Improved Trustworthiness and Weather-Independence of Conditionally Automated Vehicles in Mixed Traffic Scenarios

Automated vehicle technology has the potential to be a game changer on the roads, altering the face of driving as we experience it by today. Many benefits are expected ranging from improved safety, reduced congestion and lower stress for car occupants, social inclusion, lower emissions, and better road utilization due to optimal integration of private and public transport. Many cars sold today are already capable of some level of automation while higher automated prototype vehicles are continuously tested on public roads especially in the United States, Europe, and Japan. Automated vehicle technology has arrived rapidly on the market and the deployment is expected to accelerate over the next years. As a matter of fact, most of the core technologies required for fully automated driving (SAE level 5) are available today, however, reliability, robustness, and finally trustworthiness have to be significantly improved to achieve end-user acceptance. System and human driver uncertainty pose a significant challenge in the development of trustable and fault-tolerant automated driving controllers, especially for conditional automation (SAE level 3) in mixed traffic scenarios under unexpected weather conditions. The TrustVehicle consortium gathers European key partners who cover the entire vehicle value chain and form a European eco-system: OEMs, Tier1 suppliers, semiconductor industry, software, engineering, and research partners to enhance safety and user-friendliness of level 3 automated driving (L3AD) systems.

TrustVehicle aims at advancing L3AD functions in normal operation and in critical situations (active safety) in mixed traffic scenarios and even under harsh environmental conditions. TrustVehicle follows a user-centric approach and will provide solutions that will significantly increase reliability and trustworthiness of automated vehicles and hence, contribute to end-user acceptance. The main fields of research and innovation will cope:

  • Intrinsic self-diagnostics of sensors and systems (software and hardware layer);
  • Increased real-time detection accuracy of VRUs in urban areas by at least 3% compared to the current state-of-the-art due to the introduction of infrared-based time-of-flight cameras along with embedded machine-learning;
  • Fail-operational (fault-tolerant) system behaviour and controller design supporting both driver-in-the-loop and driver-off-the-loop scenarios by introducing 24/7 available hardware and software redundancy (optimal use of complementary sensor modalities, robust in-vehicle sensor fusion, fault-tolerant control algorithms);
  • Managing driver failures by the system due to the continuous calculation of all possible safe trajectories (“online safety corridor prediction”) using a many-core platform in order to warn the driver and to trigger “machine takeover scenarios”;
  • Supporting the driver with novel HMI technologies (redundancy due to gesture, voice, touch, knob handling) able to detect hand movements through infrared cameras, thus providing corresponding reaction and vocal feedback for each action while improving user’s acceptance among different populations;
  • Toolchain-based evaluation and assessment of L3AD functions with respect to reliability.