How autonomous networks can increase safety and efficiency in mining industry

How to protect employees in a hazardous environment has always been a major challenge for mining companies. Advances in autonomous vehicles now make it possible for companies to send unmanned large machinery to operate in dangerous environments.

However, the vehicles need reliable, uninterrupted connectivity, and this is difficult to achieve using public cellular networks in remote locations.

The TM Forum Catalyst, “AI Agent & 5G: Mining with drones and autonomous trucks”, aims to help achieve this goal by augmenting a 5G-Advanced private network with drones for network backup and AI-based autonomous network management to enable seamless, reliable connectivity.

The challenges

CHN Energy (China Energy) has partnered with China Telecom to implement 5G private networks for autonomous transportation at its open-pit coal mines in the country.

At the East Junggar mine in Xinjiang Province, CHN Energy is trialling a 5G-A private network from China Telecom to support drones for emergency rescue activities and manage a fleet of unmanned dump trucks provided by autonomous transportation company Tage-I Driver.

Currently, although the automated trucks drive on their own, an engineer rides in the passenger seat and can take over manual operation in an emergency. China Energy hopes to remove the engineers from the vehicles in 2025 for fully autonomous driving if the trucks can be controlled remotely.

The private network at the Xinjiang open-pit coal mine comprises low-earth-orbit satellites and terrestrial base stations to provide 5G-A coverage on the site, which connects the trucks’ sensors and cameras with the control center.

The drones perform multiple functions, including site surveys for planning and safety inspections as well as carrying 5G-A base stations for emergency connectivity.

The Catalyst addresses several challenges for the energy company:

• The networking environment is challenging and needs optimized planning and construction to eliminate communication issues.
• The unmanned equipment and the network require real-time inspection, maintenance and monitoring, given that traditional manual processes are error prone.
• The network complexity makes fault identification and repair difficult. If faults cannot be fixed quickly, an emergency response could be activated.

The solution

The project leverages technologies across multiple domains, including AI, autonomous driving, big data analysis, large-scale network visualization models, and 5G and vehicle-to-everything (V2X) connectivity.

The result is an automated and intelligent network operation for open-pit mines.

The solution provides service level agreement (SLA) assurance for autonomous driving by creating an integrated “vehicle, ground, sky, and cloud” network.

Data is collected in real time from base stations, trucks, and drones for fault diagnosis and root cause analysis. The network status is also visualized via a 3D simulation model that reproduces the communication status across the mining site in real time.

“We monitor very closely each service connection and do closed-loop service assurance. If there’s a network issue, service priority is adjusted in real time to make sure the critical service is up and running. For example, high-definition video streams from cameras could be lowered to standard definition”, said Zhichao Yuan, CEO at Primforce, a provider of intelligent transport network solutions and Catalyst participant.

In addition, AI algorithms enable dynamic network planning by analyzing high-value site locations and providing intelligent recommendations for constructing the network with optimal 5G coverage.

Emergency network planning is also enabled via drones that survey the site and can provide and temporary network connectivity.

“All of this is driven by the AI autonomous network. We’re increasing the service quality as well as the anomaly detection, diagnosis and self-optimization. The drones are the emergency solution that carry mini base stations for temporary connectivity so that the service stays up and running,” said Yuan.

Results

The solution is expected to improve safety and efficiency for mining companies. The automated network management can increase mining efficiency by 60% and lower operational costs by 30%.

For example, the service quality assurance API has been integrated on the unmanned truck to work closely with the truck’s main control and communication system for dynamic service scheduling and adjustment based on real-time network quality monitored by the API. This has resulted in a 90% reduction in the service dropout rate and 30% operational cost savings.

The solution also enhances the safety of workers and the autonomous machines. The real-time decision-making enables better forecasting and proactive problem resolution.

“Our solution helps the mining industry in terms of providing very robust networks all the time,” said Yuan.

The Catalyst leverages multiple assets from TM Forum as well as APIs from the Linux Foundation’s open source CAMARA project. For example, it adheres to the TM Forum’s Autonomous Networks Framework, uses the Private 5G Network Use Cases asset and implemented CAMARA’s Quality on Demand API. The group also plans to contribute to Open Digital Architecture (ODA) its work on real-time network anomaly detection and root cause analysis, which is powered by the Network Quality Assurance API.

The project is championed by China Energy, China Telecom, China Unicom and Tage I-driver. Other participants include Primforce, ZTE and research and development institutions Beijing University of Posts and Telecommunications, China Telecom Fufu Information Technology Company and Hangzhou International Innovation Institute of Beihang University.