In order to validate that the BRAVE concepts and prototypes meet the user requirements of individual drivers, other road users and any other stakeholders, it was necessary to develop and assess the methodologies to test and evaluate the use cases considering pedestrian and other vehicles movements and intentions. UTAC CERAM and UAH worked together in order to test the realised prototype in a safe environement. Recommandations were made by UTAC CERAM to UAH regarding their prototype so that it could become even more refined for the final demonstration in Barcelona.
Firstly, a primary indication of the automated vehicle’s performance was provided within the February 2020 test campaign and it helped UAH to fine-tune its behavior. Then, the automated car was on-track tested within the July 2020 final test campaign, within a Euro NCAP testing frame. A total of 76 tests were done.
5 realistic use cases, split into 12 different configurations, were defined to reproduce accident-like situations with which the vehicle will be face within it lifetime:
- VRU – 1 GRAIL à pedestrian use cases, with the use of an external HMI to indicate the vehicle’s behavior
- VRU – 2 Pedestrian à crossing pedestrian use cases, with and without obstruction
- VRU – 3 Cyclist à cyclist use cases
- VEH – 1 Cut-in à cutting-in vehicle use cases (preceding car changing lane suddenly)
- VEH – 2 False positive à crossing vehicle use cases to evaluate the prototype’s tuning
BRAVE vehicle can be controlled by software, including all control commands such as the steering wheel, the acceleration pedal and braking pedal. Overall, despite small speed stability issue and lateral deviation from the path, the vehicle – in its automated driving mode – was able to face the testing requests.
Along with its autonomous behavior, the BRAVE test vehicle also presented some novelties such as its exterior IHM, called GRAIL (Green Assistant Interfacing Light). According to the situation, the color of the lights adapts and communicate different messages depending on the vehicle’s detection and analysis. In addition, the BRAVE test vehicle is equipped with a high definition camera that detect information from far away. Such an early detection is double-edged: it can generate many overreactions, but its also enables early-decisions making and smoothness in the vehicle behavior. Overall, the high performance of the BRAVE vehicle should be highlighted, with full points awarded for all common ENCAP scenarios that were conducted.
The next big objective will be being able to quantify the level of anticipation. Too much anticipation would mean overreaction and possibly creating accidental situations, but too little anticipation would lead to more emergency-like and brutal reactions.