Mechanical Delay Compensation and Synchronization for Robotic Musicians Using Neural Networks

Technology has influenced musical activities, mainly in the way of human interaction. One case is human-robot interaction where humans can control robotic musicians in real-time performances. However, robots have mechanical delays caused by electromechanical components. These delays need to be compensated to keep the robots in synch. Thus, this paper presents an approach to learn and compensate robots’ mechanical delays using neural networks. Also, clock synchronization using the simple network time protocol is applied. As proof of concept, a robotic bongo was developed for tests. The experimental results for the mechanical delay compensation technique showed a reduction of 55.33% in the average delay between two robots.