The application of systems approach for road safety policy making, Deliverable 8.1 of the H2020 project SafetyCube
reportposted on 13.01.2017, 11:43 by Ashleigh Filtness, Pete Thomas, Rachel Talbot, Claire Quigley, Eleonora Papadimitriou, George Yannis, Athanasios Theofilatos, E. Aigner-Breuss, S. Kaiser, Klaus Machata, Wendy Weijermars, Ingrid van Schagen, T. Hermitte
The present Deliverable (D8.1) describes the co-ordination of the analysis of risks and measures using a systems framework within the SafetyCube project. It outlines the results of Task 8.1 of Work Package (WP) 8 of SafetyCube. This has involved (i) defining the systems approach to be used within SafetyCube, (ii) developing a taxonomy of risks and measures, (iii) identifying a common set of accident scenarios and (iv) initiating work on the Decision Support System (DSS) development. WP8 of the SafetyCube project has a number of specific aims, including developing the European DSS for supporting evidence-based policy making. It also aims to co-ordinate analysis undertaken in other WPs ensuring integrated research outputs, compilation of the project outputs into a suitable form to be incorporated within the DSS and the European Road Safety Observatory, and finally to develop tools to enable the continued support of evidence based road safety policies beyond SafetyCube. Evidence-based policy making enables policy makers to make justified decisions in the complex reality of road safety interventions. It refers to the use of objective, scientifically-based evidence in all stages of the policy making process. Two important pillars for evidence-based road safety policy making are road safety data and statistics and scientific knowledge (Wegman et al, 2015). This type of policy making can be beneficial (e.g. helps to identify road safety problems and select most appropriate interventions) but also has it’s challenges (e.g. a lot of information at varying levels of detail is required to inform decisions). The DSS that is being developed within SafetyCube aims to support decision makers as well as other stakeholders in their evidence-based policy making. In addition to evidence-based policy making, SafetyCube and in particular the DSS is grounded in the systems approach. The systems approach aims to steer away from the more traditionally ‘human error’ blame focussed approach to road safety, and instead takes into account all ‘components’ in a system (i.e. road users, vehicles, roads) which contribute to a risk of an accident occurring. In SafetyCube, the systems approach is being integrated in the DSS in two main ways. First, the risk factors which relate to the road user, the road or the vehicle will be linked to measures in any or all of these areas if appropriate. Second, to clarify the added value of complementary measures rather than measures in isolation, where appropriate, a description of a measure will pay special attention to and link to supporting measures. The SafetyCube DSS is underpinned by four taxonomies; Road User Behaviour (WP4), Infrastructure (WP5), Vehicles (WP6) and Post Impact Care (WP7). The taxonomy is a main structural part of the DSS system, it can be used as a search option in the DSS, it creates a uniform structure over all work packages and it can be used as a basis for linking risk factors with their corresponding measures. The structure consists of three levels, which are topic, subtopic and specific topic. Thirteen main topics were identified for Road User Behaviour (WP4), 10 main topics for Infrastructure (WP5) and six main topics for Vehicle (WP6). Four topics (based on the DaCoTA webtext on Post Impact Care, 2012), were included in WP7 (Post Impact Care). As expected, there was found to be some overlap between risk factors in one taxonomy and risk factors in another (e.g. is poor vehicle maintenance a Vehicle or Road User-related risk factor?), and some overlaps where a topic could be a risk factor or a countermeasure. Discussions between WPs ensured decisions could be made about how to overcome these ambiguities. Accident scenarios are used within SafetyCube. These are considered to be a classification system for crashes whereby crash types may be grouped according to similar characteristics under a particular scenario heading, creating specific clusters. In total, nine high level accident scenarios will form an entry point to the DSS. Each high level has multiple sub-levels which provide more detailed information about the conflict situation (before the crash). A total of 63 sub-level scenarios are considered. The task of linking risks and measures is currently underway within the SafetyCube project. The accident scenarios will provide a useful and systematic way by which to link risks and measures. They will be used, in order to generate a meaningful set of links, between risks related to specific situations, and measures to address them. The primary objective of the DSS is to provide the European and Global road safety community a user friendly, web-based, interactive Decision Support Tool which will enable policy-makers and stakeholders to select and implement the most appropriate strategies, measures and cost-effective approaches to reduce casualties and crash severity for all road users. It consists of information such as risk factors, road safety measures, cost-benefit, casualty reduction effectiveness estimates. In order to develop the DSS, a review of current existing Decision Support Systems was carried out to provide a first insight into such tools (e.g. Crash Modification Factors Clearinghouse, PRACT Repository, Road Safety Engineering Kit, iRAP). No European DSS were found in the search and of the DSS reviewed, the majority focussed on infrastructure and no risk factors were included. The SafetyCube DSS addresses these gaps. To understand user needs better, three stakeholders workshops were carried out, which allowed participants to comment on the proposed DSS and suggest ‘hot topics’ (i.e. important risk factors) to address in SafetyCube, and the findings of these workshops found that the DSS should be suitable for use by a wide range of users, should be impartial, include robust data and access to all studies used and generated results. A comprehensive common SafetyCube methodology was designed, which included: a complete taxonomy of human behaviour, infrastructure and vehicle; a detailed and recorded literature review and the development of a template for coding research studies and existing results to be stored in a database linked to the DSS. The DSS is being created on the basis of a number of design principles (e.g. modern web-based tool, ergonomic interface, simple, easily updated…). As well as a consistent layout the content itself is also of high importance (e.g. quantitative results over qualitative, methodologically sound, clarity). The DSS itself consists of the backend (relational database), the front end (website) and the way they integrate (queries). The heart of the DSS consists of the searchable/dynamic and static aspects, which consists of five entry points and three levels. The design principles of the DSS ensure a smooth integration of the Work Packages in two ways, firstly that the SafetyCube common methodology is applied and secondly that the fully linked search allows the end user to better perceive the interactions between various components in road safety. There are five entry points into DSS: ‘text search’, ‘risk factors’, ‘road safety measures’, ‘road user groups’ and ‘accident scenarios’. Once a search has been undertaken using one of these five entry points, a results page is shown to the user, which consists of a table listing the available synopses1 (overview of the topic created by synthesising findings from the coding of existing studies), meta-analysis and other studies in the database. From this, the user can then also access the individual study pages for each study listed in the results. Finally, a Tools page allows the user to access other SafetyCube tools (e.g. cost-benefit calculator, methodology information, glossary). 1 More details about the synopses can be found in the Milestone M3.1 (Martensen 2016). So far, more than 500 studies have been analysed in the area of road risks with more than 3,500 risk estimates, summarised in more than 60 synopses (including approximately 10 meta-analyses), and the related measures analyses are in progress. This wealth of information will all be incorporated into the DSS and become its core outputs. The overall design of the DSS is finalised and is currently available, with the next stage being the DSS development, including all risk factors and measures. The DSS Pilot Operation will occur later in the project, followed by the final opening of the DSS, with continual updates from the end of the project onwards. The SafetyCube DSS is intended to have a life well beyond the end of the SafetyCube research project.
This work was co‐funded by the Horizon 2020 Framework Programme of the European Union (Grant agreement No 633485 ‐ SafetyCube ‐ H2020‐MG‐2014‐2015/ H2020‐MG‐2014_TwoStages)