Développement d’outils à l’évaluation de la robustesse et de la fiabilité des véhicules intelligents et autonomes pour un transport routier sécuritaire

Développement d’outils à l’évaluation de la robustesse et de la fiabilité des véhicules intelligents et autonomes pour un transport routier sécuritaire

The purpose of this project is to bring Quebec to a leading position worldwide in the field of testing, certification and homologation of intelligent vehicles, to help saving lives on our roads and to decrease the impact of road traffic on the environment. Much hope arises by the advent of information technologies embedded in modern vehicles. Advanced Driver Assistance Systems (ADAS), connected vehicles and automated driving functions are rapidly being developed and deployed. However, this rapid development has also raised a growing need to cope with the performance evaluation and effectiveness of these highly complex systems. Before we can put these new intelligent vehicles on our roads, we need to test them thoroughly to make sure the safety benefits are not penalized by unpredicted or dangerous system behaviour. Indeed, road safety is directly related to the reliability and the robustness of the embedded sensors and systems. The testing and validation processes must evolve to cope with this increasing complexity and new innovative solutions need to be found. To evaluate the performance of processing and decision-making algorithms in intelligent vehicles, we need to develop and use an extensive suite of simulation and testing tools that can jointly operate in real time or accelerate time and match closely real vehicular and driving scenarios. Hence, this project focuses on the development of simulation and physical test protocols as well as the exploitation of field operational tests (FOT) data. For this purpose, new algorithms, mathematical models and software tools are required for extracting, processing and analyzing the corresponding huge amount of data. The vision of this project is based on the concept of « extended » testing procedures for intelligent vehicles involving a « physical » part with instrumented vehicles on test tracks, open databases gathering the data from field operational tests (FOT) on the roads and a part based on modelling and simulation of virtual multi-vehicular scenarios. It is worth noting that those components are all interrelated and must be conducted in parallel while considering feedback loops to enhance one another method, thus reflecting the data exchanges used to improve the testing method performance. Keywords: Vehicular technologies, Automotive and road safety, Intelligent transportation systems, Sensors and complex systems, ADAS, Crash test protocols, testing intelligent vehicles, autonomous vehicles, automated driving, connected vehicles

Retour haut de page