The most innovative and relevant use cases were selected, regarding a vehicle safety assessment protocol enhancement purpose. Throughout the BRAVE WP5, five recommendations are to be retained.
USE OF ARTICULATED PEDESTRIAN DUMMY
Through the articulated pedestrian dummy – and its more detailed face – the representativity of a human pedestrian is improved. Deeper testing scenarios could be set up in the future, and focus on the assessment of a vehicle’s anticipation level for example, which is currently not assessed. These would deal with pedestrian’s reaction anticipation.
VRU – 1 GRAIL: LONGITUDINAL/CROSSING COMBINATION
The pedestrian combination of a longitudinal and crossing scenario should be the first milestone of advanced anticipation scenarios. This use case is based on face recognition to anticipate the pedestrian’s behavior to best react in a smooth way.
Usually, longitudinal scenarios have a small lateral error, due to the use of a passive platform that is pulled by some belts. With a robotised platform, the lateral error is constantly corrected by D-GPS positioning and remain minor.
At the beginning of the test, the pedestrian is walking straight forward (5 km/h), in the same direction as the VUT (30, 40 and 50 km/h). Upon a certain time to collision value, the pedestrian will complete a 90°-head turn whilst entering a left turn phase. Then the pedestrian comes to a full stop and awaits a trigger to start crossing.
VRU – 2 PEDESTRIAN: OBSTRUCTED CROSSING
This ENCAP-inspired obstructed pedestrian crossing represents an interesting evolution of what is already being used nowadays. It reflects the use case where an adult pedestrian (5 km/h) would come out behind a parked car – perpendicularly to the VUT direction – and start crossing in front of a rolling vehicle (30, 40 and 50 km/h).
Similarly to the previous scenario, it represents a second way of assessing the smoothness of the vehicle behavior. As part of the torso and head of the dummy will stick out from over the obstruction car bonnet, the tested vehicle should be able to early detect a possible danger.
VEH – 2 FALSE POSITIVE: C2C CROSSING
This false positive use case addressing a car-to-car crossing is constituted of two sub-scenarios and represents the approach of a vehicle on an intersection, with either crossing or turning of another vehicle in front of it.
This use case should quantify to what extent the vehicle anticipates. If a system is correctly tuned, it should be able to react and avoid collision with the crossing GVT, but also pursue its course in the turning configuration.
An important asset of next generation vehicles is their capacity of providing a smooth and comfortable ride to their occupants, whilst dealing autonomously with exterior disruptions. It might be interesting to implement some sort of assessment on this criterion within the future test protocols, that would come and adjust the grade obtained. If protocols take the lead by implementing such ponderation, we can be sure that the automotive industry will follow and adjust their production to it.