Pick by vision for a warehouse

The next pick stays in view on a strip that follows your head, the right bin lights up, and a pinch confirms it.

A capture of the project in play mode in the Unity Editor, taken by the same check the agent uses; on a headset the real room shows where the sky is here. Play mode in the Unity Editor: a warehouse rack of twelve bins labelled A1 to C4, bin C2 outlined in green, a dark strip in the middle of the view reading C2, Spark plugs x 4, pick 3 of 6, and a cart with a blue tote holding the first picked boxes.

See its pageA native app for the headset. Continuing from it needs Unity 6 on your computer.

The request, word for word

This is the text that was sent to the agent to start the project. Nothing was edited for this page.

Turn this into pick-by-vision for a warehouse on smart glasses, the kind of app a logistics company would commission: the next pick is always in view, and the right bin lights up. Glasses have a narrow view, so keep everything that matters inside the middle of the view and use few, large words. In the world, 2 m in front of the start position: a shelf rack built from simple shapes, 4 columns by 3 rows of open bins (grey frame, cardboard-brown boxes inside each bin), with a large label on each bin: A1 to A4 on the top row, B1 to B4, C1 to C4. A small cart with a tote stands to the right of the rack. Fixed to the head (it follows where I look, about 1.2 m away, a little below centre, never more than 30 degrees wide): a strip with three things, big and high contrast: the bin ("B3"), the item and quantity ("Brake pads x 2"), and the progress ("Pick 2 of 6"). A thin arrow at the edge of the strip points toward the bin when it is not in view. The bin to pick from glows green and pulses. Confirming a pick (pinching while pointing at the bin, pressing it, or pressing the Space key in the Editor) plays a short tick, makes a box fly from the bin into the tote on the cart, and moves on to the next pick. Pointing at and confirming the wrong bin flashes it red, shows "Wrong bin" on the strip for a second and counts an error. Make up an order of six picks across different bins, with plausible car-part names and quantities. After the last pick the strip shows "Order complete", the time, picks per minute and the number of errors, and a "Next order" button resets. Remove the table, ball and button from the template scene; keep the rig and the flat-screen fallback. So that it can be checked without a headset, add a small demo mode: when the scene starts in the Editor it confirms the first three picks by itself, one every 1.2 seconds, and then waits on the fourth. On a device this is off. Title label above the rack: "Pick by vision".

How it went

Runs
One run: the request above, and it was done. A run ends only when the project builds and loads in a real browser without errors.
Started from
The Unity for XREAL template: A native Unity project for XREAL glasses: Project Aura through Android XR out of the box, XREAL One and Air through the XREAL SDK you add; hands, gaze and the real room, built as an Android app.
Notes and credits
The order and the part names are made up. The loop is six moments of the project's own demo mode in play mode. AI: Claude Code.

Build your own

Graspable is a desktop app. Describe the scene you want, and the agent builds it as a real project on your computer and checks that it runs. It is free with your own AI account.

Download GraspableYour first project

More examples

See all examples