Information and communications technology (ICT) infrastructure provider Huawei South Africa brought together mining companies, technology partners and university researchers in Johannesburg during Huawei South Africa Connect 2026 to discuss the shift from connected operations to intelligent mining. The discussion focused on the technologies needed to support that transition, including private wireless networks, underground Wi-Fi Mesh, optical fibre sensing, edge computing, intelligent video, robotics, data governance and digital twin platforms.

Participants agreed that safety, productivity, energy efficiency, asset reliability and sustainability remain the industry's priorities. They also agreed that AI will deliver value only when mines combine strong digital infrastructure with trusted operational data and experienced implementation partners.

Connectivity Becomes the Operational Foundation

Huawei outlined an approach that matches connectivity to different mining environments. In open-pit operations, private 5G can support mobile equipment, video communications, vehicle and equipment monitoring, remote operations and high-volume data transmission without recurring public-network data charges. For underground mines, Wi-Fi Mesh can extend connectivity into changing working areas where fixed fibre deployment is difficult or vulnerable to blasting activity.

The company also highlighted optical-fibre sensing as a way to extend perception beyond conventional communications. Distributed fibre sensing can be used for perimeter intrusion detection and for continuous acoustic monitoring of conveyor idlers.

Because the fibre itself acts as a continuous sensor, mines can detect unusual vibration or sound patterns and identify potential faults earlier, shifting maintenance from scheduled inspections to condition-based maintenance.

Together, these technologies allow mines to move beyond connectivity by turning operational data into faster decisions and measurable business outcomes.

AI Must Begin With Mining Value

The roundtable explored practical AI applications already relevant to Southern African mining. These included conveyor-belt inspection, oversized-material detection, equipment-health monitoring, perimeter protection, personal protective equipment detection, driver-fatigue monitoring, predictive maintenance and automated operational reporting.

Computer vision was highlighted as a practical starting point because many mines already have extensive CCTV networks. Edge AI can analyse existing camera feeds to identify unsafe behavior and equipment issues, including conveyor belt damage. Large language models could also help employees search operating procedures and maintenance manuals using natural language before generating reports or summaries.

“Connectivity is the starting point, not the goal. The real challenge is no longer how to connect equipment on a mine. It is how to turn operational data into safer working conditions and measurable gains in productivity. That happens when mining companies and technology partners build the digital foundations together, and when AI moves from trials into everyday operations,”  said Huawei sub-Saharan Africa Oil & Mining Business Director Charlie Li.

Robotics Links Research with Underground Safety

The University of Pretoria presented research into quadruped and humanoid robots for underground mining. The university’s Dr Gang Yu described potential uses including routine inspection, post-blast assessment, equipment handling, emergency response and data collection in hazardous or GPS-denied areas.

The research program is designed to test robotics as a use case rather than simply deploying robots underground. Its focus includes underground mobility, remote and semi-autonomous operation, sensor integration, AI-enabled perception, workflow integration, cybersecurity and ethical readiness. A proposed proof of concept at South Deep Gold Mine would assess whether robotic platforms can operate reliably in deep-level conditions while reducing human exposure and strengthening real-time operational control.

Digital Twins Provide a Common Operational View

ICT company and systems integrator BCX demonstrated a digital twin platform that brings together data from Internet of Things sensors, CCTV, PLCs, SCADA, MES and other operational systems in a single two- or three-dimensional view. Rather than replacing existing systems, the platform brings their data together through a single interface.

The demonstration showed how a mine could visualise production data, energy and water use, personnel and vehicle locations, environmental readings, equipment status, alarms, and surveillance feeds in one operational view. Open interfaces and low-code integration were presented as important for reducing the cost and complexity of connecting diverse mining technologies.

The platform could support faster decision-making, personnel tracking, threshold-based alerts, maintenance planning and future mine expansion. Speakers noted, however, that digital twins are only as effective as the quality of the data and integration behind them.

The discussion concluded that connectivity is only the starting point. The real value comes from turning trusted operational data into practical AI applications that improve safety and productivity.