ROADAR: Safe autonomous driving in any weather conditions

Early warning system for aquaplaning and black ice

Aquaplaning and black ice turn streets into dangerous slides, which often leads to serious accidents. Within the framework of the high-performance center »secure intelligent systems« (LZSiS), Fraunhofer EMFT and Intel are working together on a solution to the problem. A real-time warning system shall identify the potential danger due to water or ice on the road surface, thus enabling a predictive detection of the road conditions. Such an early warning system would increase the road safety for all traffic participants.

Privacy warning

With the click on the play button an external video from wwww.youtube.com is loaded and started. Your data is possible transferred and stored to third party. Do not start the video if you disagree. Find more about the youtube privacy statement under the following link: http://www.youtube.com/t/privacy_at_youtube

Danger of traction loss and autonomous driving

If cars are soon due to be on the way on our streets without a driver, the vehicle must be able to travel from A to B independently, despite pouring rain or black ice on the road. This is made possible by an assistance system, capable to predict the imminent change on the road surface and integrated into the vehicle control system. The vehicle will be able to automatically react to such changes and to avoid the risk. The security for all traffic participants can be further increased by connecting the assistance systems with each other. Using maps with information on the road conditions, other vehicles could be warned about risky situations early on, and their velocity could also be adapted accordingly.

Functionality and costs

The danger detection of the system is based on interpretation of the optical characteristics of water and ice, which makes the system more reliable and secure than the methods used so far. Data collected by near-infrared and polarization sensors, together with AI-assisted data analysis, enable unambiguous detection and localization of hazardous road surfaces caused by weather. Use of commercially available optical CMOS-sensors combined with optical filters keeps the system mechanically simple and cost effective.

NIR-Sensors

Assessment of the optical light absorption behavior

Tested in the laboratory

Study of ice and water on various subsurfaces

AI-supported

Data analysis supported by machine learning

Object detection

Development of algorithms based on image processing routines

Cloud Data Management

Cloud-based management of road surface data

Tested in real conditions

First use in real conditions for real-time detection