The goal of the project is exploring systems and functions that enable highly automated driving at higher speeds and in complex situations. During highly automated driving, the driver can delegate the task of driving to the vehicle and can relax and move on to other things. However, he must be able to take control again within a certain time. For the transition time it is essential that the vehicle with its sensors, computing resources and algorithms must be able to assess its environments and the traffic situation correctly. At higher speeds and in complex traffic situations this can be a challenging task.
Here the Ko-HAF project starts: Its solution proposal is based on a backend solution that enables the vehicles to constantly communicate with each other using a safety server in the cloud; for the communication they use 3G/4G mobile networks. The server collects and processes information regarding the vehicle’s environment. These data then are distributed to the cars, enabling them to create a complete virtual image of their surroundings at distances up to 300 meters. “Key to making automated driving safer is the available of current data of route sections and lanes as well as their representation”, said Ralph Rasshofer who represents carmaker BMW in the project supervisory committee. “Towards this end, the project will devise and test the respective standards”.
The project also includes procedures to get the driver back into the loop. BMW’s intention is to get the driver back into its driving routine as infrequently as possible and therefor puts special situations such as hazard areas, road works and intersections into the focus of its research. Supplier Continental who coordinates the activities will devise interfaces and standard formats for the exchange of environmental and navigation data across the safety server. Another research focus are methods to localise the vehicles exactly enough to determine the traffic lane they are driving in. In this context, the Continental researchers