Design of a Collaborative Inference System for Remote Robot Control via Private 5G

This paper presents the design and implementation of an image recognition-based drip robot system in a private 5G testbed composed of equipment from multiple vendors. Unlike public 5G networks, private 5G networks can be operated regionally, offering advantages such as the optimization of network parameters and coverage. Additionally, data transmitted and received within private 5G is maintained only locally, reducing concerns about data leakage when combined with MEC. By demonstrating the drip robot system, we show that the private 5G testbed operates reliably and that collaboration between MEC and end devices in performing AI tasks reduces latency.