REVEL builds technology that captures expert human skill and translates it into robot intelligence, enabling robots to perform complex tasks traditionally done by craftsmen, surgeons, and builders. By recording and modeling the motions and decisions of skilled workers, their platform creates reusable, scalable skill libraries that can be deployed across industries to address labor shortages and preserve knowledge. This approach lets robots handle hazardous or scarce‑skill work while humans focus on uniquely human activities.
Funding
Funding not disclosed
Founders
Product
Problem
Skilled manual work such as surgery, electrical installation, and fine craftsmanship is increasingly scarce due to aging workforces, declining apprenticeship pipelines, and the physical risks associated with many tasks. When experts retire or leave, their tacit knowledge often disappears, limiting the ability to maintain critical infrastructure and services.
Solution
Revel provides a platform that records the motions, decisions, and sensory feedback of expert practitioners and translates that data into machine‑readable models. Using advanced motion capture, force sensing, and computer‑vision systems, the platform creates detailed task representations that can be used to train industrial robots. The trained robots can then perform the same operations in hazardous or labor‑intensive environments, extending the reach of human expertise while preserving it indefinitely. By decoupling skill from the individual, Revel enables organizations to redeploy human workers to higher‑level activities and maintain service continuity even as the skilled labor pool shrinks.
Target Audience
Primary customers are manufacturers, healthcare providers, and service companies that rely on highly skilled manual labor and seek to automate hazardous or hard‑to‑staff tasks.
Features
- Multi‑modal data acquisition suite combining high‑precision motion capture, force/torque sensors, and visual recording to capture complete task context
- AI‑driven skill extraction pipelines that convert raw sensor streams into reproducible robot motion programs
- Compatibility with a range of collaborative and industrial robot hardware through standardized robot‑agnostic APIs
- Continuous learning loop that allows robots to refine performance based on real‑time feedback and operator validation
- Secure knowledge repository that stores captured skill models for long‑term reuse and version control