In-operation real-time simulation for predictive online collision monitoring of production systems with motion controls (SimPrO)

• Applicant: Professor Dr.-Ing. Alexander Verl
• Subject Area: Production Automation and Assembly Technology
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 545637120

In the operating phase of production systems with motion controls, collisions represent both an economic risk and a danger to people. In dynamic motion sequences of multi-axis kinematics and especially during the cooperation of several kinematics, the risk of collisions is particularly high. Due to changes and interventions in control programs and parameters at runtime by humans, adaptive systems or due to the situation-dependent behavior of motion controls, a motion sequence in operation can deviate from previously tested motion sequences. Therefore, it is necessary to be able to safeguard collisions in the operating phase, in real time and predictively. In preliminary work it is exemplarily shown that by using simulation models from virtual commissioning a simulation-based online collision avoidance in the operation phase can be realized. This approach is characterized in particular by its continuous embedding in the engineering of production systems. However, questions concerning the systematic transfer of simulation models, the comparison and the quality of prediction methods, as well as the real-time capable design of collision models remain unanswered. The goal of this project is the realization and investigation of a continuous simulation-based online collision monitoring. On this basis, the open questions can be researched. In the investigation of prediction methods, the integration of virtual controls and the use of control-integrated data (from motion planning) are particularly promising. In the area of real-time capable collision calculation, good results could be achieved with the voxelization approach in preliminary work. Therefore, this approach will be investigated in more depth within the scope of this project. The developed partial solutions, as well as the resulting overall solution, will be validated on a real production system in order to determine reliable statements about the solutions and their industrial applicability.

• DFG Programme: Research Grants
• Co-Investigator: Dr.-Ing. Armin Lechler