Why the recorded position does not point to the place
Two physical beacons several hundred metres apart that broadcast the same major:minor identifier are recognised by the system as one entry. A vehicle detected at either of them is placed in one saved location. The position in the database is therefore the place where the device is supposed to be, not the place where the error appears.
This is why searching in the field by the recorded position ends at the right device, which is not the culprit. The duplicate is the second copy, of which the database knows nothing.
Step 1: indicate the route points adjacent to the detection
The place where the vehicle actually was at the moment of detection is indicated by the neighbouring route points of the same locator, not by the recorded position of the beacon. For each jump you take the point before and the point after the detection of the duplicated identifier.
If you collect several dozen such points, a cluster forms on the map at the location of the second device. This cluster is the first approximation of the installation place.
Criteria for recognising a route jump
Step 2: calculate the centre of the cluster and convert it to hall coordinates
The centre of the cluster of distant readings is converted to coordinates using the georeferencing of the hall, with a convergence correction for the PUWG92 grid. Without the correction the point is off by tens of metres, and the technician searches in the wrong place.
Mark the result on the hall plan as an uncertainty circle, not as a point. The beacon lies in the area that the technician checks one by one: from the centre outwards.
Step 3: tell a duplicated beacon apart from a stop
A gap of 5 seconds between the true point and the jump points to a duplicated beacon, because in that time a vehicle cannot cover several hundred metres. A long pause means that the vehicle was stopped, not that the beacon is faulty.
Before going onto the floor, filter out the jumps with a short time gap. Keep the jumps after a long pause in a separate set for assessment.
Step 4: change the identifier and record it in the installation register
In the field, find the device, read its identifier and change it in one of the two physical copies. Moving the position of the entry in the database does not remove the error, it only moves it to the second place in the installation.
Record the new identifier, the place of installation and the date in the register. Do not remove inactive spare entries from the database, because the device may start broadcasting again.
Step 5: check the effect after the intervention
After the visit, check the number of route points recorded over the following minutes: an excluded beacon must not add points, while the others do. If jumps still appear, the second device is still broadcasting the old identifier.
Old routes keep their jumps, because the points are unchanging records of the moment of the signal. The effect of the fix is visible only in the routes recorded after the intervention.
Checklist
Video
Frequently asked questions
Is it enough to move the beacon position in the database?
No. Moving the position of a duplicated identifier does not remove the error, it only moves it to the second place in the installation.
How do you tell a duplicated beacon from a vehicle stop?
By the time gap. Around 5 seconds points to a duplicated beacon, a long gap means a stop.
Why is the PUWG92 correction needed?
Without the grid convergence correction the calculated point is off by tens of metres and the technician searches in the wrong place.
Will old routes improve after changing the identifier?
No, they keep their jumps. Only the recording of routes from the moment of the intervention improves.
We will show the duplicated beacon on your plan
We will review the route data and mark the search areas. Book a talk or a demo.
Let's talk