Model-based system engineering: from sensor technology to safe autonomous driving functions
Working together with students from Technical University Munich (TUM) as part of their graduate theses and internship activities, fortiss scientists are conducting research into components such as joining or leaving a platoon and car-to-car communication, as well as ADAS functions like adaptive cruise control systems and lane/emergency braking assistants. To do that the teams are developing corresponding behavior models in the fortiss AutoFOCUS3 open source systems engineering tool and are carrying out an initial validation of the models in a functional simulation environment. The dynamic the vehicle and its environment is then co-simulated with other open source tools such as OpenModelica, ROS and Gazebo and represented in a 3D simulation. As a final step, the developed functions are implemented via code generators in the fortissimo Rover, a 1:10 scale model vehicle equipped with sensor technology such as cameras, ultrasound and laser-based distance meters.
Right image: fortissimo Rover hardware platform - model vehicle equipped with 3D-printed components, Raspberry Pi controllers and sensor technology.
The fortissimo demonstrator is furthermore used to examine current research issues. These include a model-based method for deriving so-called assurance cases for validating the functional safety of the vehicles in line with ISO-26262, degradation and reconfiguration strategies for safeguarding critical driving functions (e.g., against hardware faults), and analytical and simulation-based processes for dimensioning and validating the vehicle hardware and software architecture.
