InnerWeb
Skip to main content

Why does a route jump on the map? FAQ for dispatchers and system administrators

A delivery route on the online map is a path between the positions of beacons detected by the driver's phone, not a GPS track. That is why a sudden jump of the line to a distant place almost always comes from the data of one beacon, not from the vehicle's movement. Below are the answers to the questions dispatchers and system administrators ask when they see a line on the map that the driver never drove.

151 of 651
jumps disappeared after silencing one beacon
Source: card
approx. 5 s
interval between points for real duplicates (400 to 650 m)
Source: card

Where the route line on the map comes from

The first question is usually: is this a GPS track? It is not. The delivery route on the online map is a path connecting the positions of beacons that the driver's phone detected along the way. The system does not draw the geometry of the driven road, only the successive points at which the device saw a known beacon, connected by lines.

This gives a simple diagnostic rule. If the line suddenly jumps to a distant place, check which beacon was detected at the point of the jump and what position it has in the database. The driver's device itself is rarely the cause.

The name shown on the map next to a point comes from the location assigned to the beacon, not from the beacon itself. When the name looks odd or points to a place the vehicle never visited, this is an additional sign that the assignment in the database is wrong.

A beacon by the road, but recorded inside the warehouse

The most common cause of a jump is a wrong position recorded in the database. If a beacon mounted by the road has a position recorded inside the warehouse, every delivery that detects it draws a jump to that place and back. The line goes from the road into the hall and then straight back onto the road.

Such an error shows up in many deliveries at once, always at the same point of the route. This is what distinguishes it from a single device anomaly. A dispatcher who compares several routes over the same section will quickly find the common jump point.

The good news is that correcting the recorded position of the beacon removes the jump for all future deliveries. One change in the data fixes all subsequent routes, without a site visit and without changing the software.

Jumps removed by silencing one beacon

Jumps removed by silencing one beaconDonut chart: 151 of 651 jumps disappeared after silencing one beacon, 500 remained.Disappeared after silencing one beacon: 151 jumps (23.2%)Remaining jumps: 500 jumps (76.8%)651jumpsDisappeared after silencing onebeacon151 jumps (23.2%)Remaining jumps500 jumps (76.8%)
Source: Result of one data scan from the function documentation (651 jumps)

Duplicate identifier and the major:minor pair

The second scenario is harder. Two physical beacons several hundred metres apart can broadcast an identical major:minor pair, and the system then recognises them as one beacon. The phone sees one or the other at different times, but the database holds one row with one position, so half of the detections land in the wrong place.

With a physical duplicate of the identifier, changing the recorded position does not remove the error, it only moves it to the other place. If you move the row to the first beacon, the jump appears at the second one. Here one of the devices has to be silenced or physically reprogrammed.

A separate situation is a single navigation point with two rows in the database, sharing the same minor number and different major numbers. Only one of them is actually detected, the other is a dormant spare. Which is which is established by the number of route points: the detected one has many, the spare has zero.

  • The jump is always in the same place, across many deliveries: suspect the recorded position of one beacon.
  • The jump moves after a change of position: suspect a physical duplicate of the major:minor pair.
  • Two rows with a shared minor: check the number of route points of each of them.

How to tell a duplicate from a parked vehicle

The true position of the device is established from the neighbouring route points. A point that jumps away from the previous one and to which the next point returns is an artefact, and the previous point shows the actual place of the vehicle. In one of the scans, a jump was defined as a point more than 55 units away from the previous one, after which the next point returns close to the previous one (under 21 units). The scan covered over one million route points from 14 days.

A trap is a vehicle that stood still and then started from a different place. This is not proof of a duplicate. Real duplicates had an interval between points of the order of 5 seconds, too short to cover 400 to 650 m. A vehicle that stood for an hour and then appeared a few hundred metres away has a long gap in time, not a duplicated identifier.

That is why, for every suspicion, look at two things at once: the distance between the points and the time between them. A large distance in a very short time is a signal of a duplicate, a large distance after a long pause is ordinary driving.

Apparent culprits and the one real one

Many beacons that seem guilty are only the return points of one real culprit. The line jumps to beacon A and then returns through B and C, so all three appear on the list of suspects. In one of the scans, 151 of 651 jumps disappeared after silencing a single beacon.

The conclusion for the administrator: do not fix all beacons on the list at once. Start with the one that appears in the largest number of jumps, silence it or correct it, wait for new route points, and only then assess what remains.

What to do after detection and why old routes do not change

A duplicate is silenced by changing its major number to a value that no device broadcasts. The position, minor and location remain, so restoring it is a change of one number. However, remember that both beacons of the pair then drop out of positioning until a site visit, during which one of them gets a new identifier.

After changing the identifier, the positioning cache has to be synchronised without clearing all of it, and you have to check that the silenced beacon no longer records new points while the others still do.

Route points are immutable records of the moment the signal was received, so correcting a beacon does not rewrite history. Deliveries from before the correction keep the jump in the preview. This is intended behaviour, because the history is meant to reflect what the system recorded at the time. When reporting, therefore, cut the data off at the date of the correction.

Inactive spare beacons are not deleted, because they may start broadcasting at some point. Spares can be given a name with an added zero to distinguish the pair.

Video

Frequently asked questions

Does a jump on the map mean the driver left the route?

Usually not. The route is a path between positions of detected beacons, so a jump most often comes from a wrong position of one beacon in the database or from a duplicated identifier. Check the point of the jump and the beacon that caused it.

Will correcting a beacon position fix the old routes?

No. Route points are immutable records of the moment the signal was received, so deliveries from before the correction keep the jump in the preview. The correction removes the jump for all future deliveries.

Can I delete a spare beacon from the database?

You should not. An inactive spare beacon may start broadcasting at some point, and deleting the row would leave detections without a matching record. Mark it with a name that has an added zero.

How do I know that a silenced beacon no longer records points?

After synchronising the positioning cache, check new route points: the silenced beacon should no longer create them, while the others still do. If the others also stop, the clearing covered too much.

See how the system shows the route on the map

Book a conversation or a demo: we will show you how to read the route points and how to check which beacon is behind a jump.

Let's talk