The ramp already knows what to do next.
A board sorted by departure time tells you who leaves first, not who to start on first. A quick top-up leaving at noon can sit behind a full service leaving at one. Up Next ranks the open work by when the crew has to begin, using how long each job actually takes at your field, and re-sorts itself as orders land and times slip.
A busy ramp needs more than a departure time.
Most ramp boards sort by departure time and stop there. That works until the ramp is busy. Then it gives the wrong answer two ways.
A time sort throws away what it knows
A walk-in on the ramp right now has no departure time on the ticket, so it drops below an aircraft leaving at eleven tonight. Unknown urgency is not low urgency. And once several orders are past due they all read as equally urgent, when three hours late and five minutes late are different problems. The crew can see that. The board cannot.
Leaving first is not starting first
A lav service leaving at 09:30 takes ten minutes. A fuel uplift leaving at 09:40 takes twenty-five. In departure order the second aircraft leaves late. Do the fuel first and both make it. Only the length of the job tells you that.
Three factors. You decide what each one is worth.
Every live order is scored on three things. You set how much each one counts, because what matters on your ramp is your call, not ours. A chain sets it once, for every airport it runs.
How close the aircraft is to the time the crew has to start. That is its departure time, less the work still to do on it. Past that moment the score carries on climbing, so a badly late aircraft outranks the rest. Late does not just add points. It takes over.
How long the request has sat unworked. This is what carries an order with no departure time on it, which is most walk-ins. An order nobody has touched climbs on its own.
Your own ranking of the customer, set on their record and kept private to you. It is held apart from their price tier on purpose. A customer on a keen rate can still be the one you look after first, and a discount should not reorder your ramp.
Late beats everything else
Waiting time and ramp priority stop at their weight. Departure time carries on past the due time, up to twice its own weight. On the default split that is enough for a badly late aircraft to outrank everything else put together, which is the point. When every job left is hours late, the queue falls back to whichever has been due longest.
A MedEvac goes first
A medical flight is a switch rather than a weight. Nobody wants to work out what a MedEvac is worth against a tenant customer. They want it worked first. The switch is on by default, it lifts the order above the whole queue, and the row says why it jumped.
Settings, Modules, Up Next. The MedEvac switch and the work-time allowance sit apart from the weights.
Nobody types in how long a fuel stop takes.
To know when the crew has to start, the queue needs to know how long the job takes. Most systems make you type that in once and never look at it again. This reads it from what your own crew actually did. Turn it on under Settings, Modules, Up Next and the queue switches from departure time to start time.
How long the job takes
Every finished job is a stopwatch reading you already took. The queue allows for the median of the last ninety days at your field, per service, so fuel here takes what fuel here takes. A field too new to have ninety days of its own borrows from the rest of the chain until it does.
How much of it happens at once
Two crew on one aircraft means the work overlaps, and simply adding the jobs up would rank it wrong. That gets measured too, from the same timestamps. A field that works services back to back is treated differently from one that runs them together.
And it corrects itself
There is nothing to maintain and nothing to re-enter. A crew that gets faster sees the estimate come down on its own, because it is read from a rolling window.
The Service Times report shows the same medians the queue works from, so you can see the numbers yourself. A service with too few jobs behind it says so on the row.
Every position shows the math.
Hover the position on the orders list and the whole calculation opens up. Start by 00:56. The work that time allows for. What each factor added. The measured minutes behind each service. Nobody has to take the queue's word for it, including the customer on the phone asking why they are not first.
What it deliberately does not do.
Plenty of software would sell this as artificial intelligence. Here is what it actually is, and where we drew the lines on purpose.
No AI, and nothing to train.
The rank is plain arithmetic on things you can see: a time, a clock, a number you set. It runs in the database, updates on every change, and gives the same answer twice.
It does not dispatch your people.
The queue says what is next. Who takes it is your crew's call, because the person on the ramp knows things the software does not. Nothing here pulls a fueller off an uplift. A job under way stays under way.
It will not guess how long a job takes.
A service nobody has ever timed is left out of the estimate. A service the chain has timed elsewhere borrows that figure until your own field has enough jobs behind it, and the report tells you which of the two you are looking at.
The number is per airport.
Ramp work happens at one field, so position 3 means third in line here, out of however many are live. On a list covering several airports the sort still holds, but the position is hidden, because nobody is third in line across four airports.
Bring us a busy morning.
Book a demo and describe the worst hour on your ramp: the tenant, the transient, the medical flight and the fuel truck that all want something at once. We will put it on the board and you can see what order it comes out in.
Screenshots are of RampRelay running on demonstration data. Aircraft, customers and staff shown are fictional.