Gritt exits stealth with $34 million for robots to build solar plants—then, everything else

Solar buildout has a labor problem: there just aren't enough hands to install all the panels the world wants deployed. Gritt, a startup founded by two Carnegie Mellon-trained roboticists, thinks the fix isn't more workers but smarter machines, and it just came out of stealth with $26 million in Series A funding led by Obvious Ventures, plus backing from Union Square Ventures and Active Impact Investment. Combined with an earlier seed round, that puts total funding at $34 million.

What's notable is what Gritt isn't doing. Instead of engineering robots from the ground up, the company straps its AI models onto rented, off-the-shelf hardware — skidders and robotic arms from companies like Kawasaki. CEO Puneet Puri argues that the real unlock isn't hardware at all, it's an intelligence layer generalizable enough to handle the messy, unpredictable nature of outdoor construction sites. Rivals like Luminous Robotics, Cosmic, and China's Trinabot are building proprietary robots instead, a divergence that could matter a lot as demand for installation scales and cost structures start to separate winners from also-rans.

The numbers Gritt is touting are eye-catching. Puri says an eight-person crew normally installs around 800 solar panels a day; paired with Gritt's systems, that same crew can hit 3,000 to 4,000. The company currently has two systems running in the field and is already contracted for 2.8 gigawatts of installation work over the next 18 months, with customers including three of the top 10 US power construction firms. It's aiming to get 48 systems operating within six months, which would be a serious jump from two.

A customer who spoke to TechCrunch, on condition of anonymity, was upbeat about more than just speed. Remote sites where recruiting labor is tough get easier to staff, and workers no longer need to repeatedly hoist 100-pound glass panels overhead — a detail that points to fewer injuries down the line, not just faster throughput.

Gritt isn't stopping at panel placement. The plan is to teach the same systems to fasten panels, drill posts, and build the racks panels sit on, then eventually branch into other labor-heavy construction jobs like tying rebar before concrete pours. Puri says the pace of that expansion is the real story: training the system to stack cinder blocks took weeks, but a comparable rebar-tying demo, built on the same underlying pipeline, took just a day. Longer term, the founders picture their sensor-and-AI stack doing more than physical labor — flagging an open trench before a storm hits, or catching missing inventory — turning what started as a panel-installing robot into something closer to a site-wide decision layer.