Mining companies can buy autonomous drills, remote-controlled dozers and AI software. The harder question is whether the mine has the smart mining digital infrastructure to make those systems work reliably once they leave the pilot stage.
That is turning mine-site connectivity into an operating-capex decision. The issue is no longer whether a machine can be connected. It is whether hundreds of machines, cameras, sensors and control systems can remain connected as an open pit expands, an underground mine advances, and more production-critical applications get added to the same network.
The consequences can be surprisingly tangible.
At Newmont’s Cadia operation in Australia, for example, Wi-Fi limitations affected remote dozing. The company previously struggled to connect more than two machines at distances of around 100 metres, while network instability could force operations to stop for troubleshooting. After deploying Ericsson Private 5G, Newmont reported connectivity across a much larger work area and a 50% increase in dozing capacity.
[Suggested image: a mine-site connectivity/network diagram. Alt text: “smart mining digital infrastructure network diagram”]
More Read
That is the more important story behind mine connectivity. It is not about getting a better signal. Rather, it is about preventing the network from becoming the constraint on equipment the mine has already paid for.
Smart Mining Digital Infrastructure Is Becoming an Operating-Capex Decision
Autonomous and remotely operated equipment changes the role of communications infrastructure.
When a conventional machine loses a connection, an operator may still be sitting inside the cab. When a remotely controlled or autonomous machine depends on a wireless link for command, monitoring or coordination, the network becomes part of the operating system.
Epiroc’s technical requirements for autonomous material handling explicitly call for adequate network coverage and bandwidth, network segmentation, and quality-of-service prioritisation for time-sensitive traffic such as remote machine controls, safety and access control. The company notes that it has validated its automation solutions over Wi-Fi, LTE/4G and 5G networks.
That is an important correction to the usual AI narrative. Autonomy is not just an equipment problem — it is a systems problem. The truck or drill needs positioning, perception, onboard computing and an automation platform, and all of that depends on communications infrastructure capable of supporting those functions in the areas where the machine operates.
If the network is unstable, the autonomy system inherits that instability. This is exactly why smart mining digital infrastructure has become a production question rather than an IT question.
Surface and Underground Mines Need Different Network Architectures
There is no single “best mining network.”
An open-pit mine may need coverage across kilometres of changing terrain. An underground operation has to push connectivity through tunnels, shafts and working areas that change as mining advances.
That difference affects network design.
Wi-Fi can remain useful for local, high-bandwidth applications and controlled areas. It is familiar and can be relatively straightforward to deploy where coverage requirements are limited.
But large mobile operations create a different requirement. Ericsson’s work with Newmont illustrates the difference. At Cadia, the company says its Private 5G network extended coverage from roughly 100 metres with the previous Wi-Fi arrangement to as much as 3,000 metres in the relevant deployment, while supporting high uplink throughput. At Peñasquito in Mexico, Newmont replaced dozens of relocatable Wi-Fi trailers with six cellular radio towers.
For an expanding mine, that matters because the infrastructure itself has an operating cost. Moving every trailer, repositioning every access point, and resolving every network interruption that requires technicians all represent labour, downtime, or both.
A network architecture that reduces those interventions can change the economics of the digital system.
LTE, 5G and Wi-Fi Are Not Interchangeable
The debate over private LTE versus 5G versus Wi-Fi can easily become a technology-sales exercise.
For mine operators, the better question is: which network fits the application and the physical mine?
| Network | Where it can fit | Main consideration |
|---|---|---|
| Wi-Fi | Local areas, plants, workshops and some mobile mining applications | Cost and familiarity, but coverage and mobility can become difficult across large or changing sites |
| Private LTE/4G | Wide-area mine coverage and established connected equipment | Mature private-cellular option with broad coverage and mobility |
| Private 5G | High-performance applications, remote control and expanding automation | Higher performance potential, but requires appropriate spectrum, equipment and network design |
| Fibre | Fixed backbone and high-capacity links | High performance, but physical deployment can be difficult across moving mine infrastructure |
| Mesh wireless | Mobile and changing environments | Can provide resilient connectivity, depending on architecture and application |
In practice, mines are increasingly likely to use hybrid architectures rather than replacing everything with one technology. Epiroc says mining networks commonly combine cellular LTE, Wi-Fi and wireless mesh, while 5G remains an emerging option rather than a universal replacement — which is precisely why a one-size-fits-all approach to smart mining digital infrastructure rarely works.
That is significant for procurement teams. The decision is not simply “buy 5G.” It is to design a communications layer around the mine’s production requirements, coverage conditions, equipment fleet and expected technology roadmap.
Latency Matters, But Reliability Matters First
Latency gets plenty of attention in discussions about 5G.
It matters when an operator is remotely controlling equipment and needs commands, video and machine responses to move quickly between the control centre and the machine. Ericsson reports that its Newmont Cadia deployment achieved latency below 50 milliseconds for the operator-to-machine control path, alongside high uplink performance. Another mining connectivity supplier, RADWIN, markets systems for autonomous haulage with latency below 10 milliseconds and low jitter for control, navigation and collision-avoidance applications. Those figures are vendor-specific capabilities, not universal requirements for every autonomous mining system.
That distinction matters.
A mine should not buy a network because it has the lowest advertised latency. For production-critical applications, the more fundamental question is whether the network can provide predictable performance across the operating area, including during movement, interference, changing terrain and network handovers. A theoretical peak number is less useful if a machine encounters a dead zone during a critical operation.
[Suggested image: an autonomous haul truck or drill rig on site. Alt text: “autonomous mining equipment connected via smart mining digital infrastructure”]
Sentinel Shows What Sits Underneath Autonomous Drilling
First Quantum Minerals’ Sentinel Mine in Zambia provides another useful example.
Sedna Africa supplied and commissioned a Rajant wireless network for autonomous drilling at the copper operation. The network supports the drill automation system with mobile connectivity, low-latency communication, high-capacity data transfer and remote operation.
The project moved beyond the abstract idea of a “smart mine.” The network was deployed to support an actual production application. Following trials on three drill rigs, the project reported more than a 30% increase in the number of holes drilled, alongside improvements in drilling accuracy and operational safety.
That makes Sentinel particularly relevant to the economics of digital infrastructure. The mine was not buying the communications system as its productivity product — it was buying the drilling system. But without the network underneath it, the automation could not deliver the intended operating model.
(Related reading: [link to your Smart Equipment pillar page])
Legacy Equipment Is Where the Integration Bill Starts
The newest machine on a mine site is rarely the whole fleet.
Operators often have equipment from different manufacturers, different generations of control systems and software platforms that manufacturers never designed to communicate with one another.
That creates a less glamorous but important part of digital transformation: integration.
Epiroc’s current connectivity strategy explicitly emphasises technology-agnostic solutions and interoperability, including support for different mine setups. In June 2026, Epiroc and Ericsson announced a global agreement to scale LTE and 5G connectivity for underground and surface mining, linking connectivity with automation, situational awareness, collision avoidance and data-enabled machines.
The commercial implication is becoming clearer. Mining companies are not necessarily buying a radio network as a standalone asset — vendors are increasingly packaging connectivity alongside automation, digital platforms, systems integration and managed services.
That can simplify procurement. It can also make vendor selection more strategic: a network that works perfectly for one OEM’s equipment but creates integration problems elsewhere can leave the operator with another technology silo.
The Mine Has to Be Designed for the Network to Move
There is another difference between a mine and a conventional industrial site: the site itself changes.
An office does not normally move its walls several metres every week. A mine does. Open-pit benches advance. Haul roads change. Drill patterns move. Underground headings extend. Infrastructure has to follow the operation.
That makes network planning part of mine planning.
Forsk, which provides mine-network planning technology, describes the challenge as a combination of surface terrain, underground tunnels, remote locations and changing mine geometry. It also notes that underground systems may use radiating cable or distributed antenna systems, while remote surface infrastructure can rely on microwave backhaul.
The practical lesson is that operators cannot design connectivity once and then forget it. The network needs a path for expansion.
Cybersecurity Becomes Part of the Production Conversation
More connectivity also means a larger digital attack surface.
As mines connect drills, haul trucks, sensors, cameras, control systems and remote operations centres, operational technology becomes increasingly dependent on networked infrastructure.
That makes cybersecurity part of the production conversation, and it also makes network resilience important. The question for an autonomous fleet is not only what happens when the network is working — it is what happens when connectivity deteriorates, equipment moves between coverage zones, or part of the infrastructure fails.
Automation systems therefore need appropriate segmentation, traffic prioritisation and fallback procedures. Epiroc’s automation specifications specifically call for dedicated network segmentation and quality-of-service controls for time-sensitive traffic.
For mine operators, that turns network design into a risk-management decision as well as a technology decision.
(Related reading: [link to your ESG / cybersecurity article, if you have one])
The Investment Decision Is Bigger Than 5G
The most important question facing a mining company is therefore not “should we buy 5G?” It is: what digital operating model are we trying to build, and what infrastructure will allow us to scale it?
That changes the investment calculation.
A network built to support one autonomous drill fleet may later carry equipment telemetry, worker tracking, video, environmental monitoring, remote maintenance and other applications. The infrastructure can therefore become a platform for future capital projects.
But there is no universal cost figure that determines whether private LTE or 5G makes economic sense. Public mining deployments generally do not disclose complete project capex, and network cost depends heavily on mine size, terrain, underground conditions, spectrum, backhaul, power, existing infrastructure, equipment integration and required redundancy.
That is precisely why a mine-site readiness assessment matters before procurement. Sedna Africa, which has deployed private LTE networks in mining, says it conducts detailed readiness surveys covering infrastructure, network architecture and operational requirements before deployment.
The survey is not a bureaucratic prelude to the real project. It is part of the project.
What a Mine Needs Before Buying Autonomy
For operators considering autonomous or remote-controlled equipment, the network belongs on the list of the technology project’s prerequisites.
At a minimum, the mine needs to understand:
- Coverage: Where must machines remain connected?
- Mobility: Can equipment move between coverage areas without disruptive handovers?
- Latency: What response time does the specific application require?
- Uplink capacity: Can the network carry the video, telemetry and control traffic the fleet generates?
- Quality of service: Can the network prioritise production-critical traffic?
- Backhaul: Can remote network infrastructure reliably connect to the core?
- Scalability: Can the network accommodate additional machines and applications?
- Legacy integration: Can existing equipment and control systems connect without creating new silos?
- Resilience: What happens when part of the network fails?
- Cybersecurity: How will the mine segment and protect operational technology?
- Mine progression: Can the infrastructure adapt as the pit or underground workings move?
Those questions are more useful than asking which networking technology is newest.
Smart Mining Digital Infrastructure Is Becoming the Production Architecture
The mining industry’s digital conversation has spent years focusing on what happens at the top of the technology stack: AI, autonomy, analytics and digital twins.
The investment reality is moving in the other direction. The more advanced the application becomes, the more important the infrastructure underneath it becomes.
Newmont’s experience at Cadia demonstrates the point in operational terms: replacing a connectivity arrangement that limited remote dozing with private 5G coincided with a 50% increase in dozing capacity and a much larger operating area. First Quantum’s Sentinel deployment shows the same principle from another angle, with a wireless network supporting autonomous drilling that reported a more than 30% increase in holes drilled.
These are individual deployments, not proof that every mine will achieve the same results. But they point to a larger shift: digital infrastructure is moving from being an enabling service behind mining technology to becoming part of the production system itself.
For mining companies, that changes the buying decision. The question is no longer whether AI, autonomy or connected equipment can improve an operation. It is whether the mine has built the connectivity, data architecture and network resilience needed to capture that value at scale.
The industry may still market the smart mine through its autonomous trucks and AI systems. But its competitive advantage could increasingly depend on something much less visible: the smart mining digital infrastructure underneath them.
(Related reading: http://www.skillings.net/build-a-new-private-5g-network-with-partners/ )
