The aim of this VeriCAV sub-project, was to build a Comfort Algorithm that would sit within VeriCAV’s Test Oracle and predict the occupant Comfort performance characteristics of an ADS. Connected Places Catapult designed and built a comfort algorithm, based on scientific literature. Comfort parameters were selected to be objective and tailored to the context of Autonomous Vehicles. A scoring system was devised to rate the comfort performance of ADS within simulations which sat within the VeriCAV Test Oracle. The same scoring system was designed so that it could be used within the user trials to verify the comfort algorithm’s accuracy and relevance to user trial participant’s comfort experience.
The VeriCAV project has been developing solutions to improve the evaluation of Automated Driving Systems in simulation. Our final summary report outlines approaches to: building scenarios more easily, featuring richer scenarios with intelligent behaviour in other road users, and automating the process of evaluating the driving performance. This report will be of interest to anyone in the automated vehicle community who wants to evaluate automated driving performance more efficiently and requiring less time spent per test scenario.
November 2020 VeriCAV overview presentation for 6th Symposium Driving Simulation (SDS) held on 5th November 2020. The SDS event gathers international specialists from science and industry in the field of driving simulation and autonomous driving.