Autonomous Drilling Is Now an Infrastructure Decision, Not a Tech One

Drill equipment sitting idle is not an engineering problem. It is a capital allocation problem. When a mine’s drills run at 72% availability, roughly one day in every four is lost before a single tonne of rock moves.

At Antapaccay, Glencore’s high-altitude copper operation in Peru’s Cusco region, switching to autonomous drilling pushed that figure above 92%. The gap between those two numbers is where the investment case for digital infrastructure actually lives.

Homar Lozano, director of Peru’s Institute of Mining Engineers (IIMP), reported these findings on 10 September 2026, drawing on Antapaccay’s deployment of the ARDVARC autonomous drilling system developed by FLANDERS. The timing matters. Peru’s Think Digital Report 2024 found that 76% of Peruvian mining companies have already started digital transformation, and the broader autonomous mining equipment market is estimated to have reached US$6.2 billion in 2026, up from US$3.1 billion in 2020.

This case study arrives at a moment when the industry is actively arguing about whether autonomous technology is an operational upgrade or a structural shift in how mines are built and run. Antapaccay is the anchoring evidence for a sharper question: when autonomous drilling produces world-class availability figures, does that make digital technology an infrastructure decision rather than a technology one, and what does that distinction mean for operators, investors, and the communities watching the sector evolve?

What the Antapaccay numbers actually show

Start with the raw gap. Antapaccay’s autonomous rigs recorded availability of 92% and 93%, against 72-83% for the manual drills they replaced. Utilization, the share of available time actually spent drilling, reached up to 89% during the reported periods.

Those figures mean little without a frame, so here is the frame. Under the Global Mining Guidelines (GMG), a standardised time-classification framework used to compare equipment performance across operations, overall mining equipment availability typically sits between 85% and 95%. For open-pit drill fleets specifically, world-class availability is recognised at roughly 91%, industry average sits around 80%, and the broad target is 88% or above. Caterpillar benchmarks large mining machines at 88-92% physical availability and set a utilization-of-availability target of 90%.

The structural nature of this problem is well-documented: the mining productivity gap has widened over decades as ore grades decline, haul distances extend, and manual operations struggle to sustain consistent equipment utilization, which is precisely the context that makes Antapaccay’s autonomous utilization figure of 89% so significant against an industry average of 61-65%.

Place Antapaccay against that distribution and the autonomous rigs are not merely good. They are at or above what the industry formally defines as world-class.

The utilization number is the harder achievement, and it is the one worth pausing on. Availability measures whether the equipment can run. Utilization measures what the operation actually does with that time, and average open-pit operations frequently convert only 61-65% of available time into productive drilling, with broader fleets reaching 70-85%. Antapaccay’s 89% puts it near the ceiling that even world-class OEM benchmarks aim for.

The adoption of autonomous drilling at Antapaccay resulted in equipment availability of 92% and 93% for the autonomous rigs, compared with only 72-83% for manual drills, according to Homar Lozano, director of the Peruvian Institute of Mining Engineers, reported 10 September 2026.

The precision underpinning those figures is worth noting: ARDVARC delivers hole-position accuracy of 0.5 metres in the X and Y axes and hole-angle accuracy of plus or minus 1 degree, which FLANDERS links to a 20-30% productivity gain through reduced between-hole time and the elimination of over-drilling. For anyone assessing the operational credibility of digital mining claims, these are the numbers that matter. They are real fleet figures from a live operation, benchmarkable against GMG and OEM standards rather than vendor projections, and the read you should take is that the gap between digital and manual here is the gap between average and exceptional.

How ARDVARC works and why the architecture matters

What the operator sees on screen is a hole position, an angle, and a rig that can execute it without anyone in the cab. What sits behind that screen is a spectrum of automation, not a single on-off switch.

ARDVARC operates across three levels:

• Drill Guidance: the operator drills manually while the system provides high-precision positioning and navigation.
• One Touch: semi-autonomous operation where the operator initiates a sequence and the system completes it.
• Full Auto Autonomous: the rig drills without an operator on the machine, remotely supervised.

Under remote control, a single operator can manage up to five or eight rigs depending on the configuration, which is where the labour and throughput mathematics of autonomy start to bend in the operator’s favour. Antapaccay began deploying ARDVARC in late 2022 and early 2023, and later extended it to electric rigs to achieve a global first: three autonomous, electric blasthole rigs operating concurrently. That milestone links the productivity story to the electrification story, because it demonstrates autonomy and decarbonisation advancing on the same fleet rather than competing for capital.

Autonomous surface drilling as a category has matured significantly beyond single-site pilots: the safety data, remote supervision ratios, and accuracy specifications documented across multiple open-pit deployments now provide a statistical base that individual site results alone cannot establish.

The retrofit advantage and what it changes for fleet economics

Here is the design choice that matters as much as any performance figure. ARDVARC is OEM-agnostic, meaning it retrofits onto drill rigs from multiple manufacturers rather than requiring a specific brand.

In September 2022, FLANDERS developed a digital interface with Sandvik iSeries drills that enables plug-and-play operation across multi-OEM fleets without modifying the rig itself. That detail changes the capital calculus entirely. A mine considering autonomous drilling does not have to face a full fleet-replacement decision to access it, which is precisely the barrier that keeps proprietary OEM autonomy out of reach for many operations.

This is what separates a niche solution from a scalable infrastructure layer. The demand behind the autonomous mining equipment market reaching US$6.2 billion in 2026 reflects operators who need flexible, retrofit-capable systems rather than locked-in, single-brand commitments. For readers assessing whether Antapaccay’s results are replicable elsewhere, the retrofit capability is the key variable: it is the feature that determines whether these numbers travel to other sites or stay tethered to this one.

From operational case study to infrastructure thesis

One mine is a data point. The question is whether the pattern holds elsewhere, and the comparator evidence suggests it does, across different operators, geographies, and system architectures.

At Minera Los Pelambres in Chile, Epiroc conversions delivered roughly a 10% increase in operational drilling speed, well-depth accuracy above 96%, and an autonomous utilization goal of 55% hit ahead of schedule. At Quellaveco in Peru, Anglo American runs six remote-controlled blasthole drills capable of drilling a 16-metre hole in 12-15 minutes. Vale ordered 16 AutoMine-ready Sandvik surface drills for its Salobo and Sossego copper operations in Brazil.

The Utah entry on that table is the one that changes the category. Mariana Minerals’ Copper One greenfield integrated AutoMine Surface Drilling directly into the initial mine design rather than retrofitting it later. When a mine designs autonomous drilling into its layout before a single drill has turned, the technology has stopped being an optional upgrade. It has become foundational infrastructure.

Across these operations, autonomous drilling serves three strategic functions:

1. Continuous throughput: removing shift-change downtime to sustain 24/7 operation that human-crewed fleets cannot match.
2. Safety and risk removal: pulling operators out of the cab in high-altitude, remote, or unstable-ground conditions.
3. Data platform: generating continuous machine-health and production data that feeds predictive maintenance and precise downstream planning.

The World Economic Forum estimates that wider use of autonomous machines could create US$56 billion of additional global value. That figure is best understood not as a forecast but as the aggregate of what individual operations like Antapaccay are already demonstrating at unit level. For anyone allocating capital in mining, as operator or investor, the consistency of these results is the point. The pattern now carries enough weight to reframe how the next investment decision should be evaluated.

The forces that complicate a straightforward adoption story

The optimistic reading is real, but it is not the whole picture, and the parts that resist it do not resolve cleanly.

Start with jobs. Rystad Energy projections suggest up to 400,000 roles could be replaced across North America, Europe, and Russia by 2030, though that figure comes from unverified source research and warrants caution. EY estimates workforce reductions of 40-80% at individual automated mines and 40-60% industry-wide. Set against those numbers, a Peru-specific study by the Centre for Social Responsibility in Mining at the University of Queensland (CSRM/UQ) found something quite different.

The CSRM/UQ study found that digitalization has so far had little impact on workforce size in Peru, with no union claims of job losses directly tied to process automation, and forecasts a moderate shift in skills rather than mass displacement.

This is not a data conflict to be resolved. It is a signal that outcomes depend on context: regulatory environment, implementation approach, and whether operators fund workforce transition alongside the hardware. The global projections and the Peru finding can both be true because they are measuring different things in different places.

Workforce constraints in drilling have become a structural supply-side ceiling for operations relying on manual crews, which reframes the jobs displacement debate: in markets where skilled drill operators are already scarce, autonomous systems may not be displacing available labour so much as substituting for labour that cannot be sourced at the required scale.

Three complicating factors sit underneath the headline story:

• Workforce displacement uncertainty: global projections and site-level evidence diverge sharply.
• Capital access barriers: the upfront cost of autonomy is not evenly affordable across the industry.
• Social licence erosion: centralised remote control can distance operations from the communities that host them.

The third factor deserves weight. Establishing remote operation centres hundreds of kilometres from a mine centralises control and can reduce local employment. Framed purely as cost-cutting, that shift can erode community trust. Antapaccay’s own 2024 geotechnical event, which caused mine-sequence delays and contributed to a 16% year-on-year fall in copper output to 145.8 kt, is a reminder that autonomy does not remove ground-condition risk. It relocates the operator, not the hazard.

Capital access and the two-tier adoption risk

The capital barrier is where the industry may split. Autonomous drilling demands infrastructure-level spending on hardware, software, communications networks such as LoRaWAN, and organisational change management.

Majors like Glencore, Codelco, and Vale can finance this transition at scale. Mid-tier and junior operators generally cannot match that pace, and the gap widens as autonomy becomes a baseline expectation rather than a differentiator. What began as a competitive edge is hardening into a threshold of entry.

Capital project viability assessments for mid-tier operators now have to account for the widening competitive gap that autonomy creates, because the cost of not deploying autonomous systems compounds over time as productivity differentials accumulate against majors that have already crossed the availability threshold.

There is an ESG layer to this. Access to institutional capital increasingly depends on operational transparency and digital monitoring capability. Operators who cannot afford the digital transition may find themselves disadvantaged not only operationally but in their access to responsible capital, which compounds the two-tier risk rather than easing it.

What the Antapaccay results mean before the next capital decision

The evidence points one way. The useful question now is what a rational operator or investor should watch next, and Glencore has provided the most immediate test itself.

The Coroccohuayco expansion, an estimated US$1.5 billion investment targeting roughly 250 kt of copper per year in its first decade and extending mine life to at least 2045, is where the Antapaccay results either carry forward or stay standalone. If Coroccohuayco is designed from the outset with autonomous drilling as baseline infrastructure rather than a later retrofit, that will be the clearest signal yet that Antapaccay’s numbers have altered Glencore’s own capital planning logic. The restart of the oxide leaching circuit in Q3 2025 shows the operation already layering incremental production recovery onto digital investment.

Three variables will determine whether the infrastructure thesis holds across the sector:

• Mid-tier scalability: whether operators below the majors can finance autonomy at comparable pace.
• Social licence resolution: whether the community and workforce questions can be settled in high-dependency geographies.
• Autonomy-first greenfield design as default: whether Copper One’s approach becomes the norm rather than the exception.

Digital governance as the connective tissue

The drilling story does not stand alone. The same digital infrastructure that enables autonomous rigs also enables real-time tracking of water quality, air quality, and energy consumption, which makes the governance case inseparable from the operational one.

Environmental-monitoring vendors including Eco Detection, 2SWater, KETOS, and Envirosuite now provide high-frequency, in-situ water-quality alerts across tailings and drainage sites, proving compliance in real time rather than through delayed periodic reports. A World Bank-supported digital one-stop shop for mining permits in Peru shows regulators beginning to embed digital capability into the licence-to-operate framework itself.

With 76% of Peruvian mining companies already in digital transformation per the Think Digital Report 2024, the direction is set. The signal worth monitoring is not whether autonomy works. Antapaccay has answered that. It is whether the infrastructure logic, and the US$56 billion of value the WEF attaches to it, replicates beyond the operators who can already afford it.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Past performance does not guarantee future results. Financial projections and forward-looking statements are speculative, subject to market conditions and various risk factors, and may change based on developments in the sector.