Making the invisible visible An AI based C-V2X approach to minimize fatal collisions

Important results from the project

The project´s concrete objective was to gradually integrate Eye-Net´s AI-based collision warning algorithm into Consenz´s intelligent, connected Head-up Display. This to enable a gamechanging V2X collision service that can warn of a probable collision even between vehicles outside each other´s field of view, mainly in an urban environment. The project has turned out very well and all goals have been met. The collaboration between Eye-Net and Consenz will thus most likely continue and expand both in scope and with more partners. A concrete service could reach the market within a year.

Expected long term effects

In all of the experiments performed during the project, the algorithm has shown expected results, without deviations. Consenz has been able to pick up, refine and forward the signal in the form of an intuitive warning to the driver in a well-placed display. This means that an already useful service is, in principle, ready to be implemented. Some work remains, however, to calibrate the warning signal with regard to angles for direction of travel and speed as well as sensitivity in probability calculations.

Approach and implementation

The project has followed a clear plan to gradually integrate Eye-Net´s AI-baserade algorithm into Consenz Head-up Display, first in laboratory environment and then in self-made prototypes. In-car tests have been carried out with good results according to plan. The implementation in the industrially produced units will take place in June and is not expected to result in any deviations, as the software is the same. Through regular tech and UX follow-ups between the project partners, the implementation was done according to schedule and without significant deviations.