SenseRoad_AD - Sensing the tire-road friction for safe and comfortable adaptive automated driving functions

Automated driving functions and driver assistance systems available in standard series-production vehicles are designed for the conditions on dry roads, e.g. concerning the calculation of the reference distances to vehicles ahead for emergency brake systems or adaptive cruise control. For driving functions with an automation level of SAE Level 3 or higher, the driving function at least temporarily takes on the task of monitoring the environmental status and thus the obligation to adapt the driving style to the road conditions. There is currently no technology that can determine the current road condition with sufficient accuracy and robustness for use in safety-related automated driving functions from an automation level of SAE Level 3 and higher. The aim of the SenseRoad_AD research project is to develop sensor fusion approaches to identify the current road conditions. In addition to the categories dry, wet, icy and snow-covered, a quantitative assessment of the condition using the tire-road coefficient of friction is required for use in automated driving functions. In a first step, a technology evaluation of three-dimensional time-of-flight cameras is carried out, which are currently mainly used in the vehicle interior (e.g. for gesture recognition). Preliminary work shows that the reflectivity of the received signal correlates with certain road conditions. In a second step, sensor fusion approaches are developed, which, in combination with the standard electronic stability control sensors, have the potential to enable high accuracy and robustness over wide operating areas and driving situations. All developed technologies are integrated on a real-time capable platform and are tested, optimized and validated in extensive static and dynamic tests. With the results of the research project, automated driving functions can be adapted to difficult weather conditions and can also be used in challenging environments. Combining these sensor information, it is possible to react quickly and appropriately to dangerous traffic situations such as slippery or snowy roads and thereby increasing road safety. In addition, the resilience to failures of individual sensors is increased by the sensor fusion approaches. Numerous accident statistics and studies show that these conditions are also challenging for human drivers and will be avoided as far as possible. As a result of this research project, road safety for human drivers will also increase thanks to situation-specific warning strategies without false warnings.