We shouldn’t be digging the only planet that supports us

Civilization will not stop needing metal, the IEA expects copper demand alone to grow 30% by 2040, with lithium growing fivefold.^[1] The honest question isn’t whether we dig. It’s where. And the long-term answer should not be the one planet with a biosphere.

The resources are there. NASA’s Lunar Reconnaissance Orbiter shows water-ice deposits on the Moon are more widespread than previously thought^[2], one radar survey found at least 600 million metric tons of ice in craters at the lunar north pole alone^[3], and that water splits into breathable air and rocket fuel.^[4] Farther out, the metal-rich asteroid 16 Psyche is estimated to hold iron and nickel with a theoretical value around $10,000 quadrillion, a thought experiment, but one NASA considered worth a dedicated mission arriving in 2029.^[5]

The groundwork is being laid now: the USGS has published its first off-planet mineral prospectivity map, applying a century of terrestrial resource science to lunar ice.^[6] What’s missing is the machine that can work a deposit with no crew, no atmosphere, and no rescue. Nobody is going to operate an excavator in a vacuum. Off-Earth mining is an autonomy problem, which means it’s our problem.

This is why INORE’s Earth business and space ambitions are one program, not two. Every year our fleet runs in terrestrial pits, through dust, heat, comms blackouts, and thousand-hour duty cycles, is qualification testing for regolith. The same autonomy that removes people from the pit eventually removes the pit from the planet: heavy extraction moves to dead rock that was never anyone’s home, and Earth’s remaining mines become the bridge, not the destination.

We are not romantic about this. It will take decades, and it starts in the dirt, in Perth, late this year, with a haul loop and zero people inside the fence. But the direction is set: space is the next frontier of extraction, and the machines that get there will have earned it underground. Read the full vision, or see CERES, the program pointed at it.