Advancing Robotic Systems with Distributed Multi-Agent Digital Twins: A Scalable and Adaptive Framework

Digital Twins are emerging as one of the core technologies in robotics, enabling real-time simulation, optimisation and predictive maintenance. This article proposes a Distributed Multi-Agent Digital Twin Environment, a scalable architecture to maximise robotic system performance and safety. Through the combination of real-time virtual models and physical robotic agents, Distributed Multi-Agent Digital Twin Environment enables adaptive decision-making, multi-agent collaboration and optimal task completion in dynamic environments. The architecture is geared towards the support of single-agent and multi-agent systems through advanced communication protocols that enable synchronisation and real-world usage. The system leverages the Digital Twin Protocol over UDP for real-time synchronisation, ensuring the communication between digital twins and physical robots. Experimental results, conducted on a TurtleBot, demonstrate the system’s ability to optimise the navigation of robots, avoid collisions, and improve operational effectiveness. This work provides the gateway to more intelligent, responsive, and adaptive robotic systems in numerous industrial applications.