Short answer
Warehouse layout and facility data describes the building a robot would work in: floor plans, aisle widths, rack heights and bay dimensions, door and dock positions, and floor and lighting conditions. Distributors and 3PLs hold drawings and survey records. This is context data for deployment planning, simulation and evaluation environments. It is not task demonstration data, and security-sensitive details must be removed first.
The buyer's problem
A deployment team cannot judge whether a humanoid fits a site from task footage alone. It also needs to know whether the robot can turn in an aisle, reach the top of a pick face, pass through a door, climb a dock plate or share a path with forklifts. A simulation team needs real geometry to build evaluation environments that resemble the target. A perception team wants maps to ground its localization tests.
These needs are separate from training data. Layout data does not teach a skill. It constrains where skills must work and gives evaluation scenarios their shape. Warehouse design literature treats layout, storage assignment, routing and zoning as the core design variables for order picking (de Koster et al., 2007); a buyer can read the same variables as the parameters that define a test environment.
What facility data includes
| Asset | Typical content | Used for |
|---|---|---|
| CAD floor plans | Walls, columns, aisles, docks, doors, fire exits | Path planning, simulation geometry |
| Rack drawings and specs | Bay widths, beam levels, upright spacing, load ratings | Reach and pick-height analysis |
| Slotting and location maps | Location codes mapped to coordinates | Localization ground truth, label mapping |
| Survey point clouds or scans | Measured geometry, floor flatness | Digital twins, evaluation environments |
| Floor and surface records | Material, markings, ramps, dock plates | Mobility assessment |
| Lighting and environmental logs | Illumination levels, temperature zones | Perception conditions |
| Equipment and traffic maps | Conveyor, lifts, forklift routes, pedestrian zones | Interaction planning |
Pick faces are also context: see pick-face types for how storage forms differ.
What this data can support, and what it cannot
Can support. Building simulation environments with real aisle widths and rack heights; defining evaluation sites and scenario families; checking whether a platform's dimensions and reach suit the building; planning where cameras and charging go; mapping location labels to coordinates for navigation tests; and documenting site differences when reporting results across sites.
Cannot support. Drawings and surveys are static. They do not show how people and equipment move, how items sit in racks on a given day or how a task is performed. They contain no demonstrations, no robot actions and no force data. They also drift from reality as racks are re-slotted and aisles are blocked, so "as-built" and "as-operated" differ. A simulation built from drawings is a model of the building, not evidence that a robot will succeed in it; see the warehouse pick evaluation protocol for how to separate environment context from evaluation claims.
Illustrative facility package
# Illustrative only - synthetic site, no real facility
site_id: site_071
building: single_storey, rectangular, 3 dock doors (stated)
drawings:
floor_plan: dwg/pdf, scale and date stated
rack_layout: aisle_id, bay_id, level_count
dimensions_table:
- {aisle: A03, clear_width_mm: 2800, rack_height_mm: 9000, beam_levels: 5}
survey: optional point cloud, accuracy statement, capture date
environment: {lighting_lux_range, temp_zones, floor_surface}
traffic: {forklift_routes: boolean map, pedestrian_zones: boolean map}
as_built_vs_current: date of last verification
removed: [security_systems, camera_positions, alarm_zones, access_control, high_value_storage_areas]
rights: planning and evaluation use; no redistribution; deletion on request
Security-sensitive details
Facility documents often include things a supplier would not want public: camera and alarm positions, access-control zones, vault or high-value areas, emergency routes and utility details. Ask what the supplier removes before release, and who checks it. A site plan with a customer's name on racks may also reveal the customer's presence. Treat these items as checks:
- Does the drawing set include security, fire-protection or electrical layers, and are they stripped?
- Are named areas or client cages relabeled?
- Is the address or geolocation removed or generalized?
- Is the supplier the building owner, or does a landlord control the drawings?
Supplier types
3PLs and distributors keep drawings in facilities and engineering teams; some hold laser-scan surveys from automation projects, and rack suppliers keep rack designs. Tenants may not own the building drawings, so the landlord or installer may have to agree. Look at 3PL providers as warehouse data suppliers for supplier context, and warehouse onboarding materials for related site documents. If existing documents are stale, a fresh site survey is a new collection that depends on site access; discuss feasibility using the sourcing route guide.
Quality and acceptance pointers
- Currency. Ask for the date of the last verified survey and the change history.
- Scale and units. Drawings must state scale, units and coordinate origin.
- Accuracy. For surveys, ask for the stated accuracy and the method.
- Completeness. Check that dimensions that matter to your platform are present: clear aisle width at the narrowest point, rack-face depth, door clearances, floor transitions.
- Variety. Several sites with different layouts support evaluation across environments; one site supports one environment.
- Held-out sites. If layouts feed an evaluation, keep evaluation sites out of training scenarios.
Limitations and additional preparation
Facility data is context. It needs cleaning of security layers, georeferencing decisions and conversion to formats a simulator can read. It will not replace demonstrations or evaluation protocols. It can also mislead: a drawing that shows a clear aisle says nothing about pallets staged in it. For catalog context on engineering files, see CAD and engineering drawings.
Matching facility data to a purpose
Different uses need different parts of the package. A deployment feasibility study needs clear aisle widths, door sizes, floor transitions and rack heights; precise coordinates matter less. A simulation environment needs a consistent geometry model, so a CAD plan plus a rack layout may be enough. A localization test needs a survey with a stated accuracy, and a mapping from location codes to coordinates. An evaluation that reports results by site needs a short, comparable description of each site, and the same fields for every site.
Write the purpose into the request, and ask the supplier to state which parts they can release for it. Releasing a floor plan with security layers stripped is a smaller commitment than releasing a laser scan that shows what is stored where. A scoped, staged release also lets a supplier's facilities team review what leaves the building.
Questions to ask suppliers
- Which drawings and surveys exist, in what formats, and with what dates?
- How were they produced: design intent, as-built survey or laser scan?
- What is the stated accuracy and coordinate system?
- Which layers are security-sensitive, and who removes them before release?
- Do drawings show client names, cages or high-value zones?
- Does the supplier own the building, or lease it? Who can authorize release?
- Do drawings include rack load ratings, beam heights and aisle widths per bay?
- Is a location-code-to-coordinate mapping available?
- Can the supplier describe how often racking is re-slotted?
Discuss your requirements
If you can specify the building types, dimensions and formats your planning or evaluation work needs, describe the requirement to SourceX. SourceX reviews it, follows up about potential sources among US businesses, and checks the data and the supplier's licensing permissions. A request does not guarantee a match. See the humanoid training data guide and the warehouse and fulfillment hub.
Explore sourcing and licensing options with SourceX
Sources
- Design and control of warehouse order picking: A literature review — layout, storage assignment, routing and zoning as design variables.