ZURICH — August 17, 2026 — Gravis Robotics has raised $200 million from SoftBank in a Series A financing to scale autonomous heavy-machinery systems across construction sites worldwide. The investment values the Swiss company at approximately $1 billion and supports a global rollout of its retrofit control platform.
Gravis does not manufacture excavators. Its Gravis Rack combines sensors, computing and learning-based control so contractors can add assisted or autonomous operation to mixed fleets from established equipment brands. The company says deployments have delivered productivity improvements of up to 30% in selected applications.
Autonomy for machines that reshape their environment
Construction equipment operates in conditions unlike a warehouse robot or road vehicle. An excavator deliberately cuts into soil whose density, rocks and hidden obstacles change with every pass. The control system must estimate forces, adapt the bucket path and maintain safe boundaries while the site itself is being transformed.
Gravis uses cameras, ranging sensors, positioning and hydraulic data to build that operating picture. Its software supports a spectrum from in-cab guidance and hazard detection to remote supervision and full autonomy. The company explains the system and financing on its official Series A page.
Expansion priorities
- Mixed-fleet deployment: support contractors without requiring replacement of existing machinery.
- Global rollout: expand installations with infrastructure companies across multiple continents.
- Engineering: hire specialists in robotics, controls, perception and construction operations.
- Rental access: work with equipment providers so customers can obtain autonomy as part of a standard hire.
- Site intelligence: use operating machines to collect surveying, progress and hazard data.
Capital for physical AI in construction
SoftBank describes physical AI as a central investment theme. Heavy construction offers a large installed base, persistent labour shortages and projects where small productivity gains can affect schedules and capital costs. Retrofitting provides a route to market without waiting for every equipment manufacturer to redesign its product line.
The opportunity also carries demanding safety and support obligations. Contractors need reliable performance across different soil, weather and machine models. Autonomous functions must fit site plans, operator responsibilities and local rules. Downtime on a critical earthmoving sequence can delay multiple downstream crews.
Gravis was founded in 2022 as a spinout from ETH Zurich. The company reports systems deployed with infrastructure partners on four continents. Its leadership includes chief executive Ryan Luke Johns and co-founder and chief technology officer Dominic Jud.
From pilot projects to repeatable deployment
The $200 million financing is intended to move the business beyond isolated demonstrations. Repeatable installation procedures, dealer and rental partnerships, remote support and operator training will determine whether the technology can scale across a fragmented equipment market.
Progress can be measured through productive machine hours, cycle-time improvements, fuel use, safety interventions and the number of equipment types supported. Customers will also assess how quickly a retrofit pays back relative to hiring, purchasing new machinery or continuing manual operation.
Gravis says its systems can be installed on equipment from multiple major manufacturers and adapted through the same core software. The next phase will test whether that flexibility holds as deployments expand across regions and increasingly complex job sites.
About Gravis Robotics
Gravis Robotics is a Zurich construction-technology company originating at ETH Zurich. It develops learning-based control, sensing and fleet software that turns existing heavy machinery into assisted and autonomous robotic systems.
Company contact
Gravis Robotics AG
Zurich, Switzerland
Company enquiries: www.gravisrobotics.com/contact
Website: www.gravisrobotics.com
ADI News
Leave a comment