The idea
An expert’s spatial hand guides and checkpoints stay attached to a real workpiece as a learner turns it and follows each step.
How it works
- An expert records hand guides around a workpiece, with views that remain attached as it rotates.
- Each checkpoint identifies the correct part and the expected visible state before continuing.
- A learner can recover from a mistake through a reviewed alternate step rather than scrubbing a video.
AI’s role: Search and adapt reviewed lessons to the current step; authoring, prerequisite checks, visible checkpoints, and recovery paths remain explicit and validated.
First demonstration
Help a newcomer assemble a ten-piece mechanism without scrubbing through a video; show the next hand move and verify each visible checkpoint.
What to solve next
How should a lesson explain hidden joins, force, or texture when a phone camera cannot directly observe them?