The rehabilitation of roads causes considerable traffic restrictions and latencies in inner-city areas and especially on the highway. The associated maintenance costs overwhelm many municipalities in the Federal Republic. The introduction of a new type of automated road-saving driving system that prevents serious road damage would make a significant contribution to reducing the annual maintenance costs of several billion euros in Germany.
It is based on automatic early detection of impending damage in the deeper layers of the road by means of ground penetrating radar. At the same time, the radar programs provide useful information for localization.
The aim of the present research project is to develop a cloud-based traffic service for health monitoring for traffic routes and highly automated road-friendly driving.
This will be achieved by using AI-based data analytics and a 3D map of traffic routes created by fusing GPS and GPR data.
For this purpose, on the one hand, a ground radar sensor with dual operating frequency will be developed and integrated into a prototype test vehicle of automated driving. On the other hand, a digital twin for traffic systems will be designed and implemented on the cloud to enable novel off-board localization and traffic services such as predictive road maintenance and HAF assistants for lateral and longitudinal guidance, e.g., for small-scale bypassing of previously detected severe roadway underpass damage.
In particular, the bypassing vehicles should also benefit from the highly accurate localization of the GPR-equipped vehicle by means of Car2Car communication.
Tasks 3D Mapping Solutions
The task of 3D Mapping is to record the selected test sections with measurement vehicles to provide the reference data for road condition analysis. The basic evaluation includes the processing of all acquired raw data of the measurement system, i.e. the trajectory calculation, temporal synchronization and the project-specific measurement data provision. 3D Mapping also has effective tools for measurement data extraction. The following objects are evaluated in the measurement data, i.e. scanner data and camera data: All lanes, including inbound and outbound lanes as well as hard shoulders, especially with course, elevation profile, exact lane width, cross slope, etc. The central element for relating the measurement data to the vehicle is the definition of axes for the respective lane with course, elevation profile, transverse slope, longitudinal slope and exact curvature.