SoftBank bets $200 million on robots that drive diggers

Key points

• SoftBank invested $200 million in Gravis Robotics' Series A round, which the company claims is the largest in construction robotics history.
• The Gravis Rack, a bolt-on hardware and software kit, has been installed on excavators from ten manufacturers including Caterpillar, John Deere, Volvo and Hitachi.
• Gravis says its system improves productivity by up to 30% compared with a skilled human operator at peak performance.
• Founded in 2022 as a spin-out of ETH Zurich, the company has already deployed its technology across four continents.
• An $8 million UK government contract will put autonomous excavators to work through Flannery Plant Hire on Britain's $716 billion infrastructure pipeline.

A digger doesn't steer around the world. It tears it apart. That basic fact is why construction sites have resisted automation that swept through warehouses, car factories and delivery fleets over the last decade. Wait: that's a triad too. Let me redo.

A digger doesn't steer around the world. It tears it apart. That's why construction sites have resisted automation that transformed warehouses and car factories over the last decade.

Gravis Robotics, a Zurich-based startup that teaches excavators to operate themselves, thinks it has cracked the problem. SoftBank, the Japanese investment giant, just backed that belief with $200 million. We noted SoftBank's recent $55 billion commitment to OpenAI in August; this bet is smaller but pointed at concrete and dirt rather than data.

What does Gravis actually make?

The core product is the Gravis Rack, a retrofit kit of sensors and control hardware that bolts onto excavators a contractor already owns. Once fitted, the machine operates across a spectrum: a human stays in the cab with real-time 3D guidance, or hands full control to the AI, or anything between.

The kit works on machines from at least ten brands, including Caterpillar, Case, John Deere, JCB, Hitachi, Sumitomo, Yanmar and Volvo. That matters because most construction companies run mixed fleets and nobody wants to scrap working equipment to switch software providers. Under full autonomy, one person can supervise an entire fleet rather than sit inside a single cab.

How does it learn to handle unpredictable ground?

Driving a car is hard enough for a robot. Digging is a different problem, because the ground fights back differently every few centimetres, shifting from soft clay to rock-packed soil without warning.

Gravis trained its software in simulation first, moving what it describes as billions of cubic yards of virtual material before the system touched real earth. "Skilled operators read the earth through subtle physical feedback, listening to the engine strain, sensing the machine vibration and reacting to hydraulic resistance," said Dominic Jud, co-founder and CTO of Gravis Robotics, in comments given to The Robot Report. "Our AI takes that same physical input and grounds it in machine telemetry, responding to varying subterranean forces at microsecond speeds." The company reports productivity gains of up to 30% against a peak-performing human, though no independent verification has been published.

What does the SoftBank money pay for?

Gravis plans to hire engineers and expand to more jobsites worldwide. It also recently won an $8 million UK government-backed contract with Flannery Plant Hire, Britain's largest equipment rental provider, as part of a push to accelerate the country's $716 billion infrastructure pipeline.

For construction workers the near-term picture is less about replacement and more about supervision. Someone still needs to read the site and make judgment calls. The real watch-point here isn't whether the AI can dig straight; it's whether insurers and site safety regulators will let fully autonomous machines operate without a human in the cab. That question will shape how fast the $200 million actually moves iron.