Ex-Tesla Engineer’s Startup Automates Copper Mine with Pronto

While domestic manufacturing commands significant focus, former Tesla engineer Turner Caldwell argues the foundational minerals and metals supply chain remains overlooked. This belief drove him to launch Mariana Minerals in 2024, a startup aiming to modernize and expand mining operations through comprehensive automation. His primary objective is straightforward: increase the flow of refined metal into the global economy.

A major step toward that goal involves vehicles. This week, Mariana announced a partnership with Pronto, a developer of self-driving systems for off-road haulage trucks. This marks Pronto’s first deal since its acquisition by Atoms, the robotics venture led by Uber co-founder Travis Kalanick. The move reunites Kalanick with Pronto founder Anthony Levandowski, a pivotal figure in the early development of autonomous vehicle technology.

Under the agreement, autonomous haulage trucks will begin operating next week at Copper One, a Utah copper mine Mariana purchased last year. The financial terms were not disclosed. For Caldwell, however, the partnership extends beyond simply deploying driverless trucks. Pronto’s autonomy system will integrate directly into Mariana’s proprietary MineOS software platform. This integration will enable the fully autonomous dispatch and routing of trucks without human intervention, a key component of Caldwell’s vision for a modern mine.

He envisions a network of interconnected operating systems using reinforcement learning to automate and eventually coordinate all aspects of a mining operation. Caldwell draws a direct parallel between today’s major mining firms and legacy industries disrupted by technology. “The big Western mining companies look exactly like Ford and GM before Tesla. They look a lot like NASA before SpaceX,” he said. He argues that innovation in mining is stifled because operational teams, meeting their key performance indicators with traditional methods, lack incentive to adopt new technologies.

This stagnation, in his view, caps production and ignores major efficiencies. Caldwell also sees it as an existential threat. “Because Western mining companies don’t build a lot of net new infrastructure, the talent pool hasn’t been actively attracted to it, and so the labor force is diminishing,” he explained. The result is an industry forced to do more with less, a problem Caldwell believes Mariana’s software-first approach can solve.

Success could create opportunities beyond Mariana’s own operations. Caldwell acknowledged that selling the company’s coordination software to other mines is a future possibility once the system is proven. His immediate focus, however, is on production. “The core business should be selling metal,” he stated. He compares the strategy to SpaceX, noting it would not have become a major company by merely selling rocket landing software to NASA.

Owning and operating the mine is also critical for the reinforcement learning loop. It provides superior control and higher-fidelity data, which could eventually uncover operational insights invisible to human managers. Caldwell likened this potential to DeepMind’s AlphaGo, which famously developed novel chess strategies after extensive data training.

Despite the emphasis on automation, Caldwell is not seeking to eliminate human roles. He contends that Mariana’s model will actually help expand the shrinking mining talent pool. “Part of this is a labor cost reduction, but that’s not really the goal,” he clarified. “The goal is actually enabling more productivity with the constrained labor pool that we have. Automation, and autonomy, is going to create more jobs, because we will have more mines that are operating.”