Integrating 5G-Enabled Robotics for Remote Supervision in Smart Factories: A Virtualized Framework

Abstract

Modern manufacturing demands flexible automation to handle variable tasks, but many legacy machines lack digital interfaces. This creates a need for autonomous mobile robots (AMRs) with manipulators to interact within the manufacturing cells. But there is a gap in such an environment: It requires the presence of humans onsite for handling exceptions and deviations, which is inefficient. The GigaBot project addresses this gap by proposing a virtualized supervision framework that blends robot control, digital twins, and private 5G networks. This framework enables such AMRs to perform routine work autonomously while escalating to a remote operator for novel teleoperation needs. A digital twin environment, implemented in Ignition Gazebo, provides a simulation-first workflow where operator actions are validated virtually before being applied to the physical system. Supervision is coordinated through a Remote-Control Centre (RCC) that merges live manufacturing data, simulation feedback, and policy-based safety controls. Private 5G networks provide predictable connectivity, isolation of traffic classes, and integrated positioning, with a complementary Virtual 5G Lab (V5GL) extending these capabilities into simulation for communication-aware development and scenario testing. This establishes the foundation for network-aware, human-in-the-loop automation enabling resilient human–machine collaboration in smart factories. Beyond general smart-factory scenarios, the framework is directly applied to CNC machining, enabling automated tool handoff, machine tending, and in-cell inspection via an AMR supervised over private 5G; its validity is demonstrated in a machining cell with a synchronized digital twin and bounded-authority remote teleoperation.