PRYSTINE final event

2021-12-03T10:01:48+00:00

PRYSTINE final event was the crowning jewel of 42 months of work performed by 60 partners across the Europe. During this time, consortium partners worked tirelessly in creating fail-operational systems for autonomous vehicles in urban and rural settings.

PRYSTINE final event took place in Università degli Studi di Modena e Reggio Emilia in Modena, Italy on 26^th – 27^th of October 2021. Project Management team presented PRYSTINE accomplishments live to Project Officer Anton Chichkov, while 53 consortium partner’s representatives, and project reviewers Lars Fredrik Rudolf Dahlgren and Dirk Wolfgang Friebel joined online. Final review event, including live demonstrations, was streamed via Webex platform to remote participants.

Over the course of the project, consortium partners produced 30 demonstrators, that showcased technological developments of the project, and the most impactful ones were showcased during the review. Public demonstrators are available online on this page (please see below) as well as on the project’s YouTube channel.

From several live demonstrators presented during the review, physical demonstrator car encompassed all technological developments. Maserati prototype was showcased on track during autonomous driving tests and performed demonstration of everyday road situations (traffic light turning green, approaching an intersection with pedestrian partially occluded, and stopping in emergency lane). Project management team and project officer were able to do a ride-along with UNIMORE representatives. The ride-along experience was also streamed to remote participants, so that they can get a glimpse of demonstrator car capabilities.

Several live simulators were presented in the review as well.

One of the presented simulators showcased real-time traffic state prediction. In this simulator decision-making tasks are made by Artificial intelligence, which have been “learning” from driving behaviour of a human driver on the highway and in urban setting and modelled human driving behaviour. The traffic state prediction provides a complete consistent traffic view for the area, and functions as an external additional sensor to the car, which input is used by the decision maker.

Another simulator presented a road safety alert system. This system uses machine learning based recognition engine to detect predefined set of hazardous objects on the road (e.g., wild animals, potholes) and warns other vehicles by using reliable and low latency alert.

The last simulator showcased trajectory planning and vehicles dynamics control, allowing a seamless switch between manual driving and autonomous driving mode.

PRYSTINE project developments have made a sensible impact to the development of fail-operational behaviour systems in self-driving vehicles, that can be built upon in further research.