Lupa Robotics
About
About Lupa Robotics
Lupa captures human manipulation through instrumented gloves that record visual context, tactile contact and hand motion as a person completes everyday physical tasks. Synchronising these signals preserves not only where the fingers moved, but also how contact developed and how the hand adjusted around an object.
The resulting dataset is intended to train robot-hand foundation models rather than one policy for one end effector. By representing dexterity in a form that can transfer between different robotic hands and tasks, Lupa aims to give hardware developers a common manipulation-data layer for grasping, tool use and other contact-rich behaviours.
Timeline
Events
- 2026-01-15 - Award - Lupa Robotics selected for 2026 Healthcare Robotics Catalyst cohort
The five-month programme connects participating companies with corporate mentors, clinicians and the Boston robotics ecosystem. Lupa was selected for work on a transferable human-dexterity dataset combining glove-captured vision, touch and motion. The company is using that dataset to develop a robot-hand foundation model for real-world manipulation.
Source
Context
Comparable profiles
- ForgeAI
stack adjacency - Embodied Foundation Model Developer · Robot Foundation Model Platform
- Alibaba Group
stack adjacency - Embodied Foundation Model Developer · Robot Foundation Model Platform
- RLWRLD
stack adjacency - Embodied Foundation Model Developer
- Sereact
stack adjacency - Robot Foundation Model Platform
- Physical Intelligence
stack adjacency - Robot Foundation Model Platform
- XYZ Embodied AI
stack adjacency - Embodied Foundation Model Developer
- ACE Robotics (大晓机器人)
stack adjacency - Robot Foundation Model Platform
- Delta Intelligence
stack adjacency - Embodied Foundation Model Developer
- NC AI
stack adjacency - Embodied Foundation Model Developer
- Tommoro Robotics
stack adjacency - Robot Foundation Model Platform
- Field AI
stack adjacency - Robot Foundation Model Platform
- Intelligence Factory
stack adjacency - Physical AI Training Data
Context tags
- Embodied foundation model developer - Stack Layer - Reviewed July 15, 2026 against the company overview and structured profile evidence: embodied foundation model developer.
- Robot foundation model platform - Stack Layer - Reviewed July 15, 2026 against the company overview and structured profile evidence: robot foundation model platform.
- Physical AI training data - Stack Layer - Reviewed July 15, 2026 against the company overview and structured profile evidence: physical ai training data.