Increase Road Safety Volkswagen Optimizes ADAS and AD With Data From Real Traffic Situations

Related Vendors

Volkswagen AG, Volkswagen Nutzfahrzeuge

FAULHABER Drive Systems

Wuxi InfiMotion Technology Co., Ltd.

Volkswagen will work on improving driver assistance systems and automated driving functions using data from real traffic situations to enhance traffic safety. To this end, the company is expanding its program for using sensor and image data from customer vehicles to around 40 European countries. The initiative began in January with models from the Volkswagen passenger car brand; it will be followed by Volkswagen Commercial Vehicles, Audi, Cupra, Škoda, and Porsche.

The Volkswagen vehicle fleet already generates high-resolution maps using anonymized swarm data. This helps vehicles, for example, with lane guidance on roads without lane markings. Precise driving instructions and hazard information, which can be narrowed down by local weather, are also possible.

To continuously optimize driver assistance systems, data from real driving situations, which are significantly more practical than tests with prototypes or simulations, will now be used. The aim is to design assistance functions in such a way that they are perceived as effective by the user and ideally remain permanently activated. Active systems increase safety not only for the driver and passenger but also for all road users in their surroundings.

Specific data transmission in complex situations

For the work, the automaker focuses on situations where driver assistance systems are particularly useful. These can include traffic situations with cyclists and pedestrians, such as intersection traffic near schools and kindergartens, or busy, unclear supermarket parking lots.

Triggers for data transmission can include the emergency braking assistant, manual full braking, and sudden evasive maneuvers. In this context, certain sensor, function, and image data play a particularly important role. These include camera images of the vehicle's surroundings and detection results from the environment sensors, as well as the driving direction, speed, and steering angle. Information about weather, visibility, and lighting conditions is also relevant. The following example illustrates this: The vehicle should analyze movements at pedestrian crossings and sidewalks as effectively as possible. If the camera detects pedestrian movements toward the street—such as those of playing children—the vehicle can build up braking pressure preventively to brake even faster in an emergency.

Data processing only with consent

A permanent data transmission for this purpose does not take place. User consent is the fundamental prerequisite for the transmission and processing of the data. This consent can be provided in various ways and is customized by the brands, for example, as an option in the user's own profile. This consent can be revoked at any time.

The data collection and transmission can also involve other vehicles or road users such as pedestrians and cyclists in the immediate surroundings. This is important because camera-based systems must visually classify objects and people clearly even under adverse conditions and correctly assess complex traffic situations. Individual information about people in the traffic environment is not relevant. (se)