In short: The aircraft turnaround — the choreography of fuelling, catering, cleaning, baggage and boarding that happens at the stand between arrival and departure — is where airport punctuality is won or lost. Airport Collaborative Decision Making (A-CDM) is supposed to coordinate it, but A-CDM is only as good as the data feeding it, and most of that data is generated by vehicles and crews on the apron where the network barely reaches. Private 5G is what turns the turnaround from a series of phone calls into a live process.
Key Takeaways
- The turnaround is a dozen processes racing one clock — fuel, catering, cleaning, water, baggage, boarding and pushback all converge on a single stand in well under an hour, and a delay in any one ripples into the next rotation
- A-CDM runs on milestones that someone has to report — the target off-block time depends on accurate, real-time confirmation that each step is done, and today that confirmation is often a radio call or a back-filled timestamp
- The apron is the worst-connected place at the airport — open ground, metal aircraft, dense moving vehicles and shifting demand defeat Wi-Fi, leaving the most time-critical operation on the least reliable network
In a nutshell
Where airport punctuality is actually decided
Ask most travellers where delays come from and they will point at air traffic control, the weather, or the airline. All of those matter. But a very large share of European short-haul delay is generated, or recovered, on the ground — in the 25 to 50 minutes an aircraft spends on stand between one flight and the next. That window is the turnaround, and it is the single most concentrated piece of operational choreography at any airport.
Consider what has to happen at a single narrowbody stand in well under an hour. The aircraft is marshalled in and chocked. Ground power and, often, pre-conditioned air are connected. Passenger steps or an airbridge dock. Inbound passengers disembark while the inbound hold baggage is unloaded. The cabin is cleaned and catering is swapped. Potable water is replenished and the lavatories are serviced. Fuel is loaded — sometimes with passengers still aboard, under strict conditions. Outbound bags are loaded against the final load sheet. Outbound passengers board. Doors close, the pushback tug connects, and the aircraft is dispatched.
That is a dozen separate processes, run by several different companies — the airline, one or more ground handlers, a fuelling company, a catering firm — all converging on one piece of tarmac, all racing the same clock. When they are coordinated, the aircraft pushes back on time and the next rotation starts clean. When they are not, the delay does not stay put: it follows the aircraft through the rest of its day and spreads across the network.
A-CDM: a good idea that depends on data nobody owns
The industry's answer to this coordination problem is Airport Collaborative Decision Making — A-CDM — now implemented at Heathrow, Gatwick, Manchester, Stansted and a growing list of UK and European airports, and tied into the EUROCONTROL network so that one airport's departures are sequenced into the wider European flow.
A-CDM works by tracking a sequence of milestones for each flight and using them to calculate and continuously refine a Target Off-Block Time. The logic is sound: if everyone shares accurate, real-time information about where each turnaround actually is, the airport can sequence departures intelligently, allocate stands and de-icing better, and stop wasting slots on aircraft that are not really ready. The whole edifice rests on one assumption — that the milestones are reported accurately and promptly.
This is exactly where it strains. Many of the milestones that drive A-CDM are generated out on the apron, by people and vehicles, and reported through whatever channel happens to be at hand: a radio call to a control room, a tap on a handheld that may or may not have signal, a timestamp entered after the fact. When the "boarding complete" or "fuelling complete" milestone is a human relaying information over VHF, the data feeding the system is only as timely and accurate as that call. A-CDM does not fail loudly when this happens; it quietly degrades, the off-block predictions drift, and the airport makes worse decisions on the back of stale information.
The interesting point is that A-CDM is not really a software problem at this stage of its maturity. The platforms exist and are good. It is a data-capture problem, and data capture on the apron is a connectivity problem.
Why the apron defeats the obvious networks
The apron is, from a network engineer's point of view, close to a worst case. It is wide open ground measured in hundreds of metres, so signal has a long way to travel. It is full of large metal objects — the aircraft themselves — that move, reflect and shadow radio constantly. It is dense with vehicles, each of which is increasingly a connected device in its own right. And the demand is non-uniform and bursty, spiking hard around each turnaround and moving from stand to stand through the day.
Public Wi-Fi was never designed for this and does not survive it. Coverage from a handful of masts is patchy between aircraft; handover as a vehicle drives across the apron breaks sessions; and contention collapses exactly when a bank of departures has every stand working at once. Public mobile networks help with wide-area coverage but offer no guarantees, no control over capacity at the moments that matter, and no ability to prioritise a safety-critical pushback feed over someone streaming video in the terminal.
Private 5G is a much better fit for the physics. A network designed for the apron provides consistent coverage edge to edge, seamless handover as vehicles move between cells, the capacity density to handle every stand turning around at once, and — crucially — the ability to slice and prioritise so that the data that drives the turnaround is never competing with everything else for room. It is the difference between hoping the milestone report gets through and knowing it will.
From phone calls to a live process
Put the right network across the apron and the turnaround stops being a sequence of phone calls and becomes a live, observable process. The change is less about any single gadget than about everything being connected to one reliable layer at the same time.
Every ground-support vehicle — the tug, the belt loader, the fuel bowser, the catering truck, the potable water unit — can report its position and status continuously, so the control room sees the turnaround assembling in real time rather than being told about it. Each milestone can be captured automatically or with a single confirmed tap that actually reaches the system. Stand-mounted cameras and sensors can confirm chocks on, airbridge docked, and doors closed without anyone radioing it in. The load sheet, the final figures and any last-minute changes move cleanly between the handler, the airline and the aircraft.
The payoff is twofold. The airport gets A-CDM milestones it can trust, which means off-block predictions that hold up and departure sequencing that reflects reality. And the handler gets a real-time view of whether each turnaround is on plan, so that the one process running late — the catering truck stuck two stands away, the bag load behind schedule — is visible while there is still time to recover it, not after the aircraft has missed its slot.
The honest trade-offs
We are wary of overselling this, so it is worth being clear about the limits. A private network on the apron does not, by itself, make a late turnaround quick. If the catering company is short-staffed or the inbound aircraft landed forty minutes down, no amount of connectivity recovers the time. What the network does is remove the information delay — it makes the true state of every turnaround visible early enough to act on, and it makes the data A-CDM consumes accurate enough to be worth consuming.
There is also a genuine multi-party complication. The apron is shared by the airport operator, several handlers and the airlines, and a network that is going to carry their operations needs to be neutral, governed and trusted by all of them rather than owned by one party to the disadvantage of the others. That is an organisational question as much as a technical one, and it is one of the reasons apron connectivity has lagged terminal connectivity despite being more operationally important.
Finally, the case is strongest where the turnaround pressure is highest. A regional airport running a dozen movements an hour has a different calculus from a slot-constrained major hub where every minute on stand has a measurable cost. The technology scales down, but the return scales with how tight the clock is.
Where we would start
Our view is that the apron, not the terminal, is the part of the airport where private 5G earns its keep first — because that is where the most time-critical, most multi-party, worst-connected operation actually happens. We would start by mapping the turnaround as it really runs at a given airport, identifying which A-CDM milestones are currently reported by hand or back-filled, and putting a managed network across the stands that turns those into automatic, trustworthy data.
Ultimately however, the turnaround is a problem the industry has been trying to solve with better processes and better software for two decades, while leaving the layer underneath — the network on the apron — largely untouched. We think that is the wrong way round. Get the connectivity right at the stand, and A-CDM finally has the real-time truth it was always designed to use.