Fleet fuel management for trucking carriers.

The five jobs a fuel program has to do

A fuel program is not a report and it is not a discount card. It is five jobs that have to happen in order, every day, for every truck. A program that does four of them leaks through the fifth.

• Decide where the truck should fuel

  Not the nearest station and not the cheapest sign on the interstate — the stop that is best once your contracted price, the detour off the lane, the driver's remaining hours and the fuel already in the tank are all taken together. This is a routing decision with a price attached, and it changes hour by hour as the truck moves.

• Get that decision to the driver before the tank forces it

  A correct stop that arrives after the driver has already pulled off is worth nothing. The decision has a deadline, and the deadline is set by the fuel gauge, not by the office. This is where most programs actually fail: the analysis is right and it arrives late.

• Confirm the fill actually happened

  A driver can be sent to a stop and not take it — hours ran out, the lane changed, the station was full. Without a confirmation step the program cannot tell a driver who ignored the assignment from one who was rerouted by dispatch an hour earlier.

• Reconcile against the fuel-card statement

  The fuel card is the only complete record of every fill, including the ones nobody planned. Matching card transactions back to assigned stops turns the program's claims into arithmetic: this stop was assigned, this transaction is the same fill, this is the difference against the alternative.

• Coach the fleet on what the reconciliation shows

  Compliance is not evenly distributed. Most fleets find that a small number of drivers account for most of the off-network fills, usually for reasons that are fixable once somebody knows. Without per-driver and per-lane numbers, coaching is a fleet-wide email that changes nothing.

Where fuel money leaks

Carriers with a negotiated discount network rarely lose money because the contract is bad. They lose it in the gap between the contract and the pump.

A driver at an exit with a quarter tank does not know which of the two stations in front of them is in-network, and has no way to find out that does not involve stopping and looking it up. The default is the nearest one. Repeated across a fleet and a month, the aggregate of those defaults is the leak — and because it shows up as ordinary fuel spend rather than as a line item, nothing in the accounting flags it.

The second leak is timing. A truck that runs the tank down too far loses the choice entirely; whatever is at the next exit is the stop. The third is the reverse, a truck topping off in small amounts at convenient stations, which keeps the tank comfortable and the discount unused.

None of these is a driver problem. They are all information problems, and they are solvable at the point where the decision gets made.

How DispatchFuel.ai runs it

We take the decision on for the fleet rather than handing it back as a dashboard. That is the difference between this and fuel software: nobody at the carrier has to log in for the program to work.

• We watch the tank, not the calendar

  Position and fuel level come from your telematics. When a truck is coming up on a fill, that is when the stop is worked out — against the lane ahead, your pricing and the driver's hours.

• The driver gets one message

  A Telegram message with the stop, the exit and the reason, answered with one tap. Drivers already use Telegram with dispatch, so there is no app to install and no training day.

• The fill is confirmed at the pump

  The driver confirms and keeps rolling. That confirmation is what separates a stop that was taken from a stop that was only sent.

• The card transaction closes the loop

  When the fuel-card transaction arrives it is matched back to the coordinated stop, so the savings line is recorded history rather than a projection. Actual and estimated savings are labelled separately and never blended into one number.

• Exceptions come to a person

  Stale telematics, unmatched transactions, a stop that was sent and not taken — these surface as exceptions for an ops desk to work, rather than quietly degrading the numbers.

What it connects to

Telematics for position and fuel level. Your fuel-card transaction feed for reconciliation. Your discount network pricing, in whatever form you hold it today. Telegram for driver communication. Google Sheets for import and export.

Named providers are confirmed with each carrier during the pilot rather than advertised here, because an integration list on a marketing page is a promise about somebody else's API.

Who this fits, and who it does not

It fits carriers running roughly 10 to 500 trucks that already hold a fuel-card contract with negotiated pricing, and that have nobody whose full-time job is making sure drivers use it. That combination — a good contract and no one minding it — is where the gap is widest. Larger and multi-terminal fleets are in scope too; the pilot is sized to one division so the rollout stays controlled.

Below roughly ten trucks the economics get thin, and we would rather say so on a call than after you have signed something. If you have no discount network at all, there is nothing yet for the program to route against; that contract comes first.