How Gravis Robotics is Reshaping the Physical World with Its Autonomous Earthmoving Technology

Gravis Robotics, a Zurich-based robotics company turning any earthmoving machine into a robot, led by Ryan Luke Johns, Dr. Dominic Jud, Burak Cizmeci, and Marco Tranzatto, has raised $200 million in Series A from Softbank. As the largest Series A in construction robotics history, the funding accelerates Gravis’s mission to advance physical AI and scale autonomous heavy machinery across jobsites worldwide.

Global infrastructure is undergoing a massive expansion, driven by demand for energy networks, AI data centers, housing, transit, and climate-resilient infrastructure. Construction has become a critical bottleneck, constrained by a severe labor shortage and limited automation. Heavy construction remains one of the world’s least automated industries, relying largely on machinery based on decades-old technology.

Addressing this gap requires translating advanced AI into precise, real-world physical execution. That is the challenge Gravis is building to solve.

The Gravis Approach: AI Built to Change the World, Not Just Navigate It.

Founded in 2022 as an ETH Zurich spinout, Gravis is bringing software-defined intelligence to an industry long dominated by mechanical machinery. The company is developing autonomous robotic systems that turn heavy construction equipment into intelligent machines.

Most physical AI operates in relatively controlled environments, navigating around obstacles or manipulating objects without significantly changing the surroundings. Construction machinery presents a fundamentally different challenge: excavators actively reshape the environment, breaking through variable soil, rocks, and other unpredictable terrain with every movement.

Gravis addresses this challenge with AI models trained on extensive simulated environments designed to bridge the sim-to-real gap. Billions of cubic yards of virtual terrain, from soft clay to rock-filled soil, can be processed in simulation, helping the technology bring greater precision to unpredictable construction sites.

Rather than replicating the behavior of individual operators or machines, Gravis’s AI world model incorporates the performance characteristics of equipment from multiple manufacturers, creating a more adaptable and generalizable approach to autonomous heavy machinery.

“To build the future, the physical world has to change. Whether it is housing, energy infrastructure, or data centers, every project starts with moving earth. That foundational work has become a bottleneck for the broader built environment. Gravis is building machines designed for the unpredictable reality of live jobsites, where traditional automation often falls short. With SoftBank’s backing, the company can accelerate hiring, deploy Gravis-powered autonomy across major jobsites, and scale faster than previously possible.”

— Ryan Luke Johns, CEO and Co-Founder, Gravis Robotics

How Gravis Works.

The heavy equipment market is highly fragmented, with roughly two-thirds of global demand outside the top three manufacturers. Contractors often choose machinery based on regional service relationships and existing fleet investments, making a closed, single-brand ecosystem impractical.

Gravis acts as an operating system for mixed fleets, with its software running on the Gravis Rack autonomous control kit. The system has been deployed across machinery from Caterpillar, Case, Develon, John Deere, JCB, Hitachi, Sumitomo, Yanmar, Volvo, and others. Learning-based models adapt to the characteristics of different machines, allowing the same core software to operate equipment ranging from compact excavators to machines five times larger without custom reprogramming.

This machine-level understanding enables up to a 30% increase in jobsite productivity compared with peak manual operation, while also improving worksite safety.

The software supports a spectrum from AI-assisted manual operation to full autonomy. Gravis Copilot provides operators with real-time 3D guidance and hazard detection from inside the cab, while full autonomy allows operators to supervise multiple robotic machines remotely. Across both modes, every Gravis-equipped machine also functions as a continuous site sensor, automating surveying and hazard mapping as work progresses.

The Road Ahead: What Today’s Funding Unlocks.

With systems already deployed alongside global infrastructure leaders across four continents and proven across diverse machinery and jobsites, Gravis is positioned to scale rapidly.

The new investment will support a global rollout, bringing physical AI to construction sites and embedding autonomous capabilities into heavy equipment fleets across a trillion-dollar industry constrained by labor shortages.

Gravis is also leading an $8 million UK government-backed CAM Pathfinder project with Flannery Plant Hire, the country’s largest provider of operated heavy-equipment rentals. The partnership will retrofit excavator fleets with the Gravis Rack, giving contractors on-demand access to autonomous machinery and supporting Britain’s $716 billion infrastructure pipeline.

“Skilled operators interpret subtle physical feedback, engine strain, machine vibration, and hydraulic resistance to understand changing ground conditions. Gravis’s AI combines those physical signals with machine telemetry, enabling it to respond to variations in soil and subterranean forces at microsecond speeds. Rather than simplifying the complexity of real jobsites, the technology is designed to understand and respond to it with a level of precision beyond what can be perceived from inside the cab.”

— Dominic Jud, CTO and Co-Founder, Gravis Robotics

𝐑𝐞𝐦𝐨𝐭𝐞 𝐇𝐢𝐫𝐞 𝐖𝐢𝐭𝐡 𝐔𝐬: 𝐌𝐞𝐧𝐥𝐨 𝐓𝐚𝐥𝐞𝐧𝐭 helps technology companies hire exceptional remote talent from India. Build your team with carefully curated engineers, product leaders, designers, GTM professionals, and more. Learn More At: https://www.menlotimes.com/menlo-talent