Gritt exited stealth with $32M in funding for robots that build solar farms faster than human crews. The company was founded by Carnegie Mellon roboticists, and three of the ten largest U.S. power construction firms already use its systems. The round was announced on July 21, 2026.
Key takeaways
- Funding: $26M Series A, $32M total including an earlier seed round
- The Series A was led by Obvious Ventures, with Union Square Ventures and Active Impact Investment
- A crew using Gritt installs 3,000–4,000 panels a day versus 800 without automation
- The company is contracted to install 2.8 GW of capacity within 18 months
- Founders: Puneet Puri (CEO) and Vishal Dugar (CTO), both from Carnegie Mellon
How Gritt's robots work
Gritt does not build its robots from scratch. Instead it pairs off-the-shelf hardware — rented skidders?skidders: Heavy machines for dragging and hauling large loads, here rented and AI-controlled rather than operator-driven. and robotic arms — with its own AI models. That setup gets it onto the job site faster and avoids the cost of manufacturing hardware.
The systems unload panels, move them, and place them on mounting frames with sub-millimeter precision. As reported by TechCrunch, an eight-person crew using Gritt installs 3,000–4,000 panels a day. A manual crew of the same size installs about 800.
Funding and scale
The $26M Series A was led by Obvious Ventures. Union Square Ventures and Active Impact Investment joined. The earlier seed was backed by First Round Capital, Climactic, Congruent Ventures and VSC Ventures, bringing the total to $32M.
The company has two systems in the field today and is contracted to install 2.8 GW of solar capacity within 18 months. Within six months it wants 48 systems at work. Its customers are three of the ten largest U.S. power construction firms — Gritt withholds names for competitive reasons.
Competition in construction robotics
Gritt is not alone. Luminous Robotics, Cosmic and Trinabot are chasing the same market and also automate panel installation. Gritt's bet is generalization: the same AI models should eventually handle other construction tasks. Puneet Puri argues that five years ago "making a system for one solution was still possible", but today the edge lies in AI that handles many tasks.
Plans: from panels to the rest of the build
Ultimately Gritt wants to move beyond solar. The company plans to enter rebar tying?rebar tying: Binding steel reinforcement bars with wire before concrete is poured — a tedious, repetitive manual job. and other construction work. The logic is simple: if the AI models generalize, the same stack can handle many tasks on site.
Why it matters
Gritt's approach marks a shift in construction robotics. Rather than designing an expensive, specialized robot, the company takes cheap rented hardware and builds its value in software. That lowers the barrier to entry and shortens the path to the job site.
The key question is whether the AI models really generalize from panels to other tasks. Solar installation is repetitive and predictable — ideal for automation. Rebar tying and reinforcement work are less structured. If Gritt's stack can handle them, the company stops being a single-service vendor and becomes a construction-robotics platform.
For the solar industry, speed is what counts. 2.8 GW in 18 months is a scale manual crews cannot match without multiplying headcount. With a growing shortage of skilled installers, automation stops being optional and becomes a condition for meeting schedules.
What's next
- Gritt plans to grow its fleet to 48 systems within six months (company statement)
- The company plans to enter rebar tying and other construction work
- The contracted 2.8 GW is due within 18 months — the nearest test of the model's scalability
