Warehouse layouts
Which layout works best in the same building? Several warehouse layouts with the same robots, the very same orders, deliveries and incidents, simulated and compared: throughput, order and putaway times, metres driven, waiting.
How it works
Compares warehouse layouts in the same building.
Each line of Layouts is one layout in the hall of the Warehouse fleet block (its width from Rack pairs, its depth from Rack blocks × Rack length): blocks=9 cuts the racks into more, shorter blocks with more cross-aisles (shorter trips, less storage), blocks=3 into fewer, longer ones; stations=4 or 12 changes the pack-out stations; stations_at=both splits them between the south and north walls; inbound and inbound_at do the same for the receiving stations (by default 4 on the north wall), levels sets the shelf levels; slotting=random puts popular products anywhere instead of near the stations.
Every layout gets the same robots, the very same orders and deliveries (arrival times, lines, classes) from the same starting stock and the same incidents (when, what kind and how long; where depends on the layout), for Simulated hours after a warm-up, once per seed, with routes planned on the worker's accelerator.
Results per layout: orders and lines per hour, packages stored per hour, reaction and putaway times, deliveries waiting, share of the demand served, average and 95th-percentile order time, metres driven per item, robots busy, waiting, charging, time stopped by faults, incidents, orders left waiting, pick faces and metres of racks (storage). layouts.csv for a Table, layouts_chart.csv for a Chart (bars by layout), a plan of each layout coloured blue to red where robots waited (for an Images view), and layouts_summary.json.
Near the robots' limit (Orders per hour) the layouts differ most; well below it they all keep up and differ in order time and driving. To watch one of them, put its line in the Layout setting of Warehouse fleet.
Outputs
| Output | Gives | What it is |
|---|---|---|
Resultsresults | Table csv | layouts.csv: one row per layout (throughput, order time, metres per item, waiting, storage…) |
For a chartchart | Table csv | layouts_chart.csv: orders per hour, order time, metres per item and waiting by layout |
Plansplans | Files png, json | A plan of each layout from above, coloured where robots waited (PNG), and its layout (JSON) |
Summarysummary | Metrics json | The best layout for throughput, order time and driving, and each layout against the first |
Live telemetrytelemetry | Live telemetry | What the block sends while it runs, for views such as Chart or 3D view. |
Settings
| Setting | Default | What it does |
|---|---|---|
Layoutslayouts |
Today: blocks=5 stations=8 inbound=4 · 3 long blocks: blocks=3 · 9 short blocks: blocks=9… | One per line, Name: key=value … with blocks (rack blocks from front to back: more blocks, more cross-aisles), stations (pack-out stations), stations_at (south, north or both), inbound (receiving stations), inbound_at (north, south or both), levels (shelf levels), slotting (abc: popular packages near the pack-out stations, or random). The first is the reference |
Robotsrobots |
50 | From 1 to 60 |
Orders per hourorder_rate |
350 | Near what the robots can do, the layouts differ most From 1 to 6000 |
Deliveries per hourinbound_rate |
0 | 0: as many as the orders take out From 0 to 20000 |
Seeds per layoutseeds |
3 | Each seed is another day of orders and incidents, the same for every layout From 1 to 20 |
Simulated hourshours |
2 | From 0.1 to 24 |
Warm-up (simulated minutes, not counted)warmup_min |
10 | From 0 to 120 |
Acceleratoraccelerator |
auto | auto: Tenstorrent cards, else an NVIDIA GPU, else the CPU of the worker that runs it One of: auto, CPU, NVIDIA GPU, Tenstorrent |
Rack pairs (sets the hall's width)rack_pairs |
20 | From 4 to 30 |
Rack blocks of the reference hallrack_blocks |
5 | From 1 to 8 |
Rack length of the reference hall (m)rack_len |
8 | With Rack blocks, sets the hall's depth; every layout fits in that same hall From 4 to 16 |
Pack-out stations (unless a layout says)stations |
8 | From 1 to 20 |
Receiving stationsinbound |
4 | Where deliveries come in on a conveyor and robots load them (north wall unless the Layout says inbound_at=south) From 0 to 20 |
Shelf levelslevels |
3 | Levels of each rack; each level holds one package per metre on each side From 1 to 4 |
Shelves full at the start (%)fill |
70 | The run starts from this stock, A packages near the pack-out stations, so it begins in a steady state From 0 to 98 |
Seconds to pick an itempick_s |
10 | From 1 to 120 |
Seconds to unload at the stationstation_s |
15 | From 1 to 120 |
Seconds to load a deliveryload_s |
6 | At the receiving station, per package; storing one on a shelf takes as long as a pick From 1 to 120 |
Robot speed (m/s)robot_speed |
1.2 | From 0.3 to 2 |
Hours on a chargebattery_h |
8 | From 0.5 to 24 |
Minutes to charge from emptycharge_min |
60 | From 5 to 600 |
Go to charge below (%)charge_below |
25 | From 5 to 80 |
Incidents per hourincidents_h |
2 | From 0 to 60 |
Robot faults (% of incidents)fault_share |
40 | From 0 to 100 |
Minutes until someone comesresponse_min |
3 | From 0 to 120 |
Minutes to clean up a spillclear_min |
4 | From 0.1 to 240 |
Minutes to repair a robotrepair_min |
8 | From 0.1 to 480 |
Random seedseed |
1 |
Where it runs
On one of your computers with the EsaSim worker: Any worker, NVIDIA GPU, Tenstorrent.
The worker installs the Python packages it needs when they are missing: torch[cuda]</code>, <code>scipy</code>, <code>pillow.
In its code, settings are ctx.params["key"], inputs ctx.inputs["id"] and outputs ctx.output("id", "name.ext"); see Custom blocks.