Tactile Sensing Glove Documentation
Operator-side tactile capture workflows for demonstrations, dexterous manipulation studies, and contact-aware data collection.
Purpose
A tactile sensing glove is most useful when teams want to capture how contact evolves during human demonstrations rather than only infer intent from vision and motion. It can help preserve pressure patterns, timing, and contact transitions that are difficult to reconstruct after the fact.
Typical Workflow
A common setup uses the glove during teleoperation, object handling studies, or dexterous demonstration capture. The glove stream should be synchronized with wrist pose, finger motion, camera feeds, and task labels so the resulting data can be reused for training, replay, and evaluation rather than only for one-off inspection.
Calibration Considerations
Before task capture, operators should establish a clean baseline, verify repeatability across fingers or sensing regions, and validate how the glove behaves during grip transitions and partial contacts. Inconsistent calibration quickly reduces the value of the dataset, even if the raw tactile signal looks rich.
Operator Behavior Matters
With glove-based tactile collection, protocol matters almost as much as hardware. Teams should define consistent grasp styles, object presentation, reset conditions, and failure logging rules. That makes the data more comparable across sessions and prevents the tactile stream from becoming too idiosyncratic to learn from.
Best use case — Glove-based tactile capture is especially valuable when teams want richer human demonstration data for dexterous or contact-sensitive tasks.