AstroForge

DeepSpace-2 is a 200 kilogram spacecraft built to fly about twenty million kilometres out, rendezvous with a metallic near-Earth asteroid and photograph it well enough to establish its shape, structure and composition. Hall thrusters running on sixty kilograms of xenon give it more than five kilometres per second of delta-v, an NVIDIA Jetson handles avionics, and it can operate alone for two years. AstroForge wants platinum-group metals specifically, and has been explicit about refusing the water-ice-for-propellant business most asteroid mining companies pitched, on the grounds that nobody is buying it.

The real argument is about cost. Odin, the company's second spacecraft, was built in under ten months for about $3.5 million, roughly seven per cent of what NASA spent on a comparable small deep-space mission. Matt Gialich has been unusually willing to say what that buys and what it does not. After Odin was lost, AstroForge published a debrief admitting the mission had been given a thirty per cent chance of success before launch, and listing the defects the team knowingly flew with.

Two missions, two failures, no asteroid reached. Brokkr-1 was defeated in 2023 by communications trouble and magnetic interference with its own attitude system. Odin launched in February 2025, made contact at about 124,000 miles, and was gone within roughly a day; the leading explanation is a solar array that never fully deployed, compounded by a ground segment that failed almost completely, with the most reliable signal of the mission coming from an unpaid amateur radio operator in Germany. Odin is still on the trajectory it was given, out of contact. The third mission has quietly shrunk as well. It was announced as Vestri, a lander that would clamp magnetically onto an asteroid and demonstrate mining and refining, and DeepSpace-2 does not land. It takes pictures. About $56 million has been raised in total, which is both why the spacecraft are cheap and why they are fragile.