Hand the optimizer a route you already have. It reads the traffic for the minute you'll actually be on that road, the delivery window on each stop, which customers come first and where the driver's break has to fall. Then it puts the stops in an order that holds up against all of that at once. Every route comes back priced, so the call that tells you what arrives on time also tells you what the day earned. Attached to your own order? It will drive that one on paper first and show you where it comes apart.
All three routes below call at every stop on the sheet. The only thing that changes between them is the order they are driven in. Same vehicle, same driver, same day. Hand-ordered is the sheet as it was handed over. The generic AI optimizer is what a large language model chose when given every stop, window and travel time and several minutes to think. Our AI optimizer is the order our own engine returns in seconds. The first two sheets went out to real drivers. The rest are generated.
Thirty stops is already awkward to sequence by hand. Sixty is guesswork with a clipboard. Every stop you add multiplies the orders you would have to weigh against each other, and the day keeps moving while you do it. The bars show the on-time rate for each route in the comparison, as it was submitted and after reordering.
The cost side moves further than most people expect, and it moves on two fronts at once. Take the first real sheet above. Reordered, the same thirty-two deliveries cover 146 km instead of 208, and the time the driver spends waiting for a window to open falls from three hours to half an hour. The paid day goes from 12h 21m to 8h 34m. Nothing was dropped and nothing was rushed. The stops were visited in a better order, and that was the whole intervention.
Which of the two gives back more depends on how full the day is. On that sheet it is the waiting, about two thirds of the hours recovered, because the driver kept arriving before the window opened and standing there. On the 59-stop sheet beside it the day was over-subscribed and there was no waiting left to take out, so five sixths of the saving came from the driving instead: 424 km down to 187. A driver costs more per hour than the van costs per kilometre, so payroll is the larger line on the first sheet and the two are level on the second. The hours are what you are buying. The litres come with them.
Send up to ten routes in one request and they get planned together.
It looks for a sequence that satisfies everything on the sheet at once: the delivery window on each stop, which customers come first, how long each drop takes, and where the break has to sit. The shortest loop almost never wins here, because the shortest loop has no idea what time it is.
Send the sequence you were going to drive anyway. It runs that exact order against the same traffic model and tells you which stops miss their window, and by how much. Nothing is reshuffled behind your back.
One call runs both passes over the same sheet and hands them back together. Your order and the suggested one are judged the same way, on the same day, against the same traffic. Every number in the comparison above came from that call.
Every route comes back with a profit figure as well as a plan. Driver, fuel, upkeep, insurance, standing costs and traffic fines on one side; the fees of the drops that were actually reachable on the other. Pay your drivers by the hour and it counts the hours the route actually takes, so a shorter day shows up as a cheaper one, and you can see which line moved. The figures are kept, so a month's profit for one vehicle is one call away.
Set the sliders to your fleet. Every number below is your inputs against the two real runsheets at the top of this page. Where the two sheets differ, the smaller saving is the one used, and the breakdown shows exactly where it comes from.
—
— a year in driver time and running costs.
Driver hours the reordering gives back, across —.
More a month on the same fleet and payroll, worth — at your fee.
| Minutes | As submitted | Reordered | Saved |
|---|---|---|---|
| Driving | — | — | — |
| Waiting for the window | — | — | — |
| At the door | — | — | 0.0 |
| Total | — | — | — |
Time at the door is yours and does not move. The saving is the driving and the standing about, which is where of it is.
| AED | Basis | Amount |
|---|---|---|
| Driver time | — | — |
| Fuel and running | — | — |
| Total | — | — |
On-time stops go from — to — on these sheets, so the hours come back without the day getting worse.
We set the account up for you. After that it is yours. Sign in, create an API key, and send your own routes through the endpoints that produced every number on this page.