The high diversity of products and goods requires flexible and scalable robot use on packaging or production lines, which increasingly leads to multiple robot manipulators sharing a common workspace. In addition, the trend is towards demand-oriented robot use, which results in lightweight robots often being used in combination with an autonomous mobile platform. To this end, the robotics group's research focus lies in the development and implementation of algorithms for cooperative robot control. This includes, for example, scheduling tasks for optimal resource allocation, online planning of collision-free robot trajectories, motion coordination between the robot arms and the mobile platform, as well as optimal path planning for free indoor navigation, including environmental perception.
The flexible and scalable use of multiple robot manipulators on packing or assembly lines leads to complex robot manipulations with the robots performing cooperative operations. Especially cooperative pick-and-place tasks involving multiple objects are frequently encountered. The aim is to optimally perform these repetitive tasks while balancing the robots' workload and ensuring collision-free trajectory planning. For the successful execution of cooperative robot tasks, hybrid control algorithms are employed, consisting of a discrete decision layer for the optimal task allocation and an underlying layer for the online generation of minimum-time robot trajectories.
With regard to demand-oriented robot deployment, mobile robots are recently used in combination with lightweight robot arms. The coordination of the individual robot movements increases the flexibility and maneuver-
ability of the entire robotic system and extends the operating range of the robot arms. To this end, we address the development of model-based predictive methods for trajectory planning for the manipulators and the mobile platform and their coordination to navigate in a dynamic indoor environment. The environment perception and, in particular, the localization, as well as the online mapping (SLAM), are a further essential part of an autonomous mobile robot. Therefore, methods and algorithms based on machine learning (ML), especially reinforcement learning (RL) techniques, are considered.