Skill Capture Glove

Sunday Robotics's glove that lets a person do housework wearing it and directly produce robot training data.

The Skill Capture Glove is a data-collection device from the U.S. home-robotics company Sunday Robotics, made public in November 2025 alongside its household robot, Memo; the company was founded by Tony Zhao (an author of ACT/ALOHA) and Cheng Chi (an author of Diffusion Policy/UMI). The glove shares the same geometry and sensor layout as Memo's robotic hand, so when a person wears it to do housework, the recorded motion and force data can be used almost directly as robot data; the remaining embodiment gap is handled by a process called Skill Transform, which the company says succeeds about 90% of the time. Each glove reportedly costs about $200, with a full teleoperation setup costing around $20,000. It's a representative example of embodiment-free, capture-execution-isomorphic data collection.

ExampleSunday says it has sent thousands of gloves to “Memory Developers” — people who wear them at home to record household chores — with data reportedly coming from about 500 households, used to train its foundation model, ACT-1, so that Memo learns chores like washing dishes, making coffee, and folding laundry.

Also called

Sunday Robotics Capture Glove

Related

Sunday Robotics · Sunday Robotics Memo · Sunday Robotics ACT-1 · Robot-free (Embodiment-free) Data Collection · Capture-Execution Isomorphism · Universal Manipulation Interface

Sources

Sunday Robotics 官网 (Chinese)
Sunday: ACT-1, a robot foundation model trained on zero robot data
Humanoids Daily: Sunday Unveils Memo, a Wheeled Domestic Robot That Learns From $200 Gloves

As of

2026-09