At the laboratory of automotive mechatronics (www.mechlab.de) a system for autonomeous driving will be developed. This is part of an european union founded project to investigate new methods of information fusion based on laser scanner and wireless connection data. The first step is the implementation of a new longitudinal control algorithm for the test car BMW i3 (Adaptive Cruise Control for urban environment - uACC). The principles of such systems are described here:
In our car, two different types of laser scanner are used. An Ibeo Lux with a 110° field of view is mounted at the front of the car. This is the previous model of the new laser scanner SCALA of the large automotive supplier Valeo. Additionally, a Velodyne 360° laser scanner is placed on the top of the car. The basic steps necessary for the development are described here in principle:
The actual layout of our system can be seen in the figure. Different electronic control units are used for the implementation of the different subtasks. The devices are connected via an ethernet connection.
For students we have different possibilities for participation. For short times (some weeks, e.g. study abroad) they can work on systematic measurements in real driving scenarios and evaluation of the results. Also the parameterization and testing of different control loops is possible.
For longer times (some months, e.g. bachelor/master thesis) students can develop and test their own algorithms for different subtasks:
- object detection using raw data of laser scanner
- data fusion using different object sources
- situation interpretation and cut-in/cut-out detection
- investigations on self-learning methods for control loops
- new concepts of visualization of ACC information
- simulations of ACC using Blender (https://www.blender.org)
For more information contact please: M.Sc. Sven Eckelmann (email@example.com)