Reference Framework for Coordinating Robots for Movement of Objects
Aditi R Deshpande, Mallarajapattana Janardana Venkatarangan · 2021
One of the common use cases in many factories is that loads of materials is to be moved from one place to another. It is desirable to have payload carriers for automating the tasks to improve productivity and reduce the burden on humans. The automation becomes even more important for hazardous environments. This framework is developed for a concept of multiple robots that are deployed by enabling a form of coordination from a centralised system. The work is an attempt to create a framework for such applications in which multiple robots work to accomplish payload movements. The master is a computer that uses a vision system and identifies the number of robots, the payload objects to be transferred, their initial positions, the destination points that they all need to be moved to and the obstacles in the environment that the robot needs to manoeuvre around. The master commands the robot motion, picking up and transfer the payloads to destination. The realisation also includes algorithms for optimisation of the paths with the objective of minimizing the distances to be traversed. In the prototype developed, the proposed system is evaluated based on one master running on a PC system and three robots working on payload transfer. The proof of concept is based on a prototype with robots and arena of area 1.5m x 1.5m with representative payloads, obstacles, and targets. The search algorithm is developed in Python for path optimisation with real time video feedback of the workspace from a camera. The framework can be extended to applications comprising of multiple robots as each of the robots will communicate wirelessly with the master. Scenarios were created based on different positions for payload and obstacles to determine the time taken for payloads to be moved to target locations. The optimal time taken was found to be 130.09s and 136.39s as the robots traverse the shortest path and at constant speed. The framework is found to work for real time environment changes in terms of moving robots and obstacles.