Origin raised $20 million from Accel and Zacua to scale construction robotics in the US and Europe. The startup plans to deploy 150 autonomous robots by the end of next year. "Replacing them with a robot from us saves them 20% of the cost from day one," said Yogesh Ghaturle.

Summary

The deeptech startup plans to deploy 150 autonomous robots by next year on a subscription model, offering Western contractors up to 20% savings amid severe labour shortages.

Origin, a deeptech startup building autonomous robots for interior construction, has raised $20 million in a round led by Accel and Zacua Ventures as it emerges from stealth, with plans to build 150 robots by the end of next year. Unilazer Ventures, Schema Ventures, Z21, ISAI and Imad Ventures also invested, as did some of Origin's customers. Accel's India fund is the only Indian entity on the startup's cap table.

The raise is an early bet on physical artificial intelligence (AI) in one of the least automated industries. However, venture funds are increasingly parking money in the sector, with robotics in construction attracting $1.3 billion in funding until Q3 2025, compared to $612 million in all of 2024, according to Pitchbook data.

Origin's pitch to its customers -- large housing contractors in the US and Europe such as D.R. Horton and Lennar Corp. -- centres on margin expansion amid high labour costs and worker shortages. These pressures have been compounded by tighter immigration curbs on white-collar and blue-collar workers in both regions.

"The cost of hiring skilled contractors in the US is very high. For a US contractor with a crew of five people, they pay $50,000 a month. Replacing them with a robot from us saves them 20% of the cost from day one," Yogesh Ghaturle, co-founder and chief executive of Origin, said in an interview with Mint.

Origin plans to have autonomous robots that can work on the entire interior construction of a home including things such as installation of heating, ventilation and air conditioning systems, plumbing, dry wall painting. Currently however, the startup's flagship robot, the OG-1, is capable of completing wall-finishing. This includes sanding, applying primer paint, final painting.

Founded in January 2025 by Ghaturle, Ashish Daga and Tushar Makkar, Origin rolled out its initial pilots and deployments in October 2025 and currently has six robots operational across the US and European markets.

The startup is currently focused on three priorities: scaling its painting operations with existing clients while finishing contracted projects and starting new pilots; stepping up manufacturing capacity; and building deployment teams across both markets. "We're planning to have built 150 robots by the end of next year," Ghaturle said.

According to a Zacua Ventures report on construction robotics published in March, "Investors are paying up for autonomy in proven layout, piling and scanning workflows, while treating interior, MEP and façade robots as earlier-stage, higher-risk bets."

Robots as a service

Origin's clients do not buy its robots outright; the startup operates on a subscription-based robots-as-a-service model instead, which Ghaturle described as "outcome as a service." Its customers pay between $25,000 and $40,000 a month for access to the robots.

While Ghaturle declined to disclose the number of contracts Origin has signed, he said current deals average $2-2.5 million. "Our average contract value will be more like $5-$10 million as we sign more contracts," he added.

A launch for Indian construction companies remains a few years away.

Building IP

Ghaturle describes Origin as a full-stack physical AI company, with intellectual property in hardware, software and AI. The robot is a modular, three-wheeled platform that is just 600-700mm wide and can work at heights up to 14 feet. It runs off-the-shelf power tools. "If you want to unlock a new tool on the platform, all you have to do is change the tool system and the end effector and a new use case will get unlocked on the platform," he said.

On the software side, Origin is building what it calls construction action models -- a class of robotics models for the industry. These help robots identify objects, navigate cluttered sites, avoid obstacles and work on ceilings.

Origin draws on three data sources: real-world data its deployed robots collect on job sites; photorealistic simulation and synthetic data, where it can randomize environments; and data from a lab where the team builds, breaks and rebuilds scenarios. Synthetic and lab data came first, but site data is now scaling the fastest.