Audi is using high-precision swarm data for the first time to improve its car-to-X service “local hazard alerts”. The new version uses a car-to-cloud application that is based on a novel procedure for estimating the coefficient of friction on the basis of the wheel slip. The technology can detect minute changes in road surface adhesion, upload the data to the cloud to be processed, and warn drivers behind of road ice, for example, in near real time.
Since 2017, cars made by Audi have been able to warn each other about accidents, broken-down vehicles, traffic jams, road ice, or limited visibility. To do this, the car-to-X service “local hazard alerts” analyzes various data. These include interventions of the electronic stabilization control (ESC), rain and light sensors, windshield wipers, and headlights, as well as emergency calls and airbag triggering.
Audi is now taking the next step and improving the service with high-precision swarm data to make the warning even faster and more precise.
Audi is the first manufacturer to use a patented solution from Swedish company NIRA Dynamics AB for this purpose. The two companies used this solution as a basis to develop the improved hazard alerts together with the Car.Software organization and HERE Technologies.
In the car, the system can estimate the coefficient of friction between the tires and the road surface on the basis of the wheel slip. To do this, it uses signals from the suspension, such as the wheel speed and acceleration values. It is permanently active in normal driving situations and not only at the physical limits that would require the suspension control systems to intervene.
The sensor data is anonymized, both in the car itself and also in the cloud, hosted by NIRA Dynamics AB, to which it is transmitted. The aggregated data is combined with metadata such as current weather information and empirical values, and then transmitted by a NIRA server to service provider HERE Technologies. Here, the data is integrated in the HERE location platform that represents the road network as a precise three-dimensional model.
The HERE servers send the warning information to those cars that are in or headed toward areas with poor conditions. The driver sees the warning in the Audi virtual cockpit or on the optional head-up display.
The greater the number of vehicles that deliver the data, the better the system can learn, analyze, and create maps, and thereby inform or warn the drivers depending on the situation. This is the basic principle of swarm data and swarm intelligence—an area in which Audi has acquired extensive knowledge over the past years.
In Europe, more than 1.7 million vehicles from the Volkswagen Group will supply current data for the hazard information service in 2021, and this number will increase to more than three million in 2022. The service is available in the new models from Audi, Volkswagen, SEAT, Škoda, Porsche, Bentley, and Lamborghini.
The Car.Software organization, a company of the Volkswagen Group, bore the main responsibility for the development. The project was designed in such a way that the greatest possible number of drivers could benefit from the safety advantages regardless of the group brand. This is also the first customer application as part of which vehicle data is used for this type of data analysis.
The project for improved hazard alerts is a good example of the great potential of cross-brand software development. Together with other Group brands and our strategic partners, we were able to develop a digital service within a few months while making use of our own software skills and economies of scale. The improved hazard information service is just the beginning; we see wide-ranging potential for the future.—Thomas Müller, Head of Advanced Driving Assistance Systems ADAS & Automated Driving AD at Car.Software
Using current friction coefficient maps based on this data pool, municipalities can optimize their snow clearing service in real time, and also reduce the environmental impact by using less road salt. Driver assist systems can precondition themselves and adjust to the condition of the road with even greater precision, and the route guidance of the navigation system can take the road conditions into account in order to offer a more accurate computation of the expected time of arrival. In the car, control of the wheel slip can enable the development of tire maintenance services, for example, by detecting the level of wear as well as the performance level of the tire.