Capture for this environment has not started, so there are no frames to show yet. The rule set and record schema are the same as the programmes already running.
What this environment asks for
Warehouse work alternates between long traverses with nothing to annotate and short dense bursts of manipulation at the pick face. The segmentation problem is therefore as much about excluding travel as about cutting actions.
This environment is in build-out. The rule set and review model are already defined, so the work ahead is capture partnerships and a bin and tote vocabulary rather than new methodology.
How the work runs
- Applying the published rule set unchanged so records stay interoperable.
- Excluding travel segments, which contain no annotatable hand action.
- Segmenting pick, scan and place as three distinct actions rather than one.
- Describing bin and tote position so a pick target is resolvable.
- Recording large camera motion across the traverse into each pick.
- Applying the same four-pass review before delivery.
What makes it hard
- Long travel segments contain no hand action and must be excluded cleanly.
- Scanning is a very short action that sits at the half-second floor.
- Bin faces hold many similar items, so position carries the object reference.
- Carried totes occlude the hands for much of a traverse.
- Warehouse lighting is uneven between aisle and pick face.
- High pick rates make reviewer consistency the limiting factor.
Result
Programme defined against the existing rule set, pending capture partnerships and a client bin and tote vocabulary.
What comes out
Same record structure as the live programmes, with travel excluded and pick targets located by bin position.
