principle arterial R-5.02 Secondary arterial R-5.03 Collector R-5.04 Local R-5.05 Other R-5.99 Unknown R-6 Function Class 2nd Road High

Page 129

Code Description Importance Sub code Elements Accident R-6.01 Principle arterial R-6.02 Secondary arterial R-6.03 Collector R-6.04 Local R-6.05 Other R-6.99 Unknown R-7 AADT 1st Road R-8 AADT 2nd Road R-9 Speed limit 1st road High R-9 Speed limit R-9.001 No speed Limit R-9.999 Unknown RA-9.501 <30 km/h RA-9.502 30-50 km/h RA-9.503 51-80 km/h RA-9.504 81-100 km/h RA-9.505 101-120 km/h RA-9.506 >120 km/h R-10 Speed limit 2nd road High R-10 Speed limit R-10.001 No speed Limit R-10.999 Unknown RA-10.501 <30 km/h RA-10.502 30-50 km/h RA-10.503 51-80 km/h RA-10.504 81-100 km/h RA-10.505 101-120 km/h RA-10.506 >120 km/h R-11 Motorway High R-11.01 Yes R-11.02 No R-11.99 Unknown R-12 Urban Area High R-12.01 Inside R-12.02 Outside R-12.99 Unknown R-13 Junction High R-13.00 Not at a junction R-13.01 At-grade - crossroads R-13.02 At-grade roundabout R-13.03 At-grade T or staggered junction R-13.04 At-grade multiple junction R-13.05 Not at grade (interchange) R-13.06 Other R-13.07 At a level crossing R-13.99 Unknown R-13.51 At a junction - Not specified R-14 Relation to Junction / Interchange Low R-14.00 Not applicable R-14.01 Approaching (20 m) R-14.02 Acceleration / deceleration lanes R-14.03 Through Roadway R-14.04 Entrance / exit ramps R-14.05 Crossing related R-14.06 Intersection R-14.99 Unknown R-15 Junction Control Low R-15.00 Not applicable R-15.01 Authorised person R-15.02 Give way signs R-15.03 Automatic traffic signal R-15.04 Uncontrolled R-15.99 Unknown R-16 Surface condition High R-16.01 Dry R-16.02 Snow, frost, ice slush R-16.03 Slippery R-16.04 Wet, damp R-16.05 Flood R-16.06 Other R-16.99 Unknown R 17 Obstacles Low R-17.01 Yes

Page 130

Code Description Importance Sub code Elements Accident R-17.02 No R-17.99 Unknown R-18 Carriageway type High Single carriageway - one-way R-18.01 street Single carriageway - two-way R-18.02 street R-18.03 Dual carriageway R-18.04 Single carriageway - not specified R-18.99 Unknown R-19 Number of lanes High Number of lanes (in one or two R-19 directions) R-19.99 Unknown RA-19 Total number of lanes R-20 Emergency Lane Low R-20.01 Yes R-20.02 No R-20.99 Unknown R-21 Markings Low R-21.01 None or faded R-21.02 Only separating travel direction Separating travel direction and R-21.03 lanes R-21.04 Only separating lanes R-21.05 Other R-21.99 Unknown R-22 Tunnel Low R-22.01 Yes R-22.02 No R-22.99 Unknown R-23 Bridge Low R-23.01 Yes R-23.02 No R-23.99 Unknown R-24 work zone related High R-24.01 Yes R-24.02 No R-24.99 Unknown R-25 Road Curve Low R-25.01 Yes R-25.02 No R-25.99 Unknown R-26 Road Segment grade Low R-26.01 Yes R-26.02 No R-26.99 Unknown Vehicles A-1 Accident ID High Same as A-1 U-1 Traffic Unit ID High U-1 Traffic unit ID U-2 Traffic Unit Type High U-2.01 Pedal cycle U-2.02 Moped U-2.03 Motorcycle up to 125cc U-2.04 Motorcycle over 125cc U-2.05 Passenger car U-2.06 Minibus U-2.07 Bus U-2.08 Coach U-2.09 Trolley U-2.10 Goods vehicle under 3.5 t MGW U-2.11 Goods vehicle over 3.5 t MGW U-2.12 Road tractor U-2.13 Agricultural tractor U-2.14 Tram / light rail U-2.15 Ridden animal U-2.16 Other motor vehicle U-2.17 Other non-motor vehicle U-2.18 Pedestrian U-2.19 Quad up to 50cc U-2.20 Quad over 50cc U-2.99 Unknown

Page 131

Code Description Importance Sub code Elements Accident UA-2.51 Two-wheel motor vehicle UA-2.52 Bus or minibus or coach or trolley UA-2.53 Goods vehicle UA-2.54 Motorcycle not specified U-3 Vehicle Special function Low U-3.00 Not applicable U-3.01 No special function U-3.02 Taxi U-3.03 SUV / off-road vehicle U-3.04 Vehicle used as school bus U-3.05 Vehicle used as scheduled bus U-3.06 Military U-3.07 Police U-3.08 Ambulance U-3.09 Fire-truck U-3.10 Dangerous goods vehicle U-3.99 Unknown UA-3.52 Special vehicle U-4 Trailer High U-4.00 Not applicable U-4.01 Without trailer U-4.02 With trailer U-4.99 Unknown U-5 Engine Power Low U-5.000 Not applicable U-5 Engine power U-5.999 Unknown U-6 Active Safety Equipment Low U-6.00 Not applicable U-6 Active safety equipment U-6.98 other U-6.99 Unknown U-7 Vehicle Drive Low U-7.00 Not applicable U-7.01 Left hand drive U-7.02 Right hand drive U-7.99 Unknown U-8 Make Low U-8.000 Not applicable U-8 Motor vehicle make U-8.999 Other / Unknown U-9 Model Low U-9.00 Not applicable U-9 Motor vehicle model U-9.99 Unknown U-10 Registration Year High U-10.0000 Not applicable U-10 Registration year U-10.9999 Unknown U-11 Traffic unit Manoeuvre High U-11.00 Not applicable U-11.01 Reversing U-11.02 Parked U-11.03 Entering a parking position U-11.04 Leaving a parking position U-11.05 Waiting to go ahead but held up U-11.06 Slowing or stopping U-11.07 Moving off U-11.08 U turn U-11.09 Waiting to turn left U-11.10 Turning left U-11.11 Waiting to turn right U-11.12 Turning right U-11.13 Changing lanes to left U-11.14 Changing lane to right U-11.15 Avoidance manoeuvre U-11.16 Overtaking vehicle on its left U-11.17 Overtaking vehicle on its right U-11.18 Going around left-hand bend U-11.19 Going around right-hand bend U-11.20 Straight forward / normal driving Entering or leaving a parking UA-11.51 position

Page 132

Code Description Importance Sub code Elements Accident UA-11.52 Waiting to turn UA-11.53 Turning UA-11.54 Changing lane UA-11.55 Overtaking Pedestrian Manoeuvres U-11.21 Crossing (on pedestrian crossing) U-11.22 Crossing (on other point) Walking on the carriageway, facing U-11.23 traffic Walking on the carriageway, back U-11.24 to traffic Standing or playing on the U-11.25 carriageway Not on the carriageway (on U-11.26 sidewalk, pedestrian road etc.) U-11.27 Lying on the carriageway U-11.28 Entering or getting out of a vehicle UA-11.56 Crossing Walking or standing on the UA-11.57 carriageway U-11.98 Other U-11.99 Unknown U-12 First point of impact Low U-12.01 No impact U-12.02 Left front U-12.03 Centre front U-12.04 Right front U-12.05 Right side U-12.06 Right rear U-12.07 Centre rear U-12.08 Left rear U-12.09 Right rear U-12.99 Unknown UA-12.51 Front not specified UA-12.52 Rear not specified U-13 First object hit in road Low U-13.00 Not applicable U-13.01 None U-13.02 Object from previous accident U-13.03 Parked vehicle U-13.04 Bridge U-13.05 Bollard / refuge U-13.06 Central island of roundabout U-13.07 Kerb U-13.08 Animal (except ridden animal) U-13.09 Other object U-13.10 Train U-13.99 Unknown U-14 First object hit off road Low U-14.00 Not applicable U-14.01 None U-14.02 Road sign / traffic sign U-14.03 Lamp post U-14.04 Pole U-14.05 Tree U-14.06 Bus stop / shelter U-14.07 Central crash barrier U-14.08 Crash barrier beside carriageway U-14.09 Ditch U-14.10 Parked vehicle U-14.11 Stone / rock / mountainside U-14.12 Fence U-14.13 Submerged in water U-14.14 Other permanent object U-14.99 Unknown U-15 Insurance Low U-15.00 Not applicable U-15.01 Insured for vehicles U-15.02 Not insured for vehicles U-15.99 Unknown U-16 Hit & run High U-16.00 Not applicable U-16.01 Not hit and run

Page 133

Code Description Importance Sub code Elements Accident U-16.02 Hit and run U-16.99 Unknown U-17 Registration Country High U-17.000 Not applicable U-17 Country code UA-17.501 National UA-17.502 Foreign Casualties A-1 Accident ID High U-1 Traffic Unit ID High U-1 Traffic ID U-1.99 Unknown traffic unit P-2 Year of Birth High P-2.9999XXXX Year of birth Year and month of birth (day P-2.99XXXXXX unknown) P-2.XXXXXXX Date of birth P-2.99999999 Unknown PA-2.0000XXXX Years and months of person P-3 Gender High P-3.01 Male P-3.02 female P-3.99 Unknown P-4 Nationality High P-4 Nationality PA-4.951 Nationality PA-4.952 Foreigner, within the EU PA-4.953 Foreigner, outside the EU PA-4.954 Foreigner, not specified PA-4.999 Unknown P-5 Injury Type High P-5.01 Fatal injury P-5.02 Seriously injured P-5.03 Slightly Injured P-5.04 Not injured P-5.99 Unknown PA-5.51 Injured P-6 Road User Type High P-6.01 Driver P-6.02 Passenger P-6.03 Pedestrian P-6.99 Unknown P-7 Alcohol Test Low P-7.00 Not applicable P-7.01 Tested P-7.02 Not tested P-7.99 Unknown P-8 Alcohol sample type Low P-8.00 Not applicable P-8.01 Blood sample P-8.02 Breath sample P-8.99 Unknown P-9 Alcohol result High P-9.00 Not applicable P-9.01 Positive P-9.02 Negative P-9.99 Unknown P-10 Alcohol Level High P-10.000 Not applicable P-10 Level P-10.999 Unknown P-11 Drug test Low P-11.00 Not applicable P-11.01 Positive P-11.02 Negative P-11.03 Not tested P-11.99 Unknown

P-12 Driving License Issue date High P-12.000000 Not applicable P-12.999999 Unknown Number of years and months of P-12.00XXXX driving experience

Page 134

Code Description Importance Sub code Elements Accident P-13 Driving License Validity Low P-13.00 Not applicable P-13.01 With appropriate driving license P-13.02 With inappropriate driving license P-13.03 Only driving lesson or driving test Invalid or suspended driving P-13.04 license P-13.05 No driving license P-13.06 No license required P-13.99 Unknown PA-13.51 Invalid (or no) driving license P-14 Safety Equipment High P-14.00 Not applicable P-14.01 Seat belt worn no airbag in vehicle P-14.02 Seat belt worn and airbag released Seat belt worn and airbag not P-14.03 released Seat belt not worn and airbag P-14.04 released P-14.05 Crash helmet worn Child safety seat facing forwards P-14.06 used Child safety seat facing backwards P-14.07 used No use of safety equipment (seat P-14.08 belt - helmet) Other (appropriate equipment for P-14.09 bikers and cyclists e.g. protective pads, reflective clothing, lighting) P-14.99 Unknown (it was not recorded) PA-14.51 Seat belt worn - not specified Child safety seat used - not PA-14.52 specified P-15 Seating Position in/on vehicle High P-15.00 Not applicable P-15.01 Driver P-15.02 Front seat P-15.03 Rear - seated P-15.04 Rear - standing PA-15.51 Rear - not specified P-15.05 Elsewhere P-16 Distracted by device Low Driver or pedestrian P-16.00 Not applicable P-16.01 Not distracted by device P-16.02 Telecommunication device P-16.03 Another electronic device P-16.99 Unknown Psychophysical / Physical P-17 Low Impairment P-17.00 Not applicable P-17.01 Good Inattention / absence of mind / P-17.02 Worried P-17.03 Tired / fell asleep Illness / Sudden illness / Lost P-17.04 consciousness P-17.05 Defective eyesight or hearing Dazzled by sunlight / vehicle P-17.06 headlights P-17.07 Others P-17.99 Unknown P-18 Trip Journey Purpose Low P-18.00 Not applicable Route to/from school - education / P-18.01 route to / from work P-18.02 Driving as part of the work P-18.03 Leisure/Entertainment P-18.04 Holiday P-18.05 Driving lesson P-18.06 Other P-18.99 Unknown P-19 Injury MAIS scale Low P-19.00 Not Injured

Page 135

Code Description Importance Sub code Elements Accident P-19.01 Minor P-19.02 Moderate P-19.03 Serious P-19.04 Severe P-19.05 Critical P-19.06 Maximum P-19.51 MAIS Minor p-19.53 MAIS 3 Plus P-19.99 Unknown

4.5.3 Regional Variations and Omissions by CADaS

The review process identified that the WB6 region is in the process of upgrading their paper-based report forms. It was originally proposed that this report would provide an assessment of these paper- based reports to identify any omissions, this however would not serve any real benefit as the assessment would only be correct at the time of writing.

The table above setting out the advanced datasets for CADaS should be used as the definitive list which each region should ensure their paper-based form aligns with. It is recommended that an agreed translation by each Regional Participant of these elements is achieved and circulated within the WB6 region through SEETO. It is also recommended the reference number for each element used by the CADaS system becomes the accepted protocol for describing the dataset; this will greatly facilitate the sharing of data.

4.5.4 Additional Data

As part of a multi-agency approach the inclusion of medical data is one which will provide the greatest ability to map the elements associated with the most serious incidents. The collection of medical data usually starts with the arrival of an ambulance and its crew at the scene of a crash. The process commences with initial medical assessment of the casualty. The quality of the data available will depend on the status of the medical staff present; first aider through to paramedic and finally to trauma doctor. In most cases a simple coding system is used developed around the Maximum Abbreviated Injury Scale (MAIS)

In 2013, the EC (European Commission) adopted a new definition of seriously injured road victims based on the International Classification for Diseases (ICD10). All road victims with a MAIS score of 3 or more (MAIS3+) are considered as severely injured. This new definition will coexist along with the conventional definition of severely injured, namely persons who stay at least 24 hours in hospital.

Within the WB6 region many medical facilities use the International Classification for Diseases (ICD10) system to define trauma injury. There are a number of technical reviews which provide an exact correlation between the two systems14.

The MAIS scale provides a medical explanation to define the injury of a casualty:  MAIS 1 Minor Injury  MAIS 2 Moderate Injury  MAIS 3 Serious Injury  MAIS 4 Severe Injury  MAIS 5 Critical Injury  MAIS 6 Un-Survivable Injury.

14 https://www.ncbi.nlm.nih.gov/pubmed/27736159

Page 136

An example would be as below a knee leg injury where:  MAIS 1 would be a sprained ankle  MAIS 2 would be a closed fracture of the Tibia bone  MAIS 3 would be open fracture of the Tibia bone  MAIS 4 would be amputation.

The type of other data that should be collected would include a diagram that is used to depict injury location.

A number of other factors Data associated with the response Collection time of the ambulance and Methodology subsequent time interval before the patient arrives at a trauma care centre that Police Medical should be collected. Internationally it has been identified that post-crash response and treatment plays a significant part in the reduction in trauma injury and death resulting from road crashes. Appendix D provides an advanced crash report form Crash Report Casualty data that includes additional data sets that could be collected from the medical sector.

The linking of the medical data with the police crash data will Police necessitate the creation of agreed protocols. Medical data is Database intrinsically linked to a specific person by Name, date of birth and ID number which are considered sensitive data. While police crash data contains casualty names, date of birth and ID number it is all referenced under a single incident number. All statistical data, once exported from the police database, has names and sensitive details removed.

To enable medical data to be linked to crash data the process needs to be undertaken prior to the exporting of the statistical data as per the adjacent chart.

4.5.5 Data Encoding

Agreement within the WB6 Region Participants will need to be reached defining the protocol for encoding the paper-based reports into the databases. Currently the following procedures being adopted can be described as: 1. The key facts associated with the crash must be encoded into the database within a limited time period; often prior to the reporting officer going off duty or within 24 hours of the incident. The subsequent encoding of the data should be completed within a given time period. 2. Only after the completion of the investigation into the crash will the crash report be encoded into the database, this is usually undertaken by a person other than the reporting officer.

To achieve a real time data sharing capability between Ministries will require the initial report of the collision to be encoded into the database prior to the investigating officer going off duty on the day of the incident.

The actual encoding of the data can be undertaken either by the investigating officer or by dedicated personnel. Both systems have advantages as well as disadvantages these primarily are associated with an actual knowledge base of the incident as opposed to a simple copying of the paper report.

Page 137

Where the encoding is undertaken by dedicated personnel the crash report form, or a photocopy of it, is passed to the support staff for entry into the computer database system. There are variations on this model, in some locations (e.g. South Africa), crash report forms are read by optical scanners to populate the database automatically. In the UK some police authorities call up an operator who takes the officer through the questions and enters the answers into the database over the phone.

The descriptive words used for each field always loses something in translation from one language to another, however the most important factor is that their meaning is correctly understood by the Investigating Officers completing the form and subsequently by the users of the database. Where someone other than the investigating officer encodes the data the use of photographs is an essential element in achieving a higher quality of the data especially if the photographs are also uploaded into the database file.

In describing a scene, vehicle or object the creation of a photograph or sketch plan enhances the ability to describe the scene, objects involved and factors immensely. Quality scene photographs provide a wealth of information far beyond any descriptive text.

The photograph above provides a complete scene description; it also provides information about:

 The semi-rural setting of the incident; bounded by trees and hedgerows with no visible junctions.  The straight and level layout of the road as well as the sight lines available.  The condition of the road; its width and the surface condition with clear road markings as well as the fact it was wet.  The presence of street lighting.  The resting position of the vehicles.  Limited damage assessment.

If a photograph had been taken from the other side of the scene to show the road layout in the opposite direct; behind the current camera position, this would have provided a comprehensive record of the incident.

4.5.6 Mandatory Fields

Mandatory fields are the important datasets that must be completed within the database record before the record can be initiated. The mandatory fields have been highlighted in the following table outlining the elements of the advanced CADaS datasets. The column labelled ‘importance’ has two values ‘High and Low’; those denoted as ‘High’ are the mandatory fields. These fields can simplistically be described as:  Time, date of the incident  Location of the incident – GPS coordinates  Severity of the crash – fatal, serious injury, slight injury and damage only  Number of vehicles involved  Number of casualties involved  Road classification / type  Junction type  Initial description of the collision type – head-on etc.

Other levels of field completion can be dictated within the protocol however the most important requirement is the mandatory fields that must be completed before the crash record can be signed off as completed. The primary objective however should be to achieve a 100% completion of all data fields.

Page 138

4.5.7 Data Quality

Having defined what should be collected the next task is to develop Define Improve Data a methodology to ensure the data recorded is of the highest quality. Data Quality Quality Failure to achieve this requirement will bring all subsequent analysis Targets of the data into question. The process diagram opposite explains the process to be followed for the continuous monitoring and evaluation of the crash data Measure Report Data Data Quality Quality The necessity to develop a quality assessment protocol has been Indicators found to be essential. This protocol should dictate the time period when quality assessments should be made and the process of Analyse Data rectifying errors and omissions in the data. It will also need to include Quality the process for designating the record as complete.

Modern database programs have been developed with automatic checking and assessment subroutines that are capable of identifying errors and omissions and producing automated reports to the investigating officer and senior management.

The assessment protocol should dictate the data sets that should be encoded within specified time periods. As an example, a damage only or minor injury crash report should be encoded and the record reviewed and completed within 7 days of the initial incident being recorded.

Where a more serious injury of fatality is involved the initial 7-day review should be undertaken to identify any errors or omissions. However, in such cases the computer record may remain active to allow additions / amendments to the data resulting from further investigations. In such occasions a review process should be repeated every 7 days until either the file is completed or a period of one month (30 days) after the incident has elapsed as this is the agreed definition of a fatality.

The process should also provide the requirements that must be met before the record can be marked complete and closed. The normal requirement is for the senior database manager to provide the final assessment and closer activity.

The initial review process should incorporate an assessment of the paper-based crash report against the encoded computer records. Where errors or omissions are identified a report should be created and passed to the investigating officer as well as their senior management. The report should outline the errors or omissions and it is the responsibility of the officer’s senior management to ensure the records are corrected.

4.5.8 Data Sharing

Having achieved a common crash data recording protocol it is proposed that the initial data sharing activity should be undertaken at a local level between the relevant Ministries and Road Safety Stakeholders. Some members of the WB6 region are already achieving a limited data sharing capability while others only provide a periodic exchange of numerical tables.

The proposal of this report is to achieve a real time sharing of all the CADaS statistical data sets with all the relevant Ministries and Road Safety Agencies within a Regional Participant. This will require, in the first instance, the drafting of a memorandum of understanding (MOU) (Appendix C) between all the parties involved outlining the exchange mechanism, security protocols and timelines. The example MOU provided in Appendix C is one that has

Page 139

been developed and accepted internationally by many Police and Ministerial organisations as appropriate. The exact wording should be altered to address local legal requirements and protocols.

The actual exchange of data will require the creation of a compatible computer linking database or ‘HUB’ which will facilitate the exchange of data between each of the data sources. The central ‘HUB’ database will combine the data into records associated with each crash and display the information using advanced GIS technology. Many ‘HUB’ databases also combine advanced analytical capabilities which will allow the users to undertake complex data analysis.

It is ultimately proposed the sharing of all the CADaS datasets with neighbouring WB6 Regional Participants and SEETO or equivalent body. To achieve this sharing will require:  The drafting of a memorandum of understanding (MOU) (Appendix C) between all the parties involved outlining the exchange mechanism, security protocols and timelines.  Each recipient will require a compatible computer linking database or ‘HUB’ which will facilitate the exchange of data between each of the data sources.  An advanced analytical database to permit the end user to view and analyse the crash records provided.

4.5.9 Data Analysis

One of the main uses of crash data is to manage hazardous locations on the network. The tasks include identifying locations where excessive crashes are occurring. This activity is designed to identify the characteristics within the crashes that may point to underlying causation, for example it could assist the engineer to identify appropriate treatments. Where such work has been undertaken a reassessment of the crash data can be used to test that they have been successful. It could also assist the police in identifying adverse driver behaviour; speeding. Following an enforcement campaign, the reassessment of the crash data will identify if the campaign has been successful.

Although this method is less advocated by the Safe System, as it is reactive rather than proactive, it remains important to monitor the network for safety problems at discrete locations or junctions. These localised road defects can potentially be treated very effectively at low cost.

The point of identifying hazardous locations is to isolate the best sites which can be treated by engineering and/or enforcement measures. However, the sites which have the most crashes may not always actually be appropriate for enforcement or the top priority for cost effective treatment. The ability

Page 140

to identify, by using GIS mapping systems, the locations where higher crash densities occur is an essential component.

Sites identified which have a lot of crashes might simply have high traffic flows and may not have a very treatable safety issue. This is because crash numbers generally relate to the vehicle flow rate at the location. At junctions the numbers of turning movements occurring are also a factor in the crash numbers. Ideally Blackspot screening identifies locations where excessive crashes are occurring and for this reason some measure of exposure should be taken into account when prioritising these for treatment. The best sites for treatment will have both high crash numbers occurring but also high crash rates per unit of traffic volume.

The simplest way to take exposure into account, at least to some extent, is to assess safety on single routes or parts of the network where the flows do not vary too greatly. Thus, a location with a lot of crashes compared to the surrounding similar lengths is more likely to have a safety problem.

Working with crash rates (per 100 Million Vehicle Km) in addition to the crash numbers can help to identify sections which have higher crash risks. However, it is wrong to rely just on crash rates per unit traffic volume because a section with a small number of crashes but a low flow can have an extremely high-risk rate (crashes per unit of flow) and yet not be worth treating from an economic perspective.

Other potential sources of problems are simply that longer sections will also tend to have more crashes (crash numbers also relate directly to section length, all other things being equal). Thus, taking the crash density (crashes/Km) into account may also be important.

It may be useful to calculate typical crash rates and crash densities on roads on a particular network against which to compare rates of identified possible blackspots.

The most sophisticated method to screen for blackspots is to create crash Safety Performance Functions (SPFs) or APMs for links/junctions. These can be used to estimate the typical expected crash rates for given flows and other features, against which to judge whether the observed crash number are excessive. (See Crash Data Analysis and Engineering Solutions for Local Agencies (2009)).

The way crashes occur inherently includes a great deal of random variation, and crash occurrence is best described by the Poisson or Negative Binomial probability statistical distributions. These distributions mean that some “problem” locations with a lot of crashes that have been identified by trawling the plotted data using GIS cluster (density) analysis, will not actually have any specific safety problem that can be treated. They are generated “randomly”. The problem tends to resolve over time – a process called Regression to the Mean. This means that some sites with a high crash occurrence at a particular point are likely to have reduced incidence in future without any interventions being implemented. This effect has also led to exaggerated estimates being made for the effectiveness of safety measures, since some of any reduction measured is likely to be due to the natural return to the longer-term mean crash rate for a location in the after period.

The effect of Regression to the Mean can be minimised by searching for clusters using 3 or better 5 years of data; over these periods the effect is likely to be smaller, especially if crash numbers involved at sites are high. Other statistical methods can be used to control for the Regression to the Mean issue, the best known being the Empirical Bayes (EB) method, which can be used to both screen for blackspots and also to estimate the true before/after savings when measures have been implemented.

The regression to the mean issue can only be detected or is a highly relevant concern when dealing with high quality data with good location coordinates and where reporting rates are high.

Best international practise is to use Economic Appraisal methods to make an estimation of how cost effective each of a number of possible treatments will be in terms of the financial return likely. This approach can be used to identify the best treatment of a number of alternatives for the same location, or to optimise returns from investment in safety for a programme of treatments when the budget is limited.

Page 141

Although this is recommended good practise for all road safety investments (big or small) it is not actually widely done (see Elvik and Veisten 2005, SWOV 2010). Problems which prevent the wider spread use of economic appraisal are the relative technical complexity of the methodology, the challenging data requirements, such as the need for estimated crash and/or casualty costs.

For what are in infrastructure development terms, generally small road safety investments, the challenging, full scale transport economic appraisal methodology (which requires environmental and all wider social and transportation costs to be included) can be considerably relaxed. For small safety targeted improvements, it is assumed that there will be little or no effect on trip generation, other social costs and also Carbon costs are seldom calculated. Methodologies most frequently used are the very simple First Year Rate of Returns, cost effectiveness and the more technically demanding Benefit to Cost Ratio calculations (Road safety Foundation 2011).

Economic Appraisal should allow the direct comparison of the economic returns likely from a range of proposed schemes so that where budgets are limited those which will result in the greatest returns can be identified and prioritised.

In order to routinely conduct economic appraisal, organisations need to have the following data available:  Typical costs of various works or measures  Crash/casualty costs by severity  Estimates of measure safety impact  Values for the national adopted “discount” rate  Values for national GDP growth and deflators.

In any situation where investments are made in order to improve a road safety problem it is important to objectively assess if the intervention has worked effectively. In the case of road safety, it is generally the aim to reduce crashes and casualties; this can be tested by comparing the occurrence of crashes before and after intervention. Because of the large random variation in crash patterns it is particularly important to apply statistics to identify if any reduction in crashes in the after period is statistically significant: that is greater than that expected through random variation at the 95% confidence level.

It is generally agreed that there are 4 main statistical methodologies available. A recent ITE (ITE 2009) document summarises the pros and cons of the main methods concisely.

The main approaches are:  Naïve Before-and-After Study o Uses before data to estimate expected after crash number  Before-and-After Study with Yoked Comparison o Compares before/after numbers at each treated site with a selected specific comparator  Before-and-After Study with Comparison Group o Compares before/after numbers at a site with a range of comparator areas  Before-and-After Study with the Empirical Bayes (EB) Approach o Requires significant archive data o Requires SPF models to be developed.

The clear recommendation is that the EB approach be used if possible but that this requires sufficient data to enable crash relationship models (e.g. SPFs) to be developed. The second-best methodology is “Before-and-After Study with Comparison Group” which accounts for extraneous trend affects but not regression to the mean. It is possible to take Regression to The Mean into account by other less accurate but easier methods (see DfT 2004).

Page 142

4.5.10 Computer programs

To achieve the more complex analysis as outlined above computer programs have been developed to automate the task. These programs are designed with user friendly interface that permits the creation of simplistic to the most complex queries of the datasets to elucidate the required factors. The results produced are viewed as either numerical values, specific crash reports or visually using the GIS interface. It is recommended that an ‘off the shelf’ system is always chosen above attempting to develop one. The implementation phase of an ‘off the shelf’ system is measured in months whereas developing one is usually measured in years.

The plotting of crash locations on a map is the simplest of these tasks and is provided by all the available ‘off the shelf’ systems. The ability to provide a sophisticated methodology of interrogating the datasets using a graphic interface provides for a more dynamic approach to the analysis process. Such a process provides the ability to define areas on a map to initiate a search or query of the dataset to identify incidents within the selected Regional Participant fitting the predetermined factors.

Cluster Analysis is a complex algorithm that identifies locations that have incidents with common predetermined factors. The results are displayed as a graphic layer over a map with highlighted areas that are colour coded to indicate importance.

Stick Analysis

Stick Analysis is a graphical way of displaying specific datasets associated with a search criterion. It provides a visual representation of the data to enable quick identification of trends or patterns.

Blackspot Analysis

There is no set definition of a blackspot15. It is generally accepted that it is a location that displays higher collision rates than other similar locations due to specific localised risk factor. The analysis to identify these locations is more complex than just identifying numbers of crashes and requires complex algorithms to deduce the plots. Any analysis process to correctly identify blackspots will require a degree of human interface and interpretation however modern analytical procedures are providing significant assistance in identifying such locations. It is essential any proposed system be fully evaluated in this important area to ensure the latest methodologies are being employed. Counter Measures

Having identified a high-risk location modern analysis programs are able to provide a list of the potential countermeasures, based on the initial review process, which could be undertaken to rectify the localised factors identified. Such programs should contain significant and proven counter measure libraries with an option to add further locally designed measures. Cost Benefit Analysis

The ability to rank identified blackspot locations with respect to the potential effectiveness and cost of implementing a countermeasure or intervention provides management with the tools to structure an intervention strategy that best utilised the limited budgets available.

15 http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.138.2070&rep=rep1&type=pdf

Page 143

4.5.11 Recommendations

The below table presents a summary of the actions and recommendations required to be addressed to achieve a common approach to creating a quality crash data recording system. Activity Actions Components An agreement that all the advanced CADaS It is proposed the WB6 Reginal Standardise statistical crash datasets version 3.6 2017 will be adopted Representatives sign a formal agreement to datasets within WB6 regional as the de-facto crash use the Advanced CADaS datasets as a data recording convention. requirement It is proposed a separate statistical crash Enhancement of paper reporting form in line data reporting form should be created to with Advanced CADaS datasets. encompass all the advanced CADaS datasets Use CADaS reference codes as de-facto Translation of advanced CADaS datasets dataset identification and develop an and manual into local language agreed translation for each dataset Elements:  Identification of lead agency responsible for completion of statistical crash report  Identification of agency /officer / department responsible for completing Standardise statistical crash Produce protocol for completion statistical each sections of report. reporting form report form  Maximum time periods permitted when form / sections should be completed  Protocol to quality audit paper form and require lead agency / officer / department to correct omissions and errors identified Develop a reference guide for the completion of the crash report form in the Produce training manual in local language local language based on the manuals that provides an explanation of the CADaS provided by EU. dataset Provide training to personnel on how to completing the crash report form There will be a need to either:  Upgrade the existing database datasets Enhancement of computer database Encoding statistical crash to match statistical crash data form or datasets to enable encoding of new data into computer database  Procure / develop a separate crash statistical datasets database capable of encoding the statistical crash data form

Page 144

Activity Actions Components Elements:  Identification of lead agency responsible for encoding the statistical crash data  Maximum time periods permitted when Produce a protocol outlining the initial record is encoded – within 24 methodology for: hours  Encoding the statistical crash data into  Maximum time periods when record the database. should be completed  Quality audit of data recorded o damage only 7 days  Protocol for marking record as complete o Serious injury 30 days and closed o Fatal defined on an incident by incident bases at least 30 days.  Protocol to identify omissions and errors  Protocol to rectify errors and omissions  Protocol to close a record as complete. Develop a multiagency approach to providing information for the statistical crash It is proposed each Region within the WB6 report form: produce a formal agreement with the  Medical various Ministries to agree on a protocol for o Ambulance the dissemination of data associated with a o Trauma centre road crash.  Fire and Rescue  Identify roles and responsibilities of each  Engineering agency o GIS mapping  Develop a protocol for the provision of o Traffic flow and speed data data. o iRAP data Medical data associated with a casualty involved in a crash is referenced to the Multi-Agency approach Ministry of Health to develop a protocol with casualty’s name and date of birth. the Ministry of Interior (police) with respect The Police database also contains the to the provision of trauma data from names and date of births of the casualties ambulance and trauma centre for a involved in a collision. causality. Proposal is to encode the medical casualty data into the police database to enable easy linking of data. Regional agreement on adopting the It is proposed each Region within the WB6 Maximum Abbreviated Injury Scale (MAIS) formal agree to adopt MAIS injury coding casualty coding system within WB6 Region system. This will require a computer system capable Ministry of Infrastructure to develop a of displaying crash data as an overlay within protocol to provide access to GIS mapping a GIS map Achieve a real time sharing of all the CADaS statistical data sets with all the

relevant Ministries and Road Safety Data Sharing Agencies within a Regional Participant Drafting of a memorandum of An example MOU has been provided in understanding (MOU) between all Appendix C participants The actual exchange of data will require the creation of a compatible computer linking Develop the technical capabilities to database or ‘HUB’ which will facilitate the achieve a secure linking capability for the exchange of data between each of the data Data Linking - local non-sensitive statistical crash data between sources. The central ‘HUB’ database will the relevant Ministries combine the data into records associated with each crash and display the information using advanced GIS technology To achieve this sharing will require:  The drafting of a memorandum of understanding (MOU) (Appendix C) between all the parties involved outlining the exchange mechanism, security Develop the technical capabilities to protocols and timelines. achieve a secure linking capability for the Data Linking - Regional  Each recipient will require a compatible non-sensitive statistical crash data between computer linking database or ‘HUB’ the WB6 Regions and SEETO which will facilitate the exchange of data between each of the data sources.  An advanced analytical database to permit the end user to view and analyse the crash records provided.

Page 145

Activity Actions Components It is proposed that to undertake this style of analysis will require the procurement of an Develop an ability to undertake advanced ‘off the shelf’ Analytical GIS capable crash Data Analysis analysis of the statistical crash data within a database. GIS environment. Many such programs also combine the ability to function as a ‘HUB’ to permit the linking of other data sources.

4.6 Conclusions and recommendations

The core focus of the report is to develop a road map for the sequence of activities that must be undertaken by the WB6 Regional Participants to enable them to achieve a common approach to crash data collection, analysis and dissemination of statistical crash to all interested parties. The core components that must be addressed, in order of priority are:

1. Standardise data collection a. Achieve full compliance with the Advanced CADaS datasets

2. Achieve a multi-agency approach to collecting data to include a. Medical data b. Engineering data, including details GIS mapping 3. Data encoding a. Agreement to encode the key facts of the crash before the investigating officer goes off duty or within 24 hours of the incident. b. Inclusion within the computer record of photographs and sketch plans of the crash scene c. Assign mandatory fields that must be completed within the database record before the initial report can be uploaded. d. Develop a review protocol for the encoded data to ensure there are no errors or omissions present. e. Work towards achieving a basic requirement that 100% of the data fields must be completed before the crash report can be assigned as complete and closed.

4. Data Quality a. Develop a quality assessment protocol to rectify errors and omissions 5. Data sharing within a Regional Participant

Page 146

a. Develop a ‘real time’ sharing capability of all the CADaS statistical data sets with all the relevant Ministries and Road Safety Agencies within a Regional Participants. b. The signing of a memorandum of understanding (MOU) between all the parties involved outlining the exchange mechanism, security protocols and timelines. c. Procurement of a computer linking database or ‘HUB’ which will facilitate the exchange of data between each of the data sources d. Procurement of an advanced analytical and GIS database to permit all the end users the ability to view and analysis the statistical data. 6. Data sharing with WB6 a. It is proposed all the CADaS statistical datasets is shared with neighbouring WB6 Regional Participants and SEETO. To achieve this sharing will require: i. The drafting of a memorandum of understanding (MOU) between all the parties involved outlining the exchange mechanism, security protocols and timelines. ii. Each recipient will require a computer linking database or ‘HUB’ which will facilitate the exchange of data between each of the data sources. iii. An advanced analytical database to permit the end user to view and analysis the crash records provided 7. Data Analysis a. The need to procure an ‘off the shelf’ advanced analytical and GIS based system.

Standardising data The requirement to align to the CADaS advanced dataset will necessitate a review of the paper report used by each Regional Participants to ensure it includes all the elements. It is recommended that an agreed translation by each Regional Participant of these elements is achieved and circulated within the WB6 Region through SEETO. It is recommended the reference number for each element used by the CADaS system becomes the accepted protocol for describing the dataset; this will facilitate the sharing of data.

The review identified all WB6 Regional Participants were in the process of upgrading their paper-based report forms; it was originally proposed this report would provide an assessment of these reports to identify omissions. However, this has not been possible. The report has produced a comprehensive outline of the data requirements together with references which are intended to facilitate the assessment by a Regional Participants of their data collection status.

It is recommended the paper-based approach should ultimately be replaced by electronic means of data collection. The prerequisite to achieving this is the need to have a communication system and computer database to receive the electronic data. Such an approach will greatly simplify the collection process and reduce the work load on the police and other collection agencies.

Achieve a multi-agency approach to collecting data to include

There is a need to collect more statistical data for the most serious injury and fatal collisions. This is based on the requirement to clearly understand the pre-impact approach paths, and the actions involved. Understanding is essential in developing a data led approach to reducing such incidents.

Data encoding

Agreement within WB6 Regional Participants will need to be reached defining the protocol for encoding the paper-based reports into the databases. To achieve a real time data sharing capability between Ministries will require the initial report of the collision to be encoded into the database prior to the investigating officer going off duty on the day of the incident.

Page 147

The inclusion of photographs and sketch plans to complement the statistical data fields will provide significant benefits where analysis of the data is undertaken by other Ministries and agencies that are not able to visit the scene.

Data Quality

There is a requirement for each Regional Participant to develop a quality assessment protocol. This protocol should dictate the time period when quality assessments should be made and the process of rectifying errors and omissions in the data. It will also need to include the process for designating the record as complete.

Where a more serious injury of fatality is involved the initial 7-day review should be undertaken to identify any errors or omissions. However, in such cases the computer record may remain active to allow additions / amendments to the data resulting from further investigations. On such occasions a review process should be repeated every 7 days until either the file is completed or a period of one month (30 days) after the incident has elapsed (the international agreed definition for a fatality).

Data sharing

It is proposed that the initial data sharing activity should be undertaken at a local level between the relevant Ministries and Road Safety Stakeholders. Some members of the WB6 region are already achieving a limited data sharing capability while others only provide a periodic exchange of numerical tables.

The proposal of this report is to achieve a real time sharing of all the CADaS statistical data sets with all the relevant Ministries and Road Safety Agencies within a Regional Participant. This will require, in the first instance, the drafting of a memorandum of understanding (MOU) between all the parties involved outlining the exchange mechanism, security protocols and timelines.

Ultimately, sharing of all CADaS datasets with neighbouring WB6 Regional Participants and SEETO or equivalent body is proposed. To achieve this sharing will require:  The drafting of a memorandum of understanding MOU (see Appendix C for an example of such an MOU) between all the parties involved outlining the exchange mechanism, security protocols and timelines.  Each recipient will require a compatible computer linking database or ‘HUB’ which will facilitate the exchange of data between each of the data sources.  An advanced analytical database to permit the end user to view and analyse the crash records provided.

Page 148

Data Analysis

To achieve more complex analysis, computer programs have been developed to automate the task. These programs are designed with user friendly interface that permits the creation of queries of the datasets, from the simplistic to the most complex, to elucidate the required factors. The results produced are then viewed as either numerical values, specific crash reports or visually using the GIS interface. It is recommended that an ‘off the shelf’ system is always chosen over attempting to develop one. The implementation phase of an ‘off the shelf’ system is measured in months whereas developing one is usually measured in years.

Recommendations

The table in 4.5.11 on page 144 presents a summary of the actions and recommendations required to be addressed to achieve a common approach to creating a quality crash data recording system.

Page 149

APPENDIX A: Relevant Documentation

Page 150

 Road Safety Declaration endorsed by Transport Ministers / representatives of the Western Balkans, meeting in Ljubljana, on 25 April 2018

 Road safety Audits and Inspection – Policy Paper prepared for SEETO members, undated and no author

 SEETO Road Safety Inspection Manual –Common Problems – Shared Solutions (Revised Version 2016), SEETO

 Conclusions and next steps from 8th RSWG meeting of SEETO, April 3-4, Podgorica

 Regulation (EU) 1315/2013. The indicative extension of the TEN-T Network to the Western Balkans Region is articulated in EC Regulation 2016/758 which amended the TEN-T Regulation.

 Building the Transport Core Network: Core Network Corridors and Connecting Europe Facility {COM (2013) 940 final}

 SEETO Multi-Annual Development Plan, Multi-Annual Plan 2018, Common problems – Shared solutions

 Multi-Annual Development Plan – Five Year Multi-Annual Plan 2016, SEETO

 Multi-Annual Development Plan – Five Year Multi-Annual Plan 2016 Update, SEETO

 Monitoring of the Road Safety Strategies in SEETO Members and Draft a Regional Short-term Action Plan – September 2015 - SEETO

 WBIF IPF Bimonthly Report 1801-1802

 Road Safety Principles, COWI for SEETO and DG-ELARG, October 2014

 Road Safety Audit Handbook, COWI for SEETO and DG-ELARG, September 2014

 Information on new/rehabilitation projects (subject to RSA): o The Bosnia and Herzegovina comprehensive list but missing costs o former Yugoslav Republic of Macedonia project lists provided but missing costs/financing institution or implementation plan o Montenegro – comprehensive list of planned projects provided but no indicative costs o Serbia a comprehensive list but missing some cost information.  Information relevant to High Risk road sections for RSI, per Regional Participant:  Albania – high risk road sections presented on maps  Bosnia and Herzegovina - complete information  The former Yugoslav Republic of Macedonia - from place (node) to place (node), totalling 224 km.  Serbia - in Serbian language, start and end points (X and Y) of road sections. Please note that, the points mentioned can't be identified, since they are not on a universal Longitude and Latitude system.  Information relevant to the RSI completed the last years or planned for the next years as follows, per Regional Participant: o Bosnia and Herzegovina - planned RSI o the former Yugoslav Republic of Macedonia - past and future RSI o Kosovo - RSI completed o Serbia - RSI plans

Page 151

 Maps on Comprehensive and Core Road Network  Road Report, SEETO Comprehensive Network  Road Safety Working Group Members  SEETO Road Safety Strategy Survey, Final, SEETO, March 2014  SEETO Road Safety Audit Implementation Report, Final, SEETO, July 2012  DIRECTIVE 2008/96/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL, of 19 November 2008 on road infrastructure safety management  PIARC Recommendation regarding the approach to take into account Directive 2008/96/EC in the context of road tunnels, PIARC World Road Association, 6/12/2013  DIRECTIVE 2004/54/EC OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 29 April 2004 on minimum safety requirements for tunnels in the trans-European road network  Commission Delegated Regulation (EU) 2016/758 of 04.02.2016

Page 152

APPENDIX B: Crash Report Forms

Page 153

Crash Data Report forms

1. Albania

2. Bosnia and Herzegovina (proposed generic form only provided)

3. Kosovo

4. The former Yugoslav Republic of Macedonia - no form received

5. Montenegro – no form received

6. Serbia – No form received.

Page 154

Albania

The below table reproduces the datasets from the crash report form created by the SweRoad project 1. Accident number 2. Date 3. Time 4. District 5. Location – free text 6. Road number 7. GPS coordinates 8. Reporting police officer 9. Collision type 1. Frontal 2. Rear 3. Side Impact 4. Overtaking 5. On a Curve 6. Overturned Vehicle 7. Hit Water 8. Collision with Other Vehicles 9. Collision with Pedestrian 10. Collision with Cyclist 11. Collision with Animal 12. Collision with Other 13. Collision Between Cyclist and Pedestrian 14. Other 10. Road Geometry 1. Straight road 2. Curve 3. Roundabout 4. T junction 5. Y junction 6. Cross roads 7. Staggered junction 8. Bridge 9. Railway crossing 10. Tunnel 11. Surroundings 1. Rural 2. Urban 12. Surface 1. Bitumen 2. Gravel 3. Soil 13. Road Condition 1. Good 2. With holes 3. Uneven surface 4. Slippery 5. Dusty 14. Weather 1. Dry 2. Raining / wet

Page 155

3. Snow / ice 4. Fog 15. Other factors 1. Stolen vehicle 2. Hit and run 3. Road works 16. Speed limit 17. Lighting condition 1. Day 2. Night 3. Dawn / dusk 18. Driving license 1. Learner 2. Holder 3. Non-holder 19. Driving experience 1. Below 3 years 2. 3-6 years 3. 6-9 years 4. Over 9 years 20. Drivers Nationality 1. Albanian citizen 2. foreigner 21. vehicle type 1. Car 2. Mini bus 8+1 3. Mini bus >8+1 4. Auto bus 5. Wagon < 3.5 T 6. Sports vehicle 7. Heavy truck > 3.5T 8. Heavy truck + trailer 9. Motorcycle 10. Animal drawn vehicle 11. Bicycle 12. Farm tractor 13. Other 22. Vehicle defects 1. Unidentified 2. Steering / front axle 3. Brakes 4. Wheels / tyres 5. Windscreen 6. Limited visibility 7. Overloaded 8. Other 23. Registration 1. Registered 2. Unregistered 3. Yearly inspection valid 4. Yearly inspection expired 5. Insured 6. Uninsured

Page 156

24. Vehicle age 1. 0-2 years 2. 2-7 years 3. 7-12 years 4. Over 12 years 25. Drivers behaviour 1. Fail stop at red light 2. Fail to give way 3. Ignored traffic sign 4. Careless overtaking 5. Careless driving 6. Careless turn 7. Unexpected change of direction 8. U turning with negligence 9. Sudden braking 10. Driving too close to vehicle ahead 11. Dangerous parking 12. Excess speed 13. Blinded by sun/ headlights 14. Changing lane 15. Fatigued 16. other 26. Alcohol test 1. No alcohol 2. Excess legal limit 3. No test 27. Pedestrian behaviour 1. Unknown 2. Crossing the road without care 3. Pedestrian crossing 4. Crossing the road other location 5. Walking on road 6. Other 28. Category of involved people (1-10) 1. Driver 2. Passenger 3. Pedestrian 4. Cyclist 5. Motor cyclist 6. Other 29. Number of vehicle in or hit by. 30. Gender 1. Male 2. Female 31. Age 32. Injury type 1. Fatal 2. Serious 3. Minor 4. Injury 33. Seatbelt / helmet in use 1. Yes 2. No.

Page 157

Federation of Bosnia and Herzegovina proposed crash report form

Page 158

Kosovo

The following are the datasets included in the paper-based form used by the police Accident No. Date of Accident Time accident Location GPS coordinates Number of Injury Deaths 1A. Other damage Name of Injured Police witness 1 Road Class 1 One lane 2 Two lanes 3 Three lanes 4 Four lanes 5 Five lane 6 Six lane 7 Seven lane A Divided B Undivided C Ramp 1A Traffic Flow 1 One-way traffic 2 Two-way traffic 2 Accident Location 1 At intersection 2 Between Intersections 3 Intersection of Road & driveway or alley 4 Bridge 5 Ferry or dock 6 Tunnel 7 Exit deceleration lane 8 Exit ramps 9 Exit intersections 10 Exit acceleration lane 11 Entrance ramps 12 Entrance intersection 13 Off highway 14 Parking lot single / multi-level 15 Rail Road Crossing 16 Industrial road 17 Transit express lane 2A Speed zone A 10 Km/h B 20 Km/h C 30 Km/h D 40 Km/h E 50 Km/h F 60 Km/h G 70 Km/h H 80 Km/h I 90 Km/h J 100 Km/h

Page 159

K 110 Km/h 3 Land usage in accident area 1 School / playground 2 Urban residential 3 Apartment residential 4 Business / shopping 5 Industrial / manufacturing 6 Agricultural / undeveloped 7 Recreational / park / camping 8 Rural residential 4 Road Type 1 Asphalt 2 Gravel 3 Oiled gravel 4 Earth 5 Concrete 6 Brick / stone 7 Wood 5 Traffic control 1 None 2 Stop sign 3 Yield sign 4 Officer / flagman / school guard 5 Rail road crossing sign 6 Lane use / turn control sign 7 Traffic signal 8 Traffic signal with advanced flasher 9 Flashing signal 10 Lane use signal 6 Road way character Horizontal 1 Straight 2 Single curve 3 Sharp curve 4 Switchback 5 Winding curves 6 Reverse curves Vertical 7 Flat 8 Some grade 9 Steep grade 10 Hill crest 11 Sag 7 Road Surface condition 1 Dry 2 Wet 3 Muddy 4 Snow 5 Slush 6 Ice 8 Weather condition 1 Clear 2 Cloudy 3 Raining 4 Snowing / sleet 5 Hail 6 Fog

Page 160

7 Smog / smoke 8 Strong wind 9 Lighting condition 1 Day light 2 Dawn 3 Dusk 4 Dark full illumination 5 Dark no illuminations 6 Dark some illumination 10 Safety Belt not used vehicle 1 1 Driver 2-100 Passenger 11 Safety Belt not used vehicle 2 12 Location of first contact A On roadway B Off roadway 13 Pre-collision vehicle action vehicle 1 14 Pre-collision vehicles action vehicle 2 1 Going straight ahead 2 Making right turn 3 Making left turn 4 Making 'U' turn 5 Starting from parked position 6 Starting in traffic 7 Slowing or stopping 8 Stopped in traffic 9 Entering parked position 10 Parked legally 11 Parked illegally 12 Avoiding object on road 13 Changing lanes 14 Overtaking 15 Merging 16 Backing 17 Skidding 18 Swerving 19 Spinning 20 Jack-knifing 21 Yaw 15 Type of accident collision 1 Other motor vehicle 2 Motorcycle 3 Pedestrian 4 Bicyclist 5 Animal 6 Rail road train 7 Street car /Trolley coach 8 All-terrain vehicle 9 Mope under 55 cc 10 Guard rail / traffic barrier 11 Crash cushion / impact attenuator 12 Sign post 13 Tree 14 Building / wall 15 Curbing 16 Fence 17 Bridge deck or parapet

Page 161

18 Snow bank / drift 19 Rock face 20 Ditch 21 Culvert 22 Fire Hydrant debris 23 Rocks ore 24 Over turned 25 Fire / explosion 26 Lake / river 16 Vehicle type 1 17 Vehicle type 2 1 Passenger car 2 Car and trailer 3 Single unit truck light 4 Single unit truck heavy 5 Combination unit truck light 6 Combination unit truck heavy 7 Combination unit tractor trailer 8 Tractor 9 Log truck & pole trailer 10 Bus school 11 Bus local 12 Bus intercity 13 Motorcycle 14 Moped 50 cc 15 Bicycle 16 Truck 17 Truck trailer 18 Motorcycle 19 Motor trailer 20 Special 16A Vehicle 1 usage 17A Vehicle 2 usage 1 Parked 2 Personal 3 Business 4 Driver training facility 5 Recreational 6 Emergency 7 Military 8 Taxi 9 Farm use 10 Government 11 Towing / towed 12 stolen 18 Pedestrian Location 1 At intersection 2 Not at intersection 19 Pedestrian actions 1 Crossing with signal 2 Crossing against signal 3 Crossing no signal marked crosswalk 4 Crossing no signal no crosswalk 5 Walking along highway with traffic 6 Walking along highway against traffic 7 Emerging from front / behind parked vehicle 8 Child getting on / off school bus / vehicle

Page 162

9 Adult getting on / off a vehicle 10 Pushing / working on a car 11 Working in roadway 12 Playing in roadway 13 Standing on sidewalk Apparent contributing Factors 20/21/22 Vehicle / driver 1 23/24/25 Vehicle driver 2 1 Alcohol involvement 2 Backing unsafely 3 Cutting in 4 Driving without due care 5 Driver inexperience 6 Drugs illegal 7 Extreme fatigue 8 Failing to signal 9 Failing to yield right of way 10 Fell asleep 11 Following too close 12 Improper passing 13 Illness 14 Sudden loss of consciousness 15 Driving on wrong side of the road 16 Pedestrian error / confusion 17 Pre-existing physical disability 18 Prescribed medication 19 Attempt to commit suicide 20 Ignoring traffic control device 21 Improper turning 22 Unsafe speed 23 Ignoring officer / flagman / guard 24 Avoiding vehicle / pedestrian / cycle 25 Accelerator defective 26 Brakes defective 27 headlights defective 28 Brake light out 29 Turn signal defective 30 Oversize vehicle 31 Steering vehicle 32 Tres failure / inadequate 33 Tow hitch failure 34 Driverless vehicle 35 Windshield defective 36 Engine failure 37 Suspension failure 38 Restraint system 39 Insecure load 40 Dangerous goods 41 Vehicle modification 42 Glare artificial 43 Glare sunlight 44 Obstruction / debris on road 45 Pavement surface defective 46 Visibility impaired 47 Weather 48 Road 49 Previous traffic accident 50 Sign obstruction 51 Domestic animal

Page 163

52 Wind animal 53 Insufficient traffic control 54 Road / intersection design 55 Roadside hazard 56 Other Vehicle 2 Insurance Co Name and address 2A policy No. Green Card No. Policy expiry date 3 vehicle type make model year Plate No. colour Vin No. Regional Participant Driver 6 DOB Gender Height Weight Hair Eyes Alcohol % Driving License Category Regional Participant Issue Damage Diagrams 8 Damage Motorcycle 9 Damage to car 10 Damage to bus 11 Damage to truck Accident description Other actions Charges Report date Officer ID No. Station

Page 164

APPENDIX C: Questionnaire

Page 165

The below questionnaire was submitted to each of the Regional Representatives requesting them to provide feedback prior to the field visit undertaken by the review team.

Information required

The explanation within the introduction provides an overview of the requirement and methodologies currently available for the collection of accident data. The following sections provide guidelines to the subject areas that should be addressed. The underlying requirement is to accurately describe the current status, within each the West Balkan Regional Participants, with respect to their accident recording and analysis capabilities. Each of the subjects covered will require documentary evidence to support the explanations and findings presented.

All meetings undertaken will require to be documented; they will also need to include the names and contact details of people spoken to. Where copies of documents are obtained the person providing and the source of the document must be recorded.

Legal statute that defines a road traffic accident

Within the traffic law there is usually a section that defines what constitutes a road traffic accident example:  It involves a mechanically propelled vehicle  On a road or other public place  Involves damage to property or injury.

Information required  A copy of the Act and section that provides this definition should be obtained and translated into English.

The legal requirements on the driver Within the traffic law there is usually a section that defines the actions required to be taken by a driver of a vehicle involved in a road traffic accident to report it to the appropriate authority.

Information required  A copy of the Act and section that provides these requirements should be obtained and translated into English.

The legal responsibility for the recording of accidents

Within the traffic laws there is usually a section that defines that the police have to investigate and record road traffic accidents.

Information required  A copy of the Act and section that provides this requirement should be obtained and translated into English.

Traffic accident reporting methodology.

The requirement is to identify what information is collected by the police following the initial report that a road traffic accident. This starts with the first communication that is received by the emergency

Page 166

services informing them of an incident through to the last police unit resuming from the scene of the incident.

Information required:  This should include the identification of where a single or multiple emergency phone number is directed to and what information is collected by the call handling centre.  How the report is passed to the individual emergency services and the protocols that dictate their response. The primary emergency services will be the ambulance, fire and rescue and the Police.  Information and examples of what incident logs are produced by each agency with specific reference to time and locational information of the initial call, allocation of the units that are to respond, their arrival times and incident report.  It is requested that detailed enquiries should be undertaken with the ambulance services to identify what additional data they collect with respect to the medical condition of the patients conveyed to hospital.

Accident scene investigation

The severity of the accident will dictate the rank of the police officer leading the investigation. A detailed understanding of the police protocols with respect to who investigates the incident and who is responsible for completing the accident report form.

Information required  Information on the police protocols with respect to who investigates the accident  Who is responsible for completing the accident report form  Existence of a dedicated traffic department  Existence of forensic accident investigation capabilities

Police accident report form

The primary source of information is that obtained at the scene of the accident by the investigating police officer. This is achieved by either a:  Totally paper based recording system  Combination of paper and electronically based system  Completely electronically based system

Information required:  The methodology currently in use to record the accident details  A copy of the paper-based form  A description of the computer-based fields required to be completed. This must include those fields that are mandatory before the record can be uploaded into the database.

Severity classification of the accident

The severity classifications are usually fatal, seriously injured, slightly injured and damage only.

Information required  Details of their classification  Details on who decides the classification o Police or

Page 167

o Medical personnel  The legal requirement to classify an accident as fatal: o Death at the scene only or o Death within 30 days of the accident due to the injuries sustained from the accident

Police accident database

This is the primary database used by the police to record all road traffic accident reports and the statistical data collected at the scene of the accident by the investigating officer.

Information required  The name and technical description of the database  The methodology currently in use to record the accident report form on the computer – input by investigating officer, dedicated police officer entering all accident reports, civilian operator copy typing the data into the database.  The quality audit protocols used to check the quality of the data recorded.  The format of the accident reference number used in the computer system  Information on the protocols for the recording of the accidents details onto the computer; for example, within 24 hours, within 7 days or longer  An assessment and description of the fields that are on the accident report form / designed within the database but consistently have missing data.  Soft copy export in excel format of non-sensitive (no identifying names address vehicle numbers etc.) data held within the data base for 2017

Analysis undertaken by the police

A comprehensive description of any processes undertaken by the police to analysis the accident data.

Information required  Description of the police unit responsible for the work  A description of their role and responsibilities  The computer programs they use to undertake the analysis  Obtain an example of their work, soft copies of any reports they have generated  Copy of yearly statistical reports  What are the main causes of accidents  What are the main causality class for serious and fatal accidents

Casualty injury profile

The requirement is to identify what additional data is available from the ambulance or medical facilities within a Regional Participant that treat casualties involved in road accidents. The information required covers information concerning the response time of medically trained personnel to the accident scene, an assessment of the medical condition of the causality at the scene, the arrival time of the casualty at a medical facility and their medical condition on arrival.

Information required  The medical qualifications of the ambulance personnel  The equipment available in a front-line ambulance  A description of the data collected by the ambulance personnel – copy of any form used.  Information on

Page 168

For any additional information or queries with respect to this questionnaire please contact: Mike Fell Road Safety Data Specialist Email: [email protected]

Page 169

APPENDIX D: Memorandum of Understanding

Page 170

Example of MoU that has been developed and accepted internationally by many Police and Ministerial organisations as appropriate. The exact wording should be altered to address local legal requirements and protocols.

Memorandum of Understanding

This Memorandum of Understanding (MoU) is made and entered:

By and Between

Ministry of Infrastructure, having its registered office at hereinafter referred to as “MOI*” (which expression shall, unless repugnant to the context or meaning thereof, be deemed to mean and include its successors and assigns) And

Ministry of Internal Affairs, having its registered office at , hereinafter referred to as “MOIA” (which expression shall, unless repugnant to the context or meaning thereof, be deemed to mean and include its successors and assigns) And is effective from the ______June 2018

Whereas Ministry of Infrastructure is the lead road agency for ***** and Ministry of Internal Affairs is the main law and order agency for the ******. The attention and focus on improving road safety in ***** has increased greatly in the last few years. Key amongst road safety developments in ***** is the development of the **(National Road Safety Strategy) **. The strategy commits **** to attain significant future reductions in fatal and serious injuries through development of a “Safe System” road transport network. Key to supporting the Safe System is the extensive use of crash and allied data to guide road safety strategies and actions to ensure they are efficient and effective; the ability to monitor and evaluate road safety performance is also a key requirement. The purpose of this MoU is to finalise the terms for transfer of crash data from Ministry of Internal Affairs to Ministry of ******. In furtherance of the cooperation, it is hereby agreed between MOIA and MOI to collaborate and work together under the following terms: 1. A secure web service link will be created by MOI for the transfer of crash data from MOIA system to MOI crash analysis system. MOI shall share the credentials and URL for accessing the web service. 2. MOIA will remove personally identifiable fields before exporting the data to MOI via the secure web service link; namely all: a. Names and Addresses, b. Vehicle Plate Numbers, c. Individual ID Numbers, d. Driver License Details, e. Insurance Policy Details. 3. Crash data will be transferred from MOIA System to MOI Crash Analysis System at 23:30 hours daily: a. The first transfer will require all the MOIA crash data records available electronically prior to the first transfer date. b. For all subsequent transfers, MOIA will transfer all:

Page 171

i. New Crash Records for the day ii. Altered or Amended Crash Records for the day. Reporting Requirements and Governance of the data transfer process.

I. In case where new, amended and or updated crash records are not transferred the Manager of ********** shall review and inform the matter to MOIA General Manager Traffic Department with the detailed pendency report. The pendency report will be manually submitted to the office of MOIA Director of National Traffic Department. II. In case if the web services are not accessible, the Manager of ****** shall review and inform the matter to MOIA Director of IT with the detailed incident report. The incident report will be manually submitted to the office of ***. III. MOI will produce and submit a quarterly data quality analysis report to the Chairman of the ***********and MOIA. MOI will submit quarterly action taken report to improve the data quality. This MoU will remain in effect for 5 years from the date of signature of this document, unless extended by mutual agreement of the parties. This MoU shall be governed by, construed and interpreted in accordance with the laws of Ukraine and all disputes and proceedings shall be subject to exclusive jurisdiction of the Courts. This MoU is duly agreed and executed by both parties on this the ___ day of August, 2018.

For For

Ministry of Infrastructure Ministry of Internal Affairs

______

Witness Witness

______

Page 172

APPENDIX E: Example of medical forms

Page 173

Example medical forms used by Trauma care teams

Page 174

Page 175

Example of ambulance crash report form

Page 176