Indoor mapping is the process of converting a physical space into a structured digital model that a navigation engine, mobile app or kiosk can use to plot routes, position a user and trigger location-based content. Unlike outdoor cartography, which relies on established geospatial datasets, indoor mapping almost always starts from scratch with each building. The work is repetitive, detail-heavy and easy to underestimate. This guide covers the practical workflow, data formats, accuracy validation and the ongoing maintenance that keeps a map useful after launch.
Indoor Mapping Workflow
A reliable indoor mapping workflow moves through distinct phases rather than trying to produce a finished file in one pass. Skipping any of these stages tends to create problems that only surface during testing, when correction is most expensive.
Scope and Requirements
Before touching any software, define what the map needs to support. A retail unit needing zone-based notifications has different requirements from a hospital requiring turn-by-turn wayfinding to specific treatment rooms. Key questions at this stage include which floors are included, whether vertical circulation (lifts, stairs, escalators) must be modelled, and whether the map needs to support routing or simply display position. Documenting these decisions early prevents scope creep and avoids over-engineering a map for needs that never materialise.
Survey and Data Capture
Physical survey comes next. Existing architectural floor plans are a starting point, but they rarely reflect the current state of a venue. Furniture, temporary partitions, moved fixtures and post-build alterations mean the survey must verify what is actually on the ground. Capture methods range from tape measurements and laser distance metres for small spaces to LiDAR scanners and photogrammetry for larger venues. The choice depends on the required precision, the budget and the complexity of the space.
Digital Drafting
The surveyed measurements are translated into a digital model. This is where the chosen data format dictates the toolchain. The draft goes through internal review, checking that corridors have realistic widths, rooms are correctly labelled and routing paths connect logically between floors.
Integration and Testing
Once the digital map passes internal checks, it is imported into the target platform — the indoor navigation engine, the mobile app or the content management system. Testing at this stage means walking the space with a device and comparing the rendered position against physical landmarks. Discrepancies between the map and reality become obvious here, and the cycle of adjustment and re-testing continues until the map performs acceptably within the measured environment.
Floor Plan Data Sources and Formats
The format you choose constrains what platforms can consume the map, how much detail you can encode and how painful future updates will be. There is no single universal standard, so the decision is driven by what your navigation engine accepts and what your integrator can work with.
Common Formats
- GeoJSON: Widely supported in web-based platforms. Stores coordinates and properties as JSON, making it relatively easy to inspect and edit manually. Suitable for polygon-based room outlines and point-of-interest placement, but does not natively encode multi-floor relationships or complex routing graphs.
- IndoorGML: An OGC standard designed specifically for indoor spatial information. It models spaces, connectivity and navigation semantics. Powerful in theory, but tooling support remains uneven and many commercial platforms do not import it directly.
- IMDF (Indoor Mapping Data Format): Apple's format for indoor maps. Structured around GeoJSON with a defined schema for venues, floors, rooms and features. If Apple Maps integration matters, IMDF is effectively required. Other platforms may or may not support it.
- SVG and PNG: Raster or vector images used as background layers with coordinates mapped manually. Simple to produce but fragile — any change to the floor plan means re-registering the image against the coordinate system. Common in early-stage pilots but problematic at scale.
- Proprietary formats: Many indoor navigation vendors use their own schema. These can be efficient within that ecosystem but create lock-in and make migration difficult.
Source Material
Architectural CAD files (DWG or DXF) are the most common starting point. They contain accurate geometry but include layers of information irrelevant to navigation, such as electrical schematics and structural annotations. Extracting the usable geometry — walls, doors, open areas — is a manual cleanup task. If CAD files are unavailable or outdated, scanned paper plans or even hand-drawn sketches can work for low-accuracy applications, provided the limitations are accepted upfront.
Whatever the source, verify the scale. A CAD file drawn in millimetres imported into a system expecting metres will produce a map that is technically correct in shape but three orders of magnitude wrong in size, breaking all distance calculations silently.
Creating and Updating Indoor Maps
Creating the first version of an indoor map is labour-intensive. Updating it is where most projects either succeed or slowly fail, because venues change far more often than operators expect.
Initial Creation
Start by stripping the source plan down to navigable geometry: walls, doorways, open corridors and defined obstacles. Assign each enclosed space a category — room, corridor, stairwell, lift, open area — because the navigation engine uses these categories to determine where routing is permitted. A storage cupboard, for example, should not appear as a valid destination or transit space unless there is a specific operational reason.
Add points of interest separately from the structural geometry. A till point, an exhibit, a meeting room label — these are attributes placed within the spatial model, not part of the walls-and-doors layer. Keeping them separate means the structure can remain stable while points of interest change frequently.
Handling Updates
Shops refit, exhibitions rotate, event layouts reconfigure and building works close corridors temporarily. Each of these changes demands a map update, and the process for making that update needs to be as simple as the initial creation allows.
Practical approaches include maintaining a master file in an editable format (such as GeoJSON or a vendor's native format) with clear layer separation, so that a change to a single room does not require redrawing the entire floor. For temporary changes — a conference layout that lasts three days — some platforms support overlay layers that can be activated and deactivated without modifying the base map.
A common mistake is treating the map as a one-time deliverable. If there is no defined process for who reports a physical change, who updates the map and who approves the update, the map will drift from reality within weeks.
Map Accuracy and Validation
Accuracy in indoor mapping is not a single number. It depends on what is being measured, the method used and the environment itself. Any claim of precise accuracy should be treated with scepticism unless it is backed by a measured survey of the specific venue.
Types of Accuracy
- Geometric accuracy: How closely the digital walls and features match their real-world positions. This is determined by the survey method and the care taken during digitisation.
- Positioning accuracy: How closely the navigation system places a user's dot on the map. This depends on the positioning technology (BLE beacons, Wi-Fi, UWB), the calibration of that system and the environment's radio characteristics. The map itself can be geometrically perfect yet the user dot still appears offset.
- Routing accuracy: Whether the calculated path is physically walkable. A route that sends a user through a locked door or across a reception desk fails regardless of how precisely the walls are drawn.
Validation Methods
Walk the space with the running system and a printed floor plan. At defined checkpoints — doorways, corridor junctions, specific rooms — compare the displayed position against where you are actually standing. Record the offset at each point. This produces a set of empirical measurements tied to that specific environment, which is far more useful than a vendor's generalised specification.
Check routing by requesting paths between a representative set of origin-destination pairs, including edge cases such as routes that cross floors, routes that must avoid certain areas and routes between adjacent spaces that should be trivially short. If the system suggests a convoluted path between two neighbouring rooms, the connectivity graph likely has a missing link.
Document the validation results. If the measured positioning accuracy in a particular zone is consistently outside acceptable bounds, that finding drives decisions about additional beacon placement, recalibration or accepting the limitation for that area.
Indoor Navigation Data Maintenance
Maps degrade. Not because the file corrupts, but because the physical world moves on while the digital version sits unchanged. Maintenance is the unglamorous discipline that separates a working system from an abandoned one.
Change Detection
Someone needs to notice that a change has occurred. In practice this means establishing a reporting chain: facilities teams, shop fit-out managers or exhibition curators notify the map owner when a physical alteration happens. Relying on the map team to discover changes by periodically walking the space is fragile and does not scale.
Scheduled Audits
Even with a reporting chain, scheduled audits are necessary. Quarterly or semi-annual walk-throughs, comparing the map against the current space, catch unreported changes and verify that previously reported updates were applied correctly. The audit frequency should reflect how often the venue actually changes. A museum with rotating exhibitions needs more frequent checks than an office building with fixed desk positions.
Content Maintenance
Points of interest, labels, descriptions and associated content need attention alongside the structural geometry. A renamed meeting room, a relocated exhibit or a changed opening hours notice are content updates, not structural ones, but they still require a process. If the content lives in a separate CMS, the map geometry and the content layer can be updated independently, which is considerably easier than rebuilding a monolithic file.
Version Control for Indoor Maps
Without version control, every edit to an indoor map is an irreversible overwrite. A mistaken deletion, a corrupted export or an update that breaks routing cannot be rolled back. For any venue with more than a handful of spaces, treating map files as versioned assets is essential.
Why It Matters
Indoor maps are typically edited by more than one person over time — an integrator during deployment, an in-house operator afterwards, possibly a contractor for a specific refit. Without a clear version history, it becomes impossible to answer basic questions: when was this room last modified, who changed the corridor width on the second floor, or what did the ground floor look like before the recent refit?
Practical Approaches
If the map is stored in a text-based format such as GeoJSON, standard version-control tools like Git provide full history, diffing and rollback. This is straightforward for technical teams but may be unfamiliar to facilities or operations staff who need to make updates.
Proprietary platforms often include their own versioning — save-as, restore-point or audit-log features. The capability exists but varies widely in quality. Before committing to a platform, check whether it supports named versions, whether rollback is granular (restoring a single floor rather than the entire venue) and whether the version history is exportable or locked inside the platform.
At minimum, maintain a dated archive of exported map files outside the editing platform. If the platform fails or the contract ends, the archive provides a recoverable starting point rather than starting from nothing.
Naming Conventions
Adopt a naming convention that encodes the venue, the date and the nature of the change. A file named with the venue identifier, date in ISO format and a brief description — for example, venueA_2024-03-15_ground-floor-refit — is immediately meaningful in a file listing or version log. Avoid names like map_final_v2_updated, which convey nothing about what changed or when.
Version control for indoor maps is not an academic nicety. It is the mechanism that allows a venue to iterate safely, recover from errors and maintain an audit trail without which operational accountability becomes impossible.
