The tracker went in on Monday. The truck was out working by lunchtime, and on the map it looked perfect. The fuel sensor fitted with it said nothing. Not zero, nothing, which on a screen looks exactly like a sensor that is broken.
The old answer
Bring the truck back. Lose a day of it. Pay a technician to open the cab, find that the sensor was paired to the wrong slot, change one setting, and leave. Multiply by every fitting, every sensor, every device that moves to another vehicle when a truck is sold.
The new answer
From the office, while the truck was on the road: read what the tracker actually held, see the setting that was wrong, change it, and have the device report back what it now holds. The fuel level appeared within the hour. The whole thing took three minutes and the truck never knew.
The part that matters is the read-back
Sending a setting is easy. Knowing it took is the work. A device that was told to change and did not, or that was told the wrong thing and accepted it, looks fine from the office until somebody notices the fuel page is empty a fortnight later. So no change counts until the device is asked what it now holds and answers with the value that was sent. A mismatch is shown as failed, never assumed to have worked.
And the record
Every change is on the device’s record: who, what, when, and what the device answered. Each device’s settings are snapshotted daily, so a change nobody made is noticed. The settings that could cut a truck off from the office are protected behind an explicit unlock and a reason. It is not glamorous. It is what makes a fleet of trackers still correct three years after the fitting.
For what can be set up remotely and what it costs per vehicle, see setting up vehicle trackers from your desk.