SoftBank Backs Gravis Robotics With $200M Series A for AI-Powered Construction Automation

SoftBank Leads Historic $200 Million Series A for Construction Robotics

Gravis Robotics, an ETH Zurich spinout, has secured a $200 million Series A led by SoftBank, marking the largest funding round in construction robotics history1

1

. The deal reportedly values the Zurich-based company at $1 billion post-money, granting it unicorn status just four years after its 2022 founding2

2

. SoftBank was the sole investor in the round, confirming earlier reports that Masayoshi Son's conglomerate was circling the young company with potential plans for a multi-stage acquisition4

4

.

The investment fits SoftBank's broader strategy to dominate physical AI, following its reported talks to back an $800 million round for Germany's Agile Robots and its acquisition of ABB's industrial unit last year1

1

. Son has positioned physical AI as the next trillion-dollar opportunity, establishing Robo HD as a new holding company and planning another robotics firm called Roze in the U.S.4

4

. "Physical AI is central to SoftBank's vision for the next phase of AI. Gravis is bringing AI-powered autonomy to construction, helping build smarter, more efficient infrastructure," said Dai Sakata, Managing Director of SoftBank Group5

5

.

Retrofit Technology Transforms Existing Heavy Machinery

Gravis Robotics doesn't build excavators—it builds the brains that make them autonomous. The company's flagship product, Gravis Rack, is a retrofit control kit that bolts onto existing earthmoving equipment from manufacturers including Caterpillar, John Deere, Volvo, JCB, and Hitachi1

1

. This manufacturer-agnostic approach allows construction firms to automate their fleets without scrapping existing machinery, addressing a critical barrier to adoption in an industry notorious for capital-intensive equipment investments.

The Gravis Rack is a flat, rectangular computing appliance mounted on the excavator's roof, housing AI chips, cameras, and lidar sensors. Additional cameras and lidar modules attach to rear vehicle masts, along with a GPS module enhanced by GNSS RTS technology that boosts location accuracy to under an inch3

3

. Each excavator also receives a Wi-Fi transmitter that syncs data to the Gravis Copilot—a companion Slate tablet enabling remote operation with real-time sensor footage and geographic overlays highlighting infrastructure like water pipes3

3

.

The system spans a full spectrum from AI-augmented manual control to complete autonomy, allowing operators to co-pilot from the cab or supervise entire robotic fleets remotely from safer environments2

2

. Gravis claims its technology delivers up to 30% productivity gains compared to peak manual operation while improving worksite safety5

5

.

Training AI to Read Earth Through Physical Feedback

Autonomous heavy machinery faces technical challenges absent in other robotics applications. Unlike autonomous vehicles or warehouse robots that navigate static environments without disturbing their surroundings, excavators must actively reshape terrain—digging through soil, subterranean rock, and unpredictable ground conditions3

3

. Skilled human operators read the earth through subtle physical feedback: engine strain, machine vibration, hydraulic resistance. Gravis trains its AI models heavily in simulated environments to replicate this sensory experience and close the notorious sim-to-real gap.

"Our AI takes that same physical input and grounds it in machine telemetry, responding to varying subterranean forces and soil mechanics at microsecond speeds," explained Dominic Jud, CTO and co-founder. "We didn't try to simplify the world for our software—we gave it the physical intuition to handle real jobsites with precision that goes beyond what any human can feel from inside the cab"2

2

. This approach addresses what Gravis frames as construction's core challenge: teaching machines factory-floor precision on muddy, unscripted sites1

1

.

Addressing Labor Shortages in a Trillion-Dollar Industry

Construction remains one of the world's least automated major industries, still relying on machines that operate exactly as they did 50 years ago2

2

. The sector faces severe labor shortages and an aging workforce, creating urgent demand for infrastructure projects—housing, energy grids, data centers—that construction firms struggle to meet on schedule3

3

. "Whether we are building housing, modernising energy grids or scaling data centres, every project starts with moving earth," said Ryan Luke Johns, CEO and co-founder. "Our machines are built for the messy, unscripted reality of live jobsites that breaks traditional automation"2

2

.

Gravis has deployed systems across four continents and was recently selected to lead an $8 million U.K. government-backed CAM Pathfinder project with Flannery Plant Hire, retrofitting excavator fleets to support Britain's $716 billion infrastructure pipeline5

5

. The company maintains offices in Zurich, Oxford, and Austin, positioning itself to scale autonomous heavy machinery deployments globally2

2

.

Questions Around Safety, Regulation, and SoftBank's Track Record

While the funding signals conviction, challenges remain. Autonomous heavy machinery carries substantial burdens around safety, liability, and regulation that capital alone cannot resolve overnight1

1

. The claimed 30% productivity boost reflects controlled site performance, not necessarily fleet-wide averages in adverse weather conditions. SoftBank's robotics record includes both hits and misses—its consumer robot Pepper was quietly discontinued, and Son's habit of writing enormous checks has produced spectacular outcomes in both directions1

1

.

The construction automation market has promised full autonomy for years while delivering it incrementally. Whether Gravis can convert this $200 million vote of confidence into excavators that reliably operate themselves across thousands of messy sites will determine if this becomes another SoftBank success story or cautionary tale. What's undeniable: the money chasing physical AI is real and accelerating, with SoftBank positioning itself as the category's biggest backer even as Son pushes back against bubble accusations1

1

. For construction firms watching labor costs climb and timelines slip, Gravis offers a retrofit path to automation that doesn't require fleet replacement—a pragmatic entry point if the technology delivers on its promises at scale.