Behavioural Analysis and Modelling for the Design and Implementation of Advanced Driver Assistance
Much R&D is focused on an increasing variety of advanced driver assistance systems (ADAS) to be introduced in road traffic. These systems involve the assistance and/or automation of various basic driving tasks (vehicle following, lane keeping, lane changing, proper speed keeping) based on modern technology for sensing, data processing, data transmission, operational decision making and task actuation. The full and integrated application of these systems implies full automation of vehicle driving, called 'automated vehicle guidance' (AVG).
Research in the past, indicates a limited applicability of most driver assistance systems (mainly motorways, dedicated lanes). This is mainly due to complex design issues related to the development of these systems. The complexity of design is caused by the rich variety of infrastructure and traffic situations. These systems should be able to handle in a reliable and accurate way the variety in driver behaviour that is generated by this variety in traffic situations. The design of such systems requires a (traffic) system approach which is opposite to the technology driver approach.
Research indicates a lack of sufficient conceptual and empirical knowledge about this behaviour. Improvement of this knowledge is necessary to better develop efficient and effective driver assistance systems. Moreover, better knowledge helps to improve the evaluation of the performance of these systems, the necessary certification and the arrangement of liability for system development, system application and system failure. The research programme continues scientific research that has been performed within TRAIL in the past 5 years. See: www.trail.tudelft.nl/
The aim of the programme:
- Improving the theoretical, and empirical and design knowledge regarding road vehicle driver behaviour in interaction with advanced driver assistance systems;
- The transfer of this knowledge to deployment strategies for these systems focusing on infrastructure design and traffic management;
- The improvement of the knowledge regarding system certification and liability regulation in a multi-actor environment.
BAMADAS is a research programme that is directed by Prof.dr.ir. R.E.C.M. van der Heijden (University of Nijmegen) and coordinated by Dr.ir. V.A.W.J. Marchau (Delft University of Technology). The six sub-projects are conducted by separate research teams that consist of senior staff and PhD researchers of Delft University of Technology, University of Groningen, University of Nijmegen, University of Leiden and SWOV Institute for Road Safety Research.
Budget is EUR 1.4 million. NWO-Connekt contribution: EUR 550 000.
In order to achieve these goals, six interrelated sub-projects have been distinguished, which represent the major scientific challenges of this programme. These challenges refer to a scientific mix of innovations in theory-building, methodology development and empirical research.
The way of thinking is: improved knowledge on micro-behaviour results in a better view on the terms of reference for design and deployment of ADAS (what and how). Next, it provides a structured view of performance standards and testing programmes for micro-behaviour, including liability regulation among stakeholders (users, automotive industries and road managers) for cases of failure. Finally, systems that have been designed and implemented, have impacts on traffic flows and infrastructure design. To study these impacts, traffic flow models are adapted and infrastructure (re)design programs are evaluated.
The six sub-projects are:
- ASTIM: Advanced Safety Criteria Specification by Traffic Interactions Modelling
- TOMAS: Testing Operational Models and Behavioural ASsumptions Included in Driving
- ROTAS: Modelling ROad Traffic Patterns Using Advanced Driver Assistance Systems
- MIDAS: Matching Infrastructure Design with Advanced Driver Assistance Systems
- RULES: Regulating the Use of ADAS: Liability and Legislation Aspects of Electronic Driver Support
- SPACE: SPAtial development and Concepts of Electronic transportation systems