Accidents involving pedestrians and cyclists are still a very important issue for road safety. They account for a very high percentage of the total number of road fatalities and serious injuries. Crash avoidance technologies with pedestrian and cyclist detection could definitely improve road safety. Some of those technologies are being developed under the PROSPECT project. Its main objective is to lay the foundations for next generation active safety systems.
Within PROSPECT we developed and assembled three vehicle demonstrators which feature novelties on the hardware side as well on the software side. On the hardware side, we integrated an increased number of recent radar and camera sensors which feature a higher field of view as well as greater resolution. On the software side, the recent advances in machine learning allow us to extract more and more meaningful features from those sensors. For instance, besides detecting the presence of the cyclist, they are also able to detect if the cyclist is indicating his intention to turn. The car is able react to those vulnerable road users in a more naturalistic human-like way and therefore overcome the challenges of current generation. At the end, we integrated those systems into the test vehicles, to measure the increase in performance compared to the current system, but on the other hand also to aid the novel test methodologies which are developed within PROSPECT.
We have been particularly interested in drivers’ acceptance for the novel technologies being developed by the PROSPECT project. So for our research we have made extensive use of partners´ driving simulator facilities, which allow us to investigate the key human factor issues in a safe and a controlled environment. We carried out a series of studies using the driving simulator to investigate the effects of false alarms on drivers’ acceptance for the PROSPECT system. A key finding of our research was a demonstration of the importance of the human machine interface. Particularly its transparency – how it explains to the human driver what just occurred during the false alarm scenario.
What we are trying to achieve on the test track is basically two things: rirst, is to evaluate the performance of our prototypes that were developed within PROSPECT, secondly, we are gaining some hands-on experience for the new intersection scenarios. In PROSPECT we are addressing a lot of challenges regarding the new trajectories and speed profile as well as robot tuning issues. The overall goal is to make consumer testing more realistic in the near future and to contribute to traffic safety.
We are speeding up with the preparations of our programme for the Final event. In the meantime, check out our PROSPECT video here: https://www.youtube.com/watch?v=hMK6L-oNF0g&t=3s