5G Remote Real-Time Control: Key Challenges, Enabling Technologies, and System Architectures

The concept of the Internet of Things (IoT) was proposed more than a decade ago and originally relied on mobile communication networks for data transmission. Early IoT device‑control scenarios were limited to simple remote operations due to network constraints, but the emergence of 5G brings high bandwidth, low latency, and local offloading, enabling richer real‑time information exchange and accelerating the adoption of remote control.

5G’s high bandwidth, low latency, and local offloading characteristics make it a key enabler for remote control applications. Typical 5G remote‑control scenarios include emergency intervention for autonomous vehicles in ports, open roads, mines, and hazardous environments such as cranes, hoists, chemical plants, and underground mines. These scenarios serve either as emergency intervention means or as normal operational modes, improving frontline worker experience.

Future industry digitization will see unmanned and remote operations become the norm in mines, ports, logistics, and consumer‑facing services like cloud‑based taxis and remote driving.

Key technical challenges for 5G remote real‑time control focus on supporting real‑time human‑machine interaction for complex equipment. In addition to traditional status data, real‑time video, audio, and other media must be synchronized. This imposes stringent requirements on perception latency, operation reliability, and timeliness.

Three major technologies address these challenges:

Real‑time audio/video communication : Ensures that audio/video latency, which can account for up to 80% of end‑to‑end delay, is minimized. Multi‑stream high‑definition video (4‑8 streams) may be required, demanding high bandwidth and low jitter.

Control‑signal synchronization : Guarantees reliable and low‑latency transmission of control commands, which directly affect device actions and must be highly reliable.

5G network optimization : Improves uplink video/audio latency and downlink control‑signal delivery through joint network‑device optimization.

These three technologies collectively target the latency and reliability pain points of 5G remote control, while system architecture also plays a crucial role.

Typical system architectures for 5G remote control are described as follows:

(1) Architecture A – Single‑Vehicle Direct Connection with Separate Video and Control Paths

Based on traditional video surveillance and CAN‑bus control, this architecture connects multiple cameras to an NVR‑like video gateway, which is linked to a 5G private network. The control link converts CAN to Ethernet and back, transmitting CAN data over the 5G IP network. While functional, it requires pre‑configured IP addresses and offers limited flexibility for large‑scale multi‑vehicle deployments, and its video latency remains relatively high.

(2) Architecture B – Single‑Vehicle Direct Connection with Integrated Video and Control

The gateway merges CAN control capabilities with video functions, allowing collection of video, audio, vibration, posture, and vehicle condition data. This improves extensibility, enriches on‑site content, and reduces video latency compared to pure video‑gateway solutions. However, scalability for multiple vehicles still faces flexibility challenges.

(3) Architecture C – Unified Forwarding

Multiple controlled devices and control terminals connect to a central remote‑control server that forwards traffic, eliminating the need for pre‑configured IP addresses. While simplifying large‑scale deployment, it introduces additional server load and forwarding latency.

(4) Architecture D – Integrated Fusion Architecture

Proposed by Tencent Cloud’s 5G team, this architecture combines the low‑latency benefits of direct connection with the scalability of unified forwarding. The remote‑control server manages control plane functions, while data plane traffic (video, audio, sensor data) uses direct connections when possible, falling back to media relay servers if needed. This design reduces server requirements and maintains low latency, representing the future trend for 5G remote control.

In the long term, as 5G remote control applications proliferate across diverse network environments (private networks in mines and ports, public networks for logistics, and edge‑computing via 5G MEC), separating control and data planes and leveraging MEC will enable flexible deployment and further latency reductions. Standardization of audio/video and control interfaces is also expected to improve interoperability among different vehicles and control cabins.