InnerWeb
Skip to main content

Checklist after changing beacon data: the administrator's control gate

After changing the position or identifier of a beacon, the administrator has four things to check: whether the positioning cache has been synchronised, whether a muted beacon has stopped recording new route points, whether the others still record them, and whether the old values were saved in case the change has to be reversed. The checklist below turns these checks into a gate that closes the change.

1 number
enough to mute a duplicate and to undo the muting (changing the major)
Source: card

Why a control gate after changing data

Changing data in a beacon record looks simple: you edit one number. The effect, however, shows up only in the route points recorded in the following hours, so saving the change proves nothing on its own. The control gate is a set of checks after which the administrator may consider the change finished.

In systems where many deliveries pass the same point, an error in the data of one beacon repeats on every route. A change made without checks can turn one error into two: fixing a jump in one place and muting a beacon that was needed.

Recording the state before the change

Before the change, record the old values: the position, the major and minor numbers and the assigned location. We mute a duplicate by changing its major number to a value that no device broadcasts, while the position, minor and location stay. Restoring it is then a change of one number, but only if you know what the previous one was.

Also note how many route points each beacon of the pair recorded before the change. From this number you will read after the change whether the right one of the pair was muted: the active one has many points, the spare zero.

Remember the side effect of muting: both beacons of the pair drop out of positioning until the field visit. Plan that visit and record which beacon will get the new identifier.

Synchronising the positioning cache

After changing an identifier, synchronise the positioning cache without clearing the whole cache. Clearing the entire cache is unnecessary and risky, because for a short time it deprives the system of data on all beacons, not only on the one whose change concerns.

After synchronisation, check whether the system sees the new values. If the reading of the beacon still shows the old value, the cause is usually an unsynchronised cache, not a write error in the database.

New route points: what should change and what should not

After the change, check that the muted beacon no longer records new points, and that the others still do. Both conditions matter equally. A lack of new points from the other beacons means the change or the synchronisation covered too much, while new points from the muted beacon mean the change did not work.

When assessing, read the time of points in one time zone. A difference in time between the record in the database and the view can be the effect of an offset of the time zone, not a failure, so before you conclude that there are no points, make sure you are searching the right time window.

Do not expect a change in old routes. Route points are immutable records of the moment of the signal, so correcting a beacon does not rewrite history, and deliveries from before the fix keep the jump in the view. Correcting the position removes the jump for all future deliveries.

Closing the change

You close the change when all checks are ticked and the documentation contains an entry: what was changed, when, from which value to which, and who is to make the field visit and when. The entry saves time at the next route jump, because it lets you tell a new error from an unfinished correction.

Spare inactive beacons are not deleted, because they may start broadcasting one day. A spare may be given a name with a zero added, to distinguish the pair.

Checklist

Video

Frequently asked questions

Does the whole positioning cache need clearing after muting a beacon?

No. It is enough to synchronise the positioning cache without clearing the whole cache, which limits the risk of losing data on the other beacons.

Why do old routes still show a jump after a fix?

Route points are immutable records of the moment of the signal. A correction of the beacon does not rewrite history, so deliveries from before the fix keep the jump in the view.

How do I undo the muting of a beacon?

Muting consists in changing the major number to a value that no device broadcasts, so undoing it means changing that one number back to the previous value. That is why the old values must be recorded before the change.

Tidy up beacon data changes

Book a conversation: we will show you how to close a change to beacon data with evidence in the route points.

Let's talk