Rhoda AI exits stealth with $450M Series A and FutureVision video-predictive robotics foundation model

Rhoda AI announced its public launch after 18 months in stealth. The company unveiled FutureVision, a foundation model for robotic intelligence based on video-predictive control designed to function in real-world environments. It also raised $450 million in Series A funding to accelerate development and industrial deployment. The system pre-trains on hundreds of millions of internet videos to build priors on motion and physics. It post-trains on smaller robot datasets to learn embodiment-specific behaviors. The Direct Video Action model operates in a closed loop by observing the environment, predicting future video states, converting predictions to actions, and re-observing the world every few hundred milliseconds. Rhoda reported autonomous operation in production environments with shifting materials and workflows. In a recent manufacturing evaluation, the technology completed a component-processing workflow in under two minutes per cycle without human intervention. The company said this performance exceeded customer key performance indicators. Chief Executive Officer Jagdeep Singh said the next era of robotics requires models that understand how the world moves. The Series A proceeds will support research, engineering, customer pilots, and team growth in generative artificial intelligence, computer vision, and robotics.