Industrial robots have long performed well in structured settings such as factory assembly lines. However, many systems still struggle to operate effectively in environments where layouts change, objects vary, and workflows are less predictable.
Based in Palo Alto, Rhoda AI said its technology is designed to address that gap by enabling robots to adapt more easily to real-world variability.
Earlier this week, the company emerged from stealth after 18 months with the launch of its robotic intelligence system and $450 million in Series A funding to accelerate development and industrial deployment.
The Palo Alto–based company said the capital will support research, engineering, and the expansion of pilot deployments with industrial partners. The funding round will support the continued development of Rhoda’s technology, as well as the expansion of its engineering team and industrial partnerships.
The startup is backed by investors including Khosla Ventures, Temasek, Mayfield, Capricorn Investment Group and Premji Invest, among others. Technology investor John Doerr is also among the backers.
Targeting robotics beyond laboratory environments
Led by Jagdeep Singh, Rhoda is building a new class of robot foundation model designed to bring general intelligence into the physical world.
The company’s system relies on large-scale video pretraining rather than primarily learning from teleoperated robot trajectories.
According to the company, its models are trained on hundreds of millions of internet videos to develop an understanding of motion, physics, and physical interaction, then refined with smaller amounts of robot-specific data.
A new architecture for robotic control
The system developed by Rhoda AI continuously observes its environment, predicts future states using video-based modeling and converts those predictions into actions. The process runs in a closed feedback loop, updating robot behaviour every few hundred milliseconds as conditions change.
The company refers to this architecture as a Direct Video Action (DVA) model, which it says is designed to connect perception and control to enable more responsive operation in dynamic environments.
The company’s platform, called FutureVision, serves as the intelligence layer for the company’s robotics systems and may eventually be licensed to partners building hardware or software for robotics platforms.
Its technology has already been tested in production settings. In one evaluation in a high-volume manufacturing environment, Rhoda reported that its robotic system completed a component-processing workflow in under two minutes per cycle without human intervention.
“In manufacturing, tasks with high variability have historically resisted automation. The real challenge isn’t solving it once, it’s delivering consistent, reliable output under real-world production conditions,” said Jens Wiese, Managing Partner at VC firm Leitmotif and former Volkswagen Group executive. “What impressed us about Rhoda’s approach is its ability to adapt to conditions that typically require human intervention. Technologies like this can dramatically expand the scope of what can be automated, playing a pivotal role in re-industrialising mature economies.”
