Muscle signals.
Finger-level force.
The Bracelet uses electromyography (EMG) to read muscle activity at the wrist and estimate the force from each finger.
Force-annotated human demonstrations for physical AI. Collect with our Bracelet and API, or use data collected by Hemlock.
The Bracelet uses electromyography (EMG) to read muscle activity at the wrist and estimate the force from each finger.
Use the Bracelet to capture your own tasks, in your own environment. Upload your recordings through our API to get force-annotated data.
Get the Bracelet
Tell us the tasks your robot needs to learn. We collect human demonstrations and deliver force-annotated data for your training pipeline.
Get collected data
From human instinct to robot skill. Force-annotated demonstrations capture the touch behind the task.
Delicate manipulation
Parts pick-and-place

Capture skills such as assembly, packing, parts handling, and tool use.

Explore tasks such as supply handling, tray setup, and equipment preparation.

From opening containers to handling delicate objects, start with the skills that matter to you.
The same movement can use a gentle touch or a firm grip. The Bracelet estimates finger force from muscle signals, adding the information video alone can’t capture.