AI Hand Engine
Palmplug’s AI Hand Engine captures, learns, and transfers dexterous skills to humans and robots.
Record
Connect Palmplug to the AI Hand Engine and capture every micro movement in the hand and fingers with precision.
Analyze
Turn raw motion data into clear, interactive charts perfect for analysis. Researchers can zoom, segment, and annotate sessions, then export datasets and figures. It synchronizes multi-sensor streams (finger flexion, abduction, adduction, acceleration, hand pose) at 120 Hz, supports filtering and event tagging, and computes summary metrics automatically.
Export
The AI Hand Engine exports CSV data at 120 Hz (120 samples/second), including:
Joint rotations (DIP, PIP, MCP)
Finger flexion
Finger abduction/adduction
Finger acceleration
Hand position
Train
Train your hand model to perform a wide range of skills from simple gestures like sign language or playing a song on the piano to complex, dexterous tasks such as opening a bottle or assembling a product. Export dexterous skills from the hand engine to NVIDIA Isac Lab.
Link
Connect your trained hand model to Robots, Apps, Devices, and other AI training models. Unlock a wide range of skills. Using the Palmplug Hand Engine API.
Includes:
Palmplug SDK
Skill Capture
AI Analytics & Research Tools
Interactive Charts & Graphs
CSV (and JSON) Data Export
Cloud Storage & Backups
Skill API (API access to skills)