Insider Brief
- Ropedia launched HOMIE Gen2, a head-mounted data-capture system designed to record real-world human activity and convert it into training data for robots and physical AI models.
- The 380-gram device provides 360-degree visual coverage, records 1080p video at 30 frames per second with a global shutter and includes four-channel spatial audio, while supporting synchronized capture across multiple units.
- Recordings are processed through Ropedia’s Human Experience Engine, which structures more than 10 data types including video, audio, depth, movement, body pose and spatial signals, with the company reporting up to 96% accuracy in capturing and interpreting human activity.
Ropedia has launched HOMIE Gen2, a wearable data-capture system designed to record human activity in real-world environments and turn it into training data for robots and physical AI models.
The Singapore-based company said HOMIE Gen2 is the fourth generation of its capture hardware developed over the past 12 months. The launch follows a recent pre-Series A funding round that brought Ropedia’s total funding to $30 million.
“LLMs have shown that scaling training data can unlock new levels of intelligence. Physical AI follows a similar scaling principle, but its training data must capture the complexity of the real world, co-founder and CEO Zhaoxi Chen said in the announcement. “Traditional fixed-camera setups and controlled demonstrations can capture individual actions, but they struggle to provide the diversity and in-the-wild data needed to train systems for the physical world. As human experience scales in volume, diversity and quality, it can provide a broader foundation for physical AI to learn how people perceive and act in real environments.”
HOMIE Gen2 is worn on the head and records a person’s surroundings and actions while they move through environments such as homes, factories and retail locations. Ropedia is targeting robotics developers that need data showing how people perceive, move and interact with the physical world.
According to the company, the device provides 360-degree visual coverage and records 1080p video at 30 frames per second using a global shutter, which captures an entire image at once rather than scanning it line by line. The 380-gram system also includes four-channel spatial audio.
A multifunction port allows multiple devices to be synchronized and is designed to support additional sensors. Ropedia said several HOMIE Gen2 units can be used at the same time to record an environment from different people and viewpoints.
The system is self-contained and designed for hands-free use, allowing activities to be captured where they normally occur rather than recreated in dedicated data-collection facilities.
Recordings are processed through Ropedia’s Human Experience Engine, which synchronizes and structures more than 10 types of information, including video, audio, depth, movement, body pose and spatial data. Ropedia said testing of the system has shown accuracy of as much as 96% in capturing and interpreting real-world human activity.
“We are at a stage for Physical AI similar to where autonomous driving was before widespread deployment,” added Chen. “Autonomous driving has millions of vehicles on the road generating real-world data. Physical AI doesn’t have that scale of robot deployment yet. Human data helps bridge that gap, giving us the real-world experience needed to move the industry forward.”
HOMIE Gen2 builds on Ropedia’s Xperience-10M dataset, released in March. The company said the dataset contains more than 10 million interaction episodes drawn from 10,000 hours of first-person recordings.
