Ropedia's HOMIE: The Wearable Tech Changing How Robots Learn
Ropedia raises $22 million to scale its head-mounted device, HOMIE, which captures how humans move so robots can learn from it.
Key points
- Ropedia raised $22 million in pre-Series A funding in October 2023.
- HOMIE is a head-mounted device with four cameras that captures human movement to train robots.
- Data-collection costs run up to 50 times lower than traditional methods.
- Ropedia collects around 1,000 hours of data weekly, with 100,000 hours gathered so far.
- The company plans to expand collection across North America and Southeast Asia.
Ropedia has raised $22 million in pre-Series A funding to scale HOMIE, its head-mounted data-collection device. Add the earlier seed round and the company's total sits at $30 million. Founded in 2025, with offices in Singapore and Mountain View, California, it's building what CEO Zhaoxi Chen calls data infrastructure for physical AI: the pipeline that turns human behaviour into robot training material.
What is HOMIE and how does it work?
HOMIE fits on your head and packs in four cameras for a 360-degree view, an audio sensor that reconstructs where sounds are coming from, and a motion unit to track how you move. The battery module is swappable. Ropedia builds most of the hardware itself, which Chen says produces cleaner data than off-the-shelf components would.
That matters because the whole point is egocentric data: footage and sensor readings captured from a human's own perspective, the same way people actually learn physical skills. Ropedia claims the approach cuts data-collection costs by up to 50 times compared with conventional methods. Our earlier look at physical AI investment on 22 July 2026 shows just how competitive that cost argument is becoming.
Why is this funding important for Ropedia?
The money goes toward three things: scaling HOMIE production, hiring across hardware, software and model development, and building out the data platform with annotation tooling and compliance infrastructure.
Right now the company collects around 1,000 hours of footage a week, 100,000 hours in total, focusing on home environments. Housekeeping, cleaning and interior design make up the bulk of the datasets. There's also a pickleball club whose members wear HOMIE during practice. Clients can either license existing datasets or commission new data captures, and it's the second option where Ropedia expects to grow.
What are Ropedia's next steps?
Expansion targets North America first, since most existing clients are based in the United States. Chen told The Robot Report the company also wants to grow its U.S. Team and increase manufacturing output.
Longer term, Ropedia plans a hand device that tracks finger movements and tactile contact, synchronised with HOMIE so the two streams align. Shoes that sense how feet interact with the ground are also in the works, aimed specifically at humanoid robot training. Chen described the vision as turning a human into a "socket" with devices on multiple body parts at once. That's either visionary or unsettling depending on your comfort with being a data-collection node, but the robotics industry's appetite for this kind of training data isn't slowing down.
Common questions
How does Ropedia ensure privacy with HOMIE?
Ropedia says it's building compliance infrastructure into the data platform to handle large-scale collection safely, though it hasn't published detailed privacy policies publicly.
Is HOMIE available for personal use?
No. HOMIE is a research and industry tool. There's no indication of a consumer release.
