Mining’s Tech Revolution: From Point Solutions to Integrated AI
Travis Kalanick has never overseen a mining operation. In 2026, following a $1.7 billion funding round led by Andreessen Horowitz, his venture Atoms acquired Pronto, an autonomous haulage company, and launched a dedicated mining unit. This positions him among the more influential new backers shaping how mines are managed today. In 2022, I argued that the energy transition is essentially a metals transition, requiring mining to undergo its own shale-style revolution to satisfy growing demand. That thesis was grounded in the sector’s inherent limitations—falling ore grades, difficult remote labor conditions, and mounting environmental scrutiny—which collectively make technology highly valuable throughout the physical supply chain.
Three years on, a flurry of industry acquisitions offers the strongest proof that the broader market has reached the same view. Caterpillar CEO Joe Creed remarked at CES earlier this year that embedding industrial AI is about driving progress in the physical world. Coming from the sector’s largest equipment maker, that statement underscores how much the conversation has shifted since 2022. Three categories of acquirers illustrate the dynamic. First, established equipment manufacturers are layering on software to defend their existing customer base and capture higher-margin, recurring revenue across their hardware portfolios. Caterpillar finalized a roughly $733 million deal for RPMGlobal in February 2026, then added Skycatch. Sandvik announced its purchase of ThoroughTec Simulation, which brings in about $17.5 million in high-margin revenue. Second, players in consumables and civil engineering are moving into the same space. Orica’s recent deals, along with Bentley Systems’ acquisition of Seequent, suggest that chemical and geospatial leaders intend to dominate the integrated mine-to-mill digital workflow. Third, venture capital—including Atoms—is entering physical automation directly, viewing heavy industrial automation as a pathway to developing real-world AI models rather than merely licensing another software product.
The capital is clearly flowing, but the more pressing question is whether it is yielding results. Despite hundreds of millions spent on acquisitions, the industry is hitting an operational wall. Miners have amassed numerous standalone software tools for geology, fleet management, and processing that function in isolation, and many are failing to achieve the returns they anticipated due to the absence of a unified data source. Acquiring point-solution software has proven straightforward; getting legacy systems to interoperate in real time is where progress stalls. A few early successes exist—Orica, for example, recently posted a notable earnings boost in its Digital Solutions arm. By effectively integrating major purchases like Terra Insights, the company drove enough recurring digital revenue to lift segment earnings growth to 29 percent. Yet across the wider sector, Orica’s performance is the outlier, and a substantial portion of the capital others have deployed over the past year remains unproductive.
This bottleneck is steering acquirers toward two types of targets next. The first focuses on generating high-fidelity, real-time data at the operational site, since AI models are ineffective without current operational information. Exum Instruments and Minpraxis are both tackling this, replacing weeks-long laboratory turnaround with on-site trace-level physical and chemical analysis. The second involves unified spatial models that consolidate geological, fleet, and environmental data into a single platform, enabling predictive simulations rather than reliance on outdated, fragmented records. Strayos and AiMinr are advancing in this direction. This friction is a common topic in industry discussions. Senior executives at top-tier mining companies, global equipment manufacturers, and major chemical and service providers all cite the same obstacle in off-the-record conversations. As one senior technical director at a major equipment maker put it, everyone recognizes that digital tools and AI are essential for meaningful performance gains, but that awareness has not resolved the issue. That company’s own situation illustrates the challenge: it has invested heavily in point solutions, machine-vision platforms, AI process optimizers, and specialized mine-planning software, each effective in its niche, yet none designed to communicate with the others. Integrating them into a cohesive system has proven far more difficult than purchasing the components.
The first genuine test of whether this approach works is unfolding at a desert copper mine in Utah. Mariana Minerals, supported by Andreessen Horowitz, Breakthrough Energy Ventures, and Khosla Ventures, restarted the long-idled Lisbon Valley copper operation in April 2026, building autonomy into production from day one. The site uses Pronto’s autonomous haul trucks and Sandvik’s autonomous drilling, all managed under Mariana’s own software layer. The premise is that a mine designed around integration from the start can outperform incumbents that are still retrofitting legacy systems. It is premature to declare this a success—the environment is harsh, equipment undergoes significant wear, and scaling a single site to industry-level volumes presents a different challenge than launching one operation. Still, it stands as the most direct live experiment for the notion that controlling the integration layer, rather than merely the tools feeding it, is what truly narrows the productivity gap.
For institutional investors, this multi-pronged consolidation offers a clearer exit landscape than the sector has seen in years. The convergence of heavy industry incumbents, adjacent consumables and engineering firms, and well-funded newcomers ensures ongoing demand for those who control core data assets and orchestration software. A year ago, an investor backing an unproven mining tech startup had few potential buyers to consider; today, that list has expanded considerably. The industry has long debated declining ore grades and slow permitting, but the more immediate constraint may be more fundamental: whether anyone can actually make all this new technology function together. Tem Tumurbat is co-founder of Nomadic Venture Partners, which invests in critical minerals and mining technology.
Recommended posts
Free Data: Insights – World
Instant access. No credit card needed.
