What RSSI measures
RSSI (Received Signal Strength Indicator) is a number that the radio chip of the worn device reports for every message received from a beacon. The values are negative, because they express the loss relative to 1 mW: the closer to zero, the stronger the signal. The result depends on the transmitter, the antenna, the orientation of the device and the surroundings.
The basic conclusion for a practitioner: RSSI is one input to the calculation, not the result. A distance in meters is produced only after many readings from many beacons are processed against the known layout of the transmitters.
Why signal strength is not distance
In theory, the signal weakens with distance. In an industrial hall, this relationship is disturbed by reflections from metal structures, attenuation by racking, the human body, forklifts and changes in humidity. Two measurements taken from the same spot a few seconds apart can give different values.
That is why a manufacturer quoting location accuracy in meters should say at what transmitter density and after what averaging it was achieved. A control question for the integrator is: from how many readings and how many transmitters is a single position calculated?
The most common sources of scatter
In practice, the same causes recur in halls, and they are worth recording in the pilot protocol.
- the employee's body between the device and the beacon, especially when the device is worn on the wrist or belt
- metal racking and machines reflecting the signal
- varying antenna orientation while the device is moving
- transmitters spaced more widely than recommended
- a beacon recorded in the system in a different place from where it physically hangs
Four RSSI readings and their average
- Reading
- Average of four
How the system limits position jumps
Signal strength is normalized and averaged in the system over a maximum of the four most recent readings. As a result, a single outlying measurement does not move the employee to the other end of the hall, while the position still follows movement, because the window covers only a short history.
The system does not rely on a single beacon. The position results from the signals of several transmitters averaged on the map, and it works without GPS, which matters inside buildings and under roofs, where GPS fails.
Example of averaging four readings
For illustration, take four consecutive readings from the same beacon: -70, -76, -68 and -72 dBm. The average is -71.5 dBm, and the deviation of the most extreme reading from the average is 4.5 dB. Without averaging, the system would use the last value, -72 dBm, or by chance the worst, -76 dBm. This is an illustrative calculation on assumed values, not a measurement from a specific plant.
Two modes of selecting the position
The simpler heartbeat mode selects the nearest location based on signal strength. The 3D mode calculates a three-dimensional position with an evolutionary algorithm and requires readings from at least 5 beacons. The choice of mode depends on whether the plant needs the name of a place (hall, door, warehouse) or a position on a building floor.
How to assess accuracy in the pilot
Accuracy is not assessed from a single walk through the hall. The pilot plan should include a control route with several points of known position, a few hours of work in the plant's traffic and a comparison of results at different times, when the number of forklifts and people changes.
The IT integrator should require a record of raw readings and of the processed result for the same moments. The difference between them shows how much work the averaging does and how much depends on the layout of the transmitters.
What to do when the position still jumps
If jumps remain after averaging, check the configuration data first, and only then the hardware: whether each beacon has a correct location and a unique identifier, and whether the density matches the recommendation (roads about every 100 m, buildings one transmitter per 100-200 m²). Additional transmitters are added only later, in places with weak coverage.
A sideways route spike followed by a return has a separate diagnosis, because it usually results from an incorrect beacon entry rather than from signal power. It is described on the page about route jumps.
Checklist
Video
Frequently asked questions
Can RSSI be converted to meters with a formula?
Formulas exist, but in a hall they give a wide scatter, because the environment changes the attenuation. That is why the system calculates the position from many transmitters and averages the readings.
Why does the position jump despite averaging?
A common cause is an incorrect beacon location entry, transmitters spaced too widely, or interference from racking and the employee's body.
Does a longer average always improve the result?
No. A longer window smooths out jumps but delays the position when the employee is moving. That is why the window is limited to the last few readings.
Sources
Check accuracy in your own hall
Together with you we will plan a control route and a measurement in the plant's traffic, so that the assessment of accuracy rests on data, not on promises.
Let's talk