Machines Learn to Drill Without Us

Sandvik and Rio Tinto announced this week they're jointly developing autonomous drill rigs that combine Rio Tinto's remote operations experience with Sandvik's AutoMine technology, starting with support drilling in Western Australia's Pilbara region . The machines will be controlled from Rio Tinto's Perth Operations Centre—hundreds of kilometers from the actual drilling. Testing begins at Sandvik's facility in Finland before moving to Rio Tinto's Australian sites .

This isn't a pilot. Autonomous drilling systems have already demonstrated productivity improvements of up to 30 percent compared to manual operations over more than a decade of deployment, according to Worley . What's changed is the scale. As of January 2026, AI has transitioned from a niche luxury for top-tier miners to a foundational operational requirement across the global sector, AzoMining reported . The question is no longer whether automation works—it's who controls the electrons powering it.

Can Mines Run Without People on Site?

Modern sites adopting end-to-end autonomous workflows are reporting up to 40% productivity gains, owing to fewer unplanned shutdowns, faster cycle times, optimized ore haulage, and higher equipment uptime . The gains come from consistency, not speed. Autonomous systems can operate around the clock without breaks, leading to higher productivity and faster drilling times , but the real value lies in precision.

Navigation in underground settings, where GPS signals are unavailable, depends on inertial navigation systems and LiDAR-based odometry—gyroscopes and accelerometers track position and orientation, providing accurate localization in GPS-denied environments . These systems integrate AI and machine learning to execute drilling cycles, using data from mine planning software like iSURE to optimize drilling patterns and adapt to real-time conditions .

The technology is maturing fast. The Asia-Pacific region holds a 40% share of the AI in mining market as of early 2026, driven by China's massive investment in "smart coal mines" and Australia's leadership in autonomous extraction . But deployment isn't frictionless. A reliable, high-bandwidth wireless network such as LTE is non-negotiable—it's the backbone for low-latency data transmission between equipment and control systems, along with high-precision navigation requiring a stable Global Navigation Satellite System supported by local RTK base stations for accurate hole placement .

What About the Energy to Power the AI?

Here's the paradox: AI is automating energy extraction while consuming unprecedented amounts of energy itself. Global electricity demand from data centers grew by 17% in 2025, while electricity consumption from AI-focused data centers surged 50% , the IEA reported. Capital expenditure from just five technology companies exceeded $400 billion in 2025 and is expected to jump by another 75% in 2026—now larger than global investment in oil and natural gas production .

The constraint is no longer capital. It's control over electrons—the quarter AI infrastructure became constrained by energy , according to analysis from Global Data Center Hub. Morgan Stanley Research forecasts U.S. data center demand could reach 74 GW by 2028, with a projected shortfall of about 49 GW in available power access . That's roughly equivalent to adding 49 nuclear power plants in two years.

Energy companies are responding by embedding AI into their own operations. AI-powered digital twins and predictive analytics tools are moving from niche applications into core operational use across factories and energy assets, as artificial intelligence enters a more operational phase in 2026 , Hanwha reported. Analyses of predictive maintenance deployments cite reductions of roughly 35% in unplanned downtime, 20% in maintenance costs, and measurable increases in asset uptime when AI is used to flag issues before equipment fails .

The applications are getting more sophisticated. AI-equipped satellites and sensors can detect incidents in critical energy infrastructure 500 times faster than traditional ground-based methods and at high spatial resolutions , the IEA noted. AI systems continuously monitor the health of critical energy infrastructure and better plan for maintenance needs—this proactive approach can reduce equipment downtime by up to 50% and lower maintenance costs by 10% to 40% , according to Brookings Institution research.

Where Do Large Language Models Fit In?

Beyond drilling and predictive maintenance, LLMs are starting to reshape how energy companies interact with geospatial data. Recent research presented an approach integrating Large Language Models, specifically GPT-4 and the open-source DeepSeek-R1, into Geographic Information System workflows to enhance the accessibility, flexibility, and efficiency of spatial analysis tasks—designing a system capable of interpreting natural language instructions and translating them into automated GIS workflows through dynamically generated Python scripts .

In experiments where 10 GIS analysts used QGIS to perform identical tasks, the automated system not only matched the accuracy and quality of manual methods but also outperformed them in speed, efficiency, and accessibility . The combination of Spark, improved algorithms, and agent systems with NLP significantly speeds up the selection of plots for renewable energy sources, supporting sustainable investment decisions .

The use cases extend across the value chain. AI-powered workflows automate the management of contracts, procurement, and regulatory filings, adapting instantly to new requirements and reducing the risk of errors or missed deadlines , Microsoft noted in a recent analysis of agentic AI in renewable energy operations.

What Changed This Week

The Sandvik-Rio Tinto partnership signals a shift from proprietary systems to interoperable platforms. Under the agreement, Sandvik and Rio Tinto will co-develop the interoperability and autonomous capabilities required for remote, multi-rig and multi-site autonomous operation via Rio Tinto's Perth Operations Centre . That matters because it suggests the industry is converging on standards rather than fragmenting into competing ecosystems. Meanwhile, AI's energy appetite continues to reshape infrastructure investment— by the first quarter of 2026, data centers had fully transitioned from digital infrastructure into integrated energy and compute platforms .

What to Watch

The Sandvik-Rio Tinto program will include field trials to validate performance against production targets, beginning with development and on-rig testing at the Sandvik Test Pit in Finland, followed by site-based testing at Rio Tinto operations in Western Australia . Watch for results from those trials in the second half of 2026—they'll indicate whether multi-vendor autonomous systems can match the performance of proprietary platforms.

On the energy side, the IEA is expected to release updated projections on data center electricity consumption later this quarter. Close monitoring, frequent updates and cooperation with the tech sector, including more systematic energy consumption disclosures, will remain important to improve the robustness of the outlook for AI's energy demand . The gap between AI's promise to optimize energy systems and its own voracious energy consumption remains the industry's central tension.