Digital Twin-Assisted Joint Communication and Control Scheme for Intelligent Robot Collaboration

In this paper, we propose a novel digital twin (DT)assisted joint communication and control scheme, in which robots can achieve much better collaboration for search and rescue (SAR) tasks in disaster-affected areas. Particularly, the design of the scheme is decomposed into two sub-developments at different timescales. First, a communication-aware robust control policy is developed for each robot in a small timescale based on nonlinear model predictive control and control barrier function constraints, which mitigates the impact of device-todevice communication delay on task execution. Second, a controlaware radio spectrum resource allocation strategy is developed at the edge server in a large timescale to improve total control task effectiveness of robots. This is achieved by a DT-assisted deep reinforcement learning (DRL) algorithm, where DTs of robots are utilized to emulate potential robot movements for DRL state augmentation. Simulation results demonstrate that the proposed scheme outperforms benchmarks for SAR tasks. A simulation demo can be found at: https://youtu.be/9Salnh9EIVg.