Vehicle-to-vehicle communication promises a large safety benefit for motorcycles. Furthermore, new motorcycles are equipped with an increasing number of vehicle dynamics sensors (e.g. wheel speed sensors, gyro sensors). These deliver information about the current driving state variables. Hence, the Institute of Automotive Engineering at Technische Universität Darmstadt and carhs.communication are in the process of researching the fundamentals of a communicationbased warning system for motorcycles. This system generates sensor-based or manually entered warning messages and sends these to other motorcyclists using wireless communication devices. In this way, riders can receive early warnings of road hazards. In order to detect hazards based on standard sensors, new methods were developed, which are presented in this paper.

An analysis of an accident database and a motorcycle rider survey revealed the following main causes for accidents that would be avoidable using a system such as that investigated here:​

• Roadway damages, e.g. unevenness, ground waves, transversal ruts, pot holes
• Obstacles on the road, such as broken down vehicles behind a curve
• Excessive speed in curves, especially in irregular road conditions
• Friction steps caused by oil, gravel sand, bitumen

Driving dynamics for the above mentioned situations were analyzed. New criteria were derived and used to generate warning messages based on vehicle dynamics sensor information. In order to validate the criteria, over 500 test drives were conducted. To detect hazards caused by individual roadway damages, a new criterion was derived based on the measurement of the vertical wheel acceleration. With this criterion, hazard detection becomes independent of hazard type and shape.

Obstacles are detected by identifying evasive maneuvers. These are distinguished from other maneuvers by means of a correlation factor, determined on the basis of a previously defined standard maneuver and the current driving state. In a previous study, the vehicle side-slip angle velocity was found to be a criterion to detect critical driving situations during cornering. These situations are caused by friction steps or by exceeding the maximum lateral acceleration. The current study adapts this criterion for use in a communication-based warning system.

Friction steps and low friction (during straight driving) are detected by evaluating braking activity and longitudinal acceleration. In addition to methods for hazard detection, a methodology to design an appropriate HMI was also developed and validated. To increase market acceptance, an additional comfort-oriented functionality was implemented und tested. This function is based on the same technology as the safety function.