Camera based road condition prediction for autonomous driving

Important results from the project

The project´s goal was to develop a first simplified demonstrator and lay the foundation for a product architecture, with the purpose of demonstrating the intended koncept to a potential customer. These goals were achieved, as well as the purpose in itself. The additional goal, to file a patent application, to protect the intended concept, was unfortunately never met since existing foreground IP and "known information" forces us to tweak the application. Because of this, the product concept has not been described for customers and some uncertainty exists.

Expected long term effects

The result of the project has led to the expected outcome, that a potential partner has shown great interest and wishes to move on in the discussion about a joint development project. However, as mentioned above, the filing of a patent application has failed and the effect of having submitted such an application and thus obtaining a priority date has unfortunately not been achieved.

Approach and implementation

As the project required winter roads, major challenges arose when there was no snow but the project deadline stayed the same. To handle this, artificial snow was created when the temperature allowed and data could be collected on limited road stretches. This was however not enough, so it was decided to focus the demonstrator on being able to distinguish asphalt road from gravel road. This alternative demo-target was successful and, despite the lack of snow, could still demonstrate the potential of the solution.