5G-Advanced and Multi-Robot Collaboration in Private Networks

Posted by on

As enterprises move from experimenting with robots to deploying them at scale, connectivity is increasingly shaping what these systems can actually do in real operational environments. Mobile, humanoid and collaborative robots are expected to move freely across factories, warehouses, ports and retail spaces, while coordinating with each other and interacting with people. This places stringent demands on private networks, particularly around uplink capacity, latency, reliability and the ability to integrate tightly with edge computing and AI workloads.

A recent GSMA case study provides a detailed look at how China Telecom is addressing these challenges through a private 5G-Advanced solution called EasyOn·Robot. Developed in collaboration with ZTE and robotics companies DroidUp and AgiBot, EasyOn·Robot is designed specifically to support multi-robot collaboration and robot training environments. While the demonstrations described in the paper are based in China, the underlying architecture and performance targets are highly relevant to private network deployments globally.

From China Telecom’s perspective, robots represent one of the most demanding enterprise applications for private 5G networks. Training robots to develop so-called physical AI requires continuous transmission of large volumes of data from onboard cameras and sensors, often from multiple robots operating simultaneously. Even after training, robots may still rely on the network to coordinate tasks, access shared intelligence, query AI models running at the edge, and upload data for verification and fine-grained positioning. In enterprise deployments, this connectivity also underpins real-time monitoring, task orchestration and coordinated or swarm-style behaviours.

EasyOn·Robot is positioned as a simplified private 5G-Advanced network architecture optimised for these requirements. Rather than deploying a full standalone 5G core, the solution uses a lightweight, plug-in edge compute module, referred to as NodeEngine, to execute essential 5G core functions locally. This approach is intended to reduce cost and deployment complexity, with the case study suggesting installations can be completed within one to two days, while still maintaining security and isolation from public networks.

Technically, EasyOn·Robot makes extensive use of 5G-Advanced capabilities to deliver ultra-high throughput, ultra-low latency and high reliability. According to ZTE, the solution can support uplink speeds of up to 2 Gbps and downlink speeds of up to 6 Gbps, with latency below 10 milliseconds and reliability reaching 99.99 percent. A key design target is guaranteeing each robot an uplink of around 200 Mbps, enabling continuous streaming of high-resolution sensor and environmental data in real time. This uplink-centric design is particularly important for robot training, coordination and precise positioning in dynamic environments.

ZTE’s role is central in providing the radio, system and edge integration capabilities that enable EasyOn·Robot. The architecture is based on an edge-device synergy model, allowing different types of robots to connect through standardised interfaces and operate as a coordinated system. This is critical in real-world deployments, where enterprises are unlikely to rely on a single robot vendor. The platform is designed to support heterogeneous robots, including humanoid robots, wheeled or legged platforms, robotic arms and even digital twin robots, working together within the same private network.

The contributions of DroidUp and AgiBot help illustrate how these network capabilities translate into practical use cases. Demonstrations supported by China Telecom’s private network include multi-robot live performances, collaborative customer-facing roles and a 24 by 7 unmanned convenience store. In the retail scenario, robots autonomously handled customer entry, personalised recommendations, checkout and delivery, with the private 5G-Advanced network ensuring millisecond-level sensor synchronisation and coordination. At events such as MWC Shanghai and the World Artificial Intelligence Conference, robots from different vendors were shown collaborating in the same environment, supported by extensive large language model and visual language model training.

From a private networks standpoint, the case study underlines a broader architectural shift. EasyOn·Robot treats connectivity, edge computing and intelligence as a single integrated system rather than separate layers. Real-time monitoring by edge nodes and intelligent scheduling allow tasks to be dynamically allocated across robots, enabling what the partners describe as team intelligence. For operators such as China Telecom, this positions the private network as a core enabler of enterprise automation rather than a background utility.

The paper also highlights practical considerations for wider adoption. Spectrum availability remains a key factor, with national policies determining whether enterprises can deploy private 5G networks directly or need to work with operators. At the same time, simplified architectures that lower deployment barriers are likely to be attractive to enterprises that want to scale robotics without building deep in-house telecoms expertise.

As the global robotics market continues to grow, driven by automation, labour shortages and advances in AI, private 5G-Advanced networks are expected to play an increasingly important role. China Telecom’s work on EasyOn·Robot, together with ZTE, DroidUp and AgiBot, provides a concrete example of how operators can evolve private networks to support uplink-intensive, latency-sensitive and highly coordinated robotic systems. It also offers a glimpse of how private networks may become a foundational platform for multi-robot collaboration and physical AI in the coming years.

The video provides further details:

Related Posts: 

Comments

Post a Comment