Lessons for technology developers from Satish Rao’s September events in Washington and Houston.
The critical minerals famine will reward the technologies that move fastest from the lab bench to field decisions. Developers who can demonstrate both field performance and commercial value will set the pace for an industry that needs new supply sooner than traditional timelines allow.
Drilling costs an exploration company $150 to $300 per meter, so a geologist commits to a target only when the evidence justifies the spend. Processing raises the stakes further, with a first-of-a-kind critical minerals plant costing $30 million to $50 million to build. At those prices, mining and energy companies need much more than a strong lab result before they will deploy a new technology.
Satish Rao, Managing Director at Clareo, took that message to audiences at two recent events. At the US Department of Energy’s ARPA-E ROCKS program kickoff in Washington, he delivered opening remarks on a panel for grant awardees developing exploration and orebody characterization technologies. ROCKS focuses on the front end of mineral development. At Pilotathon in Houston, he joined fellow panelists from venture capital and manufacturing on a critical minerals panel, speaking to an audience of founders and corporate investors. The Pilotathon discussion spanned the whole value chain, with particular emphasis on midstream processing. The advice to innovators held across both rooms.
Lab performance is only half the battle. Satish told ROCKS awardees that demonstrating hardware or algorithmic performance in a controlled setting is necessary, but accounts for only half the work. Mining and energy companies make the same point from the buyer’s side. A corporate considering a pilot already accepts that the chemistry works in a lab. Its question is whether the technology keeps working at a remote site, in real-world operating conditions, without constant recalibration.
Exploration technology faces the same test underground. A downhole sensor must maintain signal clarity through vibration, temperature, drill mud and rough borehole walls. A detection algorithm must perform against complex, heterogeneous geology without relying on slow or destructive sample preparation. Developers who design for these conditions from the outset reach the pilot stage with evidence an operator can act on.
Integration decides adoption. For processing technologies, Satish pointed to flowsheet fit. A technology that can be retrofitted into an existing processing flowsheet gives a mining company a decision it can make. However, a technology that forces a multi-hundred-million-dollar capital redesign asks the company to rebuild its operation around an unproven tool.
The exploration equivalent is the geologist’s software. Front-end exploration runs as a sequence of decision gates, from regional target selection through the drill decision to resource estimation, with each gate dependent on data and a reduction in uncertainty. Sensing technologies must produce structured digital data that feeds directly into geological modeling software and block models without creating isolated data silos. Satish’s first recommendation to ROCKS awardees was to build that integration from day one.
Technical and commercial validation must run in parallel. Satish urged ROCKS awardees to use the grant period to test unit economics, pricing structures and field deployment logistics alongside prototype development. He set a specific commercial proof point. The technology must measurably lower the total capital spent per verified drill target or reduce the in-fill drill density a deposit requires. Front-end exploration relies on patient capital that demands rapid risk reduction, so business models such as Sensing-as-a-Service, low-footprint mobile kits and shared-risk pilot gates suit lean exploration budgets better than a large up-front capital purchase.
The capital challenge is more acute in processing. In software, scaling means adding server capacity. In critical minerals processing, it can mean building a first-of-a-kind plant costing $30 million to $50 million. At Pilotathon, Satish described startups caught between traditional venture capital, which avoids high capital intensity and long development cycles, and project finance, which will not fund unproven technology without years of continuous commercial operating history. Corporates co-hosting a pilot want a clear pathway in which non-dilutive government grants, such as DOE or Department of War (DoW) funding, or strategic corporate venture funds, share the pilot expense. Closing that gap requires co-hosting models that bring founders, corporate innovation teams and system builders together.
From pilot to adoption
For front-end exploration, Satish set a practical test: data quality and speed must be reliable enough for a site geologist to change a live drilling or target-elimination decision while the rig is still on site. Off-site laboratory assays currently take weeks or months to return, and that lag stalls the decisions exploration depends on. Mine operators and exploration geologists carry severe downside risk and adopt unproven tools cautiously. Validation across diverse field conditions builds the second-mover trust that turns a single successful pilot into industry-wide adoption.
