Define routes, zones and points of interest
Bluetooth beacon-based positioning uses the signal strength measurements from multiple beacons to estimate where a device is within a space. It sits alongside simpler proximity detection, where a single beacon only tells you that a phone is nearby, not precisely where. Positioning aims to place that device on a map, or at least within a defined zone, by comparing readings from several beacons at once.
Two main methods drive most deployments. Trilateration uses the estimated distance from three or more beacons to calculate a coordinate. Those distances come from the Received Signal Strength Indicator (RSSI), which is converted to a rough distance using a propagation model. Fingerprinting takes a different approach: during a survey phase, you record RSSI values from every beacon at many reference points across the floor, building a radio map. When a user's phone reports its own RSSI readings, the system compares them against that map and finds the closest match.
Neither method delivers a fixed accuracy figure you can quote before measuring the actual environment. RSSI fluctuates with antenna orientation, the phone model, whether the user is holding the device or it is in a pocket, and how many people are standing nearby. A corridor with plasterboard walls will behave differently from an open retail hall with metal shelving. Any supplier or integrator offering a single accuracy number without a site survey is extrapolating from other buildings, not describing yours.
Beacon-based positioning also requires a receiver that can see enough beacons simultaneously. In a native app, this is straightforward: the app scans in the background and reports RSSI values to a positioning engine. In a web-based flow, the browser's Web Bluetooth API has significant limitations on background scanning and OS support, which constrains what is practically achievable without an app installed.
Design zones, paths and instructions together
Retail environments
In a supermarket or large retail store, beacon-based positioning can support aisle-level navigation rather than the shelf-level precision some vendors imply. A typical deployment places beacons on ceiling tiles or high on shelving uprights at regular intervals. The useful outcome is usually "you are near the bakery aisle" or "turn left at the end of this row", not "you are standing 30 centimetres from the tinned tomatoes". For queue detection and dwell-time analytics, zone-level positioning is often sufficient and considerably more reliable than attempting sub-metre accuracy.
Museums and galleries
Museum positioning tends to work well with zone-based approaches because visitors move between distinct rooms or gallery sections. Placing a beacon at each entrance or in the centre of a room allows the system to identify which exhibit area the visitor is in and trigger relevant audio or text content. The relatively slow pace of movement and the clear physical boundaries between zones suit fingerprinting or simple nearest-beacon logic more reliably than a complex trilateration model.
Events and temporary venues
Temporary event spaces present a particular challenge: the radio environment changes every time you set up. Staging, draping, crowd density and temporary walls all alter signal propagation. A fingerprinting survey done during setup with an empty hall will not match conditions during a packed keynote session. For events, a simpler zone model with generous boundaries and frequent recalibration checks is usually more robust than attempting fine-grained positioning.
Placement and density decisions
Beacon density depends on what you are trying to achieve. For zone detection, one beacon per zone may suffice if zones are large and clearly separated. For trilateration, a device typically needs line-of-sight or near-line-of-sight to at least three beacons, which means higher density and careful attention to what sits between beacon and phone. Mounting height matters: ceiling-mounted beacons at three to four metres give a clearer path than waist-height placements that are easily obstructed by people and stock.
When planning density, work from the requirement backwards. Define the smallest zone or the coarsest position you genuinely need, then calculate how many beacons are required to guarantee that a phone in any part of the space can hear at least three with usable signal strength. Add a margin for beacon failure, because one flat battery should not collapse the positioning grid in an area.
Keep routes accurate after building changes
Assuming uniform accuracy
The most frequent error is treating the positioning system as though it performs equally well everywhere in a venue. In practice, accuracy degrades near walls, in corners, behind large fixtures and in areas where only two beacons are reachable. If your wayfinding logic assumes a consistent accuracy across the floor plan, users will receive wrong-turn instructions precisely where they are most confused. Map your accuracy expectations by zone during the pilot, not just as a single average.
Ignoring human body attenuation
The human body absorbs Bluetooth signals significantly, particularly at 2.4 GHz. A phone held in front of the user will report different RSSI values from the same beacon compared to a phone in a back pocket or a handbag. In dense crowds, the cumulative effect of dozens of bodies between beacon and phone can shift readings enough to push a position estimate into an adjacent zone. Test with devices in realistic carry positions, not just held flat in the hand.
Not accounting for environmental changes
Retail environments change constantly. Seasonal displays, new shelving configurations, stock piled high on gondola ends and even the presence of a promotional stand can alter the radio map. If you used fingerprinting during an initial survey, the map will drift over time. Schedule periodic re-surveys or build a feedback mechanism where the system flags zones where positioning confidence has dropped, prompting a physical check.
Over-relying on RSSI without fingerprinting
Trilateration using a simple propagation model is attractive because it avoids the survey effort of fingerprinting. In an open, unobstructed space with minimal multipath reflection, it can work adequately. In most real indoor environments, the model breaks down because RSSI does not decrease neatly with distance when signals bounce off walls, floors and metal fittings. If you choose trilateration over fingerprinting, validate it thoroughly in the actual space and accept that accuracy will vary considerably across zones.
Key checks before committing
- Run a pilot with a representative set of phone models, not just the latest flagship devices. Mid-range and older handsets have different Bluetooth radios and antenna positions.
- Measure positioning performance at different times of day and with different occupancy levels, not just during a quiet morning survey.
- Test what happens when one or two beacons in a cluster fail or are removed. Does the system degrade gracefully or produce wild position jumps?
- Ask your integrator or platform provider to explain their positioning algorithm in practical terms: what happens at zone boundaries, how quickly does it update, and what is the confidence threshold below which the system suppresses a position estimate?
- Confirm whether the system requires a native app or can operate through a browser, and if the latter, verify OS and browser support for the specific devices your visitors actually use.
- Check how calibration data is stored and updated. If the radio map lives only on the device, every app update or floor-plan change requires re-downloading a large dataset.
Bluetooth beacon-based positioning is a useful tool when matched honestly to what the environment and the use case can support. It works best as zone-level guidance backed by clear signage, not as a substitute for physical wayfinding infrastructure. The next step is to compare this approach against other signal-based methods, such as Wi-Fi-based indoor positioning, which trades off different strengths and constraints.
