How positioning works
Accurate indoor positioning with zero configuration on the device itself. Blecon's BLE positioning system uses proximity-based detection — each device's position is determined by the hotspot that detects it. No device firmware changes, no complex setup, just hotspots listening and the cloud doing the work. It's a real-time location system (RTLS) that works indoors and outdoors.
The process
- Devices broadcast Bluetooth signals continuously.
- Hotspots detect those signals and measure the signal strength (RSSI).
- Hotspots report measurements to the Blecon cloud.
- The positioning engine uses the detecting hotspot's known location as the device's position.
How hotspots know their location
Hotspots determine their own position in one of three ways:
- Manual configuration — You set the hotspot's exact coordinates (GPS or local x/y/z). Most accurate.
- GPS — Hotspots with GPS modules determine their own position. Typical accuracy: 5–10 meters.
- WiFi/cellular — Hotspots estimate their location from nearby networks. Typical accuracy: 10–100 meters.
The system automatically prefers hotspots with better location accuracy and excludes those with stale data.
Position calculation
A device's position is the location of the hotspot that detects it. This gives room-level or area-level accuracy — if you know where your hotspots are, you know where your devices are.
When multiple hotspots detect the same device, the system automatically picks the best hotspot based on signal strength and location accuracy.
Collection window
The system collects measurements over a configurable time window (default: 90 seconds) before calculating a position. The positioning engine processes accumulated measurements in regular cycles. Longer windows provide more data but increase latency.
Network zones — positioning without coordinates
Where GPS isn't available (indoors, underground, dense urban areas), hotspots and reference beacons provide asset tracking without GPS through network zone positioning. Assign a network zone to a hotspot or place a reference beacon in a room, and any device detected there is reported in that network zone.
Network zones don't require surveyed coordinates — a single hotspot is enough. See Zone detection for setup details.
Accuracy
Position accuracy depends on:
- Hotspot location accuracy — Manually surveyed coordinates produce better positions than WiFi/cellular auto-location.
- Hotspot density — More hotspots means finer-grained proximity detection.
- Environment — Walls, metal, and water absorb Bluetooth signals and affect which hotspot detects the device.
| Configuration | Typical accuracy |
|---|---|
| GPS-positioned hotspots | Area level (5–10 meters) |
| WiFi/cellular auto-located hotspots | Broad area (10–100 meters) |
| Network zones (no coordinates needed) | Room or area level |
The accuracy value
Each position event includes an accuracy_meters value — an estimated accuracy radius in meters. This is not the raw GPS accuracy of the hotspot. It is a calculated value that accounts for two factors:
- Hotspot GPS accuracy — How precise the hotspot's own location fix is.
- Location freshness — How much time has passed since that fix, and how far the hotspot may have moved.
A phone standing still with a fresh GPS fix might produce a 5 m accuracy value. The same phone after walking for 10 seconds without a new fix might show 15 m — because the phone could have moved since its last known position. The accuracy value widens to reflect that real-world uncertainty.
In short: it's not that GPS is less accurate — it's that the system is being honest about the fact that a GPS reading from a few seconds ago doesn't perfectly represent where the hotspot is right now. A stationary hotspot will show accuracy much closer to its raw GPS.
You can filter out low-accuracy positions using the accuracy filter.
Two timestamps
Each position event includes two timestamps:
- Sample time — When the signal measurements were taken.
- Calculated at — When Blecon calculated the position from those measurements.
These may differ by seconds (real-time) or longer (if the device was offline and uploaded measurements later).