Data Collection

The London-based startup wants to be the infrastructure layer for robot learning, offering free cloud-native data tools to researchers to solve the "plumbing" problem.

Startup Generalist AI has unveiled GEN-0, an embodied foundation model it claims is trained on an unprecedented 270,000 hours of real-world manipulation data. The company reports its new architecture, "Harmonic Reasoning," and massive dataset have unlocked predictable scaling laws for robot intelligence.

A recent LA Times report highlights the massive, low-tech effort to gather real-world data for robots. This data bottleneck is the central challenge in robotics, and top companies are making wildly different bets on how to solve it—from human-video capture and teleoperation to massive simulation.

Neura Robotics has announced the NEURA Gym, a large-scale physical training facility designed to generate real-world interaction data from hundreds of robots. The initiative aims to solve one of the biggest challenges in robotics: creating AI models that can reliably transfer skills from simulation to the complexities of the physical world.