Bedrock Robotics, the San Francisco startup that retrofits heavy construction machines to work without an operator in the cab, has raised a $600 million Series C at a $4.1 billion post-money valuation. The round, announced on October 26, was co-led by CapitalG, Alphabet's growth fund, and Emergence Capital. It brings Bedrock's total funding to nearly $1 billion and more than doubles the $1.75 billion valuation the company reached in February, making it one of the best-funded bets on autonomous construction and physical AI outside humanoids.
The Round and Its Backers
New investors in the Series C include the US Innovative Technology Fund, Bessemer Venture Partners and Temasek. They join existing backers 8VC, Georgian, Valor Equity Partners, Xora, Eclipse, Perry Creek Capital and Nvidia, among others, according to Robotics & Automation News.
The step-up has been rapid. Bedrock emerged from stealth in July 2025 with $80 million in seed and Series A funding, then raised a $270 million Series B in February 2026. The new round alone exceeds everything it had raised before.
From Pilot to Paid Work
Bedrock was founded in 2024 by former Waymo leaders Boris Sofman, Kevin Peterson and Ajay Gummalla. Rather than building its own machines, the company sells autonomy: its flagship Bedrock Operator system retrofits existing excavators in the 20 to 80 ton class and similar heavy equipment. Contractors keep their fleets and add self-driving capability, an easier sell in an industry where machines are expensive and last for years.
The key milestone came in mid-August 2026, when Bedrock's excavators began doing paid work on live commercial job sites with no operator on board. Customer volume has more than doubled since 2025, the company says. Named customers include:
- Haydon, an employee-owned general contractor in Tempe, Arizona, running fully autonomous earthwork at a remote Texas site
- Wagner Construction in Minnesota and Smith Trucking Company in Florida
- B&S Site Development in Virginia and Native Excavating in Colorado
- Clarkson Construction in Missouri, C.W. Matthews in Georgia and ArchKey Solutions in Arkansas
Haydon plans to use the technology across building and heavy civil work, including advanced manufacturing, water and wastewater, transportation, healthcare and data centre projects, and to test trench work and larger machines.
What the Money Will Fund
Bedrock says it will use the capital to expand autonomous deployments with more contractors and to build the field, commercial and operational infrastructure needed to support machines on active job sites. It plans to extend its reach into the Midwest and South Atlantic and to add new machine types and tasks.
"We have proven that fully autonomous excavators can perform real work of value," Sofman said, adding that the next challenge is getting that capability to more contractors. The company has also added two senior hires with autonomous-vehicle backgrounds: Amy Jones Satrom, formerly of Amazon, Nuro and Uber, as vice president of operations, and Lando Juarez, previously head of legal for Uber's autonomous mobility and delivery business, as general counsel.
The Labour Shortage Case
Investors are framing the deal around a structural shortage of skilled operators. Bedrock cites estimates of about 1.7 million skilled-trades openings a year through 2035, and industry survey data showing that 87% of construction firms have openings for hourly craft roles, with 88% saying those roles are as hard or harder to fill than a year ago.
"We cannot afford to take decades developing advanced robotics for this critical industry," said Derek Zanutto, general partner at CapitalG. Emergence general partner Jake Saper described the challenge as a physical problem rather than a software one. In that framing, an autonomous excavator is less a cost-cutting tool than a way to add capacity when crews cannot be hired.
Why It Matters
Much of the attention in robotics this year has gone to humanoids, yet Bedrock shows how quickly capital is flowing to narrower, task-specific autonomy that can generate revenue today. Construction sites are semi-structured environments with clear, repetitive tasks such as digging, trenching and loading, which makes them a natural next step for self-driving technology developed in robotaxis.
The round also links two booms. The same AI infrastructure build-out that is driving hyperscaler capital spending requires enormous amounts of earthwork for data centres, power and water facilities. If Bedrock can scale its retrofit model safely across more states, autonomous heavy equipment could become one of the first places where physical AI changes how real infrastructure gets built.
