Convergence of Machine Learning and Robotics Communication in Collaborative Assembly: Mobility, Connectivity and Future Perspectives

Collaborative assemblies of robots are promising the next generation of robot applications by ensuring that safe and reliable robots work collectively toward a common goal. To maintain this collaboration and harmony, effective wireless communication technologies are required in order to enable the robots share data and control signals amongst themselves. With the advent of Machine Learning (ML), recent advancements in intelligent techniques for the domain of robot communications have led to improved functionality in robot assemblies, ability to take informed and coordinated decisions, and an overall improvement in efficiency of the entire swarm. This survey is targeted towards a comprehensive study of the convergence of ML and communication for collaborative assemblies of robots operating in the space, on the ground and in underwater environments. We identify the pertinent issues that arise in the case of robot swarms like preventing collisions, keeping connectivity between robots, maintaining the communication quality, and ensuring collaboration between robots. ML techniques that have been applied for improving different criteria such as mobility, connectivity, Quality of Service (QoS) and efficient data collection for energy efficiency are then discussed from the viewpoint of their importance in the case of collaborative robot assemblies. Lastly, the paper also identifies open issues and avenues for future research.