Embodex equips experienced electricians, plumbers, HVAC technicians, and other tradespeople with rugged, body‑worn cameras that record unedited 4K first‑person video of their work. The captured footage preserves complete workflows—including bimanual coordination, grasping, cutting, and fastening—and is encrypted for secure storage, creating high‑fidelity training data for next‑generation intelligent systems that need real‑world physical expertise.
Funding
Funding not disclosed
Founders
Product
Problem
Skilled home‑trade knowledge is largely tacit, captured only in the hands of veteran electricians, plumbers, and HVAC technicians. As these workers retire, their practical expertise—how they move, coordinate tools, and make real‑time decisions on messy job sites—is lost, leaving a gap for training both human apprentices and emerging embodied AI systems.
Solution
Embodex equips experienced tradespeople with rugged, 12‑hour battery body‑worn cameras that record unedited first‑person 4K video of their work. The captured footage preserves complete workflows, including bimanual coordination, grasping, cutting, and fastening, under real‑world conditions. Data is encrypted from capture to storage and aggregated to create a large‑scale dataset of physical expertise. This dataset supports the development of world models, embodied learning algorithms, and training resources for both AI systems and human apprentices, ensuring that critical hands‑on knowledge is retained and leveraged for future intelligent technologies.
Target Audience
Primary customers are research labs, technology firms, and academic institutions developing embodied intelligence, as well as safety‑critical organizations seeking authentic skill data from home‑trade professionals.
Features
- 4K ultra‑HD resolution with continuous 12‑hour recording for full‑day job coverage
- IP67‑rated, dust‑proof and water‑resistant design suitable for harsh jobsite environments
- Military‑grade end‑to‑end encryption protecting video data from capture through storage
- First‑person POV capture that records bimanual coordination, tool use, and task sequencing without editing
- Scalable data pipeline delivering high‑fidelity video for training embodied AI and apprenticeship programs