Implementation of path following and obstacle avoidance in omnidirectional platforms

Mobile robots, in particular AGVs, have witnessed an increased interest in the last few years. This is due to the reduced costs and increased precision, which makes them highly advantageous in industrial applications like logistics. Depending on the operating scenario and constraints, this indoor robot can be built using different kinematic models (i.e., differential drive, omnidirectional, etc.). To best exploit each model's advantage, different control and planning algorithms must be implemented. In this work, we propose a pipeline for path following and obstacle avoidance for omnidirectional robots. Our solution allows switching between different operating modes, simulating different kinematic models based on the scenario. The proposed system can achieve low error (i.e., less than 5cm) in the operating scenario.