Questions people ask

About the drawing

Why does everything ask for millimetres? One unit everywhere removes a whole class of mistakes. A 120 m building is 120000. Displays round to metres where that reads better.

Where is the Save button? There isn't one. The scenario saves as you go. Ctrl+S does nothing on purpose.

My zone was refused. Why? Zones must sit inside the building and must not overlap. The message says by how much, for example "move it 1 900 mm inside". Fix that distance and try again.

I can't find a Bulk or Dangerous goods button. They are on the Zones tab but marked "coming soon". Ask the copilot to draw one, or import it from a DXF. See Zones.

Why are some ribbon buttons greyed out? They are on the roadmap. Hover one and the tooltip says "coming soon".

Why did my racking disappear when I moved the zone? It did not disappear; it regenerated to the new shape. If the new shape fits no rows at all, the platform keeps the old racking and refuses the change instead.

About capacity

Why are some positions marked dead? No aisle reaches their face. They exist in steel but not in practice, so the platform shows them and does not count them as usable.

Why don't my staging floor positions count as capacity? Goods on a staging floor are passing through, not stored. Floor positions are reported on their own, as throughput space.

Why is there no overall score? A single score needs a weighting of capacity against cost against travel. That is a business decision, not an engineering fact, so the platform shows every dimension and leaves the weighting to you.

About simulation

Is the comparison fair? When both runs use the same seed, horizon and replications, replication i of each sees the same trucks and orders until your change makes them differ. The difference is then paired, replication by replication, and the band is a paired interval. The platform measures that the demand matched; it never assumes it.

Why three replications? Why not one? One replication cannot tell you how much of the result was luck. Three gives a first band. Ten narrows it. A demo run takes about fifteen seconds, so more replications are cheap.

What is not modelled? Truck appointments and dock schedules, one-way aisles, breakdowns, individual products in specific slots, and rolling a missed order onto the next truck. Each run's assumptions name what applied to it.

My run is marked stale. Is it wrong? No. It is still right for the inputs it read, which its Inputs panel shows. It is stale because the scenario has changed since. Run again to see the current design.

Why is the spend called "indicative"? It is rate × paid hours: wages only. It is not a cost-to-serve figure, and it says so.

About the copilot

Can the copilot break my design? It uses the same validated tools you do, so it cannot place anything the rules refuse. It can make a poor design, so check its work. For big experiments, clone first.

Can I undo what the copilot did? Its changes are not on the editor's undo list. Ask it to remove what it built, delete it yourself, or work in a clone.

Why does it ask me questions instead of building? A brief that leaves a real decision open, such as which wall the docks go on, gets a short question before any building starts.

Where does it run? On the Warehouse Studio server, with the AI provider your administrator configured. A turn keeps running if you reload the page; the panel reattaches.