The aim of KIMKO is to develop an autonomous robot system consisting of a mobile platform, two lightweight robots, and stereo cameras for use in a 3D printing farm. The main research focus of this project is the collision-free online trajectory planning for the manipulators and the mobile platform as well as their coordination in order to navigate autonomously and to cooperatively plan and perform the robot motions/tasks in a confined place. Due to the high structural flexibility, model-based predictive control strategies are used for trajectories generation in order to cope with the challenges that arise in environments with static and dynamic obstacles. The perception of the environment and, in particular, the localization as well as the online mapping (SLAM) are a further essential part of this research project. Therefore, the methods and algorithms to be developed will be based on machine learning (ML), and especially reinforcement learning (RL) techniques.

The autonomous navigation of the mobile robot platform in a dynamic industrial environment relies on the continuous sensing of the environment by a visual system consisting of three stereo cameras. The resulting information is taken into account to realize a path and finally a collision-free trajectory planning for the robot. For this purpose, a data-based modeling of the environment and a map generation are performed first, which are used by global planning algorithms to compute an optimal robot path through static obstacles. The time parameterization of the generated path as well as the local path following control is realized using a nonlinear model predictive algorithm (MPC), taking into account the ongoing sensing of the environment to update the generated path and avoid collisions with dynamic obstacles.  In addition, MPC-based algorithms are also used for collision-free trajectory planning of the robot arms, so that the robot can finally perform its assigned tasks safely in a dynamic work environment.

• Model-based planning and control of cooperative collision-free trajectories for the robot arms
• Cooperative robot control to coordinate the movements of the robot arms and the omnidirectional mobile platform for a high degree of flexibility and precision
• ML-based detection and tracking of static and dynamic objects
• Simultaneous localization and mapping (SLAM) for free indoor robot navigation
• Optimization-based path planning and navigation for the omnidirectional mobile robot
• Flexible robot end effectors for gripping and handling parts of different geometries
• Development of an experimental setup consisting of a mobile robot, two lightweight robots, stereo cameras and ultrasonic sensors

• Cooperative Robotics
• Model Predictive Control
• Optimization and Control

Projektträger: AiF-Projekt GmbH

Aug 2019 – Oct 2022

DC Vision Systems GmbH

Imetron GmbH

Prof. Dr.-Ing. Naim Bajcinca
Gottlieb-Daimler-Str. 42
67663, Kaiserslautern
+49 (0)631/205-3230
naim.bajcinca(at)mv.uni-kl.de