A robotics startup just showed an 8-person crew installing 4,000 solar panels a day instead of 800. It's not staying in solar.

Gritt, a robotics startup that came out of stealth on July 21, raised $32.4 million to attach AI-guided robotic arms to the skid steers and forklifts crews already use on site — turning rented equipment into a machine that places solar panels with millimeter precision. The company says an 8-person crew that normally installs about 800 panels a day places 3,000 to 4,000 a day working alongside its system. It's contracted for 2.8 GW of solar installation over the next 18 months, and it isn't staying in solar.

What exactly did Gritt build?

The pitch isn't a new robot — it's an attachment. Gritt's arms bolt onto equipment a crew is already renting for the job, rather than requiring a fleet of purpose-built machines. That's a capital-light structure on purpose: the company is using off-the-shelf robotic arms (from suppliers like Kawasaki) and rented skid steers instead of manufacturing hardware from scratch, according to reporting on the raise. For a solar EPC, that changes the buy decision from "invest in a robotics fleet" to "add a module to equipment already in the budget."

How big is the productivity claim?

Gritt's own numbers: 800 panels a day for a standard 8-person crew, versus 3,000 to 4,000 a day for the same crew size working with its arms — a 4x-to-5x jump, and the company says its robots have placed tens of thousands of panels without a reported breakage. Those are vendor-reported figures from a single company, not an independent field study, so treat the multiple as a ceiling case from favorable site conditions until it's tested outside Gritt's own contracted jobs.

Why solar first, and what comes after it?

Solar is a forgiving test bed: flat sites, repetitive placement, a single component shape. Gritt's founders — Puneet Puri and Vishal Dugar, both out of Carnegie Mellon's Robotics Institute — are using it to prove the arm before generalizing it. The company says it's already extending the same system into fastening panels, drilling mounting posts, and tying rebar, and naming data centers, bridges, and roads as the next markets for the same bolt-on approach. CEO Puneet Puri's framing of the problem: "Every industry has embraced automation, but outdoor construction sites are exposed, with unforgiving terrain and variable weather. You can't tell a construction site to be less muddy or to stop snowing." That's the actual engineering problem — outdoor sites don't offer the controlled conditions factory robots get — and it's the same problem any trade betting on field automation has to solve, not just solar racking.

Should a GC or sitework sub act on this now?

Not yet, and not directly — Gritt has one customer type today (utility-scale solar EPCs) and no announced product for building trades, civil work, or vertical construction. But the structure of the bet is worth watching if your firm is evaluating any field automation vendor: a bolt-on arm priced against a crew's daily output, running on equipment you're already renting, is a meaningfully lower-risk pilot than buying dedicated robotic hardware. If Gritt's rebar-tying and drilling expansion actually ships and holds up on a real job — not just a demo — that's the version of "construction robotics" worth pressure-testing against your own crew's numbers, the same way you'd pilot any new equipment attachment: side-by-side, on one crew, for two weeks, before committing a schedule to it.

Gritt's approach also lines up with Monumental's robotic bricklaying subs, which raised its own funding round the week prior — a second signal that venture money is betting on physical AI attaching to specific, repetitive field tasks rather than trying to replace a crew wholesale.

Forward this to the person on your team who's still arguing AI is overhyped.