In short: An aircraft maintenance hangar is a steel box containing the largest metal objects in civil aviation, and WiFi behaves accordingly: coverage collapses inside fuselages, under wings and behind the aircraft itself. UK MRO sites have gone paperless on the assumption that engineers can reach the maintenance system from wherever they are working, and on most airfields they cannot.
Key Takeaways
- The aircraft is the obstruction — a wide-body fuselage is an aluminium tube that blocks 5 GHz almost completely, so an engineer inside the cabin, in a wing tank or on the flight deck is working in a radio shadow created by the thing they are fixing.
- Paperless Part-145 assumes coverage — electronic task cards, digital sign-off and certificate of release to service workflows only save time if the engineer can raise a finding and clear it at the aircraft, not at a desk in the office.
- One licensed cell beats twenty access points — a hangar covered from a small number of 3.8 GHz cells under an Ofcom Shared Access licence holds up when the doors open and the aircraft moves, which a WiFi survey done on an empty floor does not.
In a nutshell
Why does WiFi fail in an aircraft maintenance hangar?
Because every assumption a WiFi survey makes is wrong the moment an aircraft is towed in.
A base maintenance hangar is a clear-span steel structure, commonly twenty-five to thirty-five metres to the underside of the roof, with a door opening wide enough to admit a wide-body. Access points get mounted high on the roof steel because that is where the cable trays and the mounting points are. That geometry works in a warehouse full of cardboard. It does not work when the floor is occupied by a 60-metre aluminium cylinder with wings, because 5 GHz signal from above cannot get into the cabin, cannot get under the wing root, and reflects hard off every surface it does reach.
Then the aircraft moves. Tomorrow's check is a different type in a different position, possibly on jacks, possibly with the tailplane where the last one's nose was. The coverage map changes with it. Add docking, staging, work platforms, a paint bay with its own metalwork, and a hangar door that is open for a tow and shut for the next six weeks, and the radio environment is different every week of the year.
The practical result is familiar to anyone who has worked a check at Cardiff, Norwich, Prestwick or Stansted: the tablet works on the hangar floor, drops out on the flight deck, and is useless in the aft cargo hold. Engineers adapt by writing on paper and transcribing later, which is exactly the behaviour the paperless programme was bought to eliminate.
What is actually running on the hangar network?
More than most airfield IT teams realise, because the applications arrived one at a time.
The maintenance system is the backbone: AMOS, TRAX, Ramco or an airline's own platform, delivering task cards to tablets, taking findings, recording parts issued, and holding the record that supports the certificate of release to service. A wide-body heavy check generates thousands of task cards and a large volume of non-routine findings, each of which is raised, planned, worked and signed off.
Around that sit the systems that have quietly become wireless. Stores and parts issue, which increasingly means barcode or RFID scanning at the bin rather than a paper requisition. Tool control, which on a Part-145 site is a safety-critical inventory problem, not a stock problem. Test equipment and NDT kit producing data files that need to reach a server. Borescope inspection, where a video feed of a compressor blade is the evidence for a repair decision. Photographs of damage, corrosion and repairs, at a file size that makes them the largest single traffic type in the hangar.
Add the people: shift handover, planning conversations, the technical services team at head office being asked to look at something now rather than tomorrow. And add the contractors, because heavy maintenance runs on subcontracted specialists in painting, interiors, structures and avionics, all of whom arrive with their own devices and their own expectations about getting online.
That is an industrial network carrying operational traffic, and at a lot of sites it is running on the same access points the hangar office uses for email.
Can a paperless Part-145 hangar work without reliable coverage?
Not in the way the business case promised.
UK CAA Part-145 approval requires that maintenance data is available to personnel at the point of work and that records are kept to a defined standard. Electronic systems are accepted and widely used; the regulation does not care whether the record is paper or digital, only that it is controlled, accurate and available. What the regulation does not do is pay for the productivity gain. That gain comes entirely from removing the walk: the engineer raises the finding at the aircraft, the planner sees it immediately, the part is ordered before the shift ends, and the sign-off happens where the work happened.
Take the walk away and the electronic system is a paper system with extra steps. Engineers cache task cards before climbing into the aircraft, work offline, and sync when they come back down. Findings surface hours late. Planning works from a picture of the check that is always slightly out of date, and on a check with a hard redelivery date, hours of latency in the finding-to-part-order loop is aircraft-on-ground time nobody budgeted for.
Where operators have timed that loop, the lag sits in the coverage rather than in the software.
What about getting data off the aircraft itself?
This is the requirement that is growing fastest and the one hangar networks are least ready for.
Modern types generate substantial volumes of recorded data per flight, and the maintenance value of that data depends on getting it off the aircraft promptly. Airlines have been doing wireless quick access recorder offload over WiFi at the gate for years, and the same mechanism is used in the other direction to load software parts and database updates into avionics, cabin systems and inflight entertainment. Loading an IFE content refresh or an avionics software part over a congested access point at the far end of a hangar takes hours that the check plan does not have.
Hangars are where the heavy loads happen, because that is where the aircraft sits still for long enough. A cabin retrofit, an avionics upgrade or a full content load moves tens or hundreds of gigabytes into a single airframe. Doing that over a shared unlicensed channel that the whole hangar is also using is a bad plan, and doing it over a temporary cable run to the aircraft is what most sites actually do.
A licensed network with a dedicated slice for aircraft data transfer solves both halves of that: predictable throughput for the load, and no contention with the engineers' tablets while it runs.
Where do drones, borescopes and remote support fit?
Drone inspection of airframes has moved from demonstration to routine at several European MRO providers, using systems from vendors such as Mainblades and Donecle to photograph a full fuselage for lightning strike and hail damage in under an hour, against most of a shift for a manual inspection with a cherry picker. The output is thousands of high-resolution images that need to reach an analysis platform quickly enough to inform the same day's work.
Remote support is the other one. An engineer wearing a head-mounted camera showing a corrosion finding to the OEM's structures desk in Toulouse or Seattle, live, gets a repair decision in an hour rather than a day. That needs upload bandwidth from inside the aircraft, which is precisely where hangar WiFi does not reach. The same is true of borescope video from an engine on wing.
Neither application is exotic. Both are blocked by the same physical constraint.
Why does licensed spectrum change the picture?
Three reasons, in order of importance.
Power and propagation. Ofcom's Shared Access licence gives a hangar operator exclusive local use of spectrum in the n77 band, 3.8 to 4.2 GHz, at power levels that unlicensed WiFi is not permitted to use. A Low Power licence costs £80 per 10 MHz channel per site per year. That power budget is what allows a small number of cells to fill a thirty-metre hangar and get useful signal into a fuselage, rather than twenty access points fighting each other from the roof steel.
No contention. The hangar is not sharing its channels with the airport's passenger WiFi, the contractors' hotspots, the neighbouring FBO or the retail estate. On a busy airfield, unlicensed spectrum is the most crowded resource on site, a problem we covered from the terminal side in why airport WiFi buckles at peak load. Inside a hangar, the interference is closer and the consequences are operational rather than reputational.
Continuity across the airfield. A private network that covers the hangars, the apron in front of them, the engine shop and the stores gives one set of credentials to a workforce that moves between all of them. Line maintenance teams working the apron are on the same network as base maintenance in the hangar, which matters at sites where the same engineers do both.
What should an MRO do first?
Instrument the current network before specifying a new one. Most hangar WiFi has never been surveyed with an aircraft in it, which is the only condition that matters. Put a survey tool inside the fuselage, in the cargo holds, on the flight deck and under the wing on a live check, and compare that with the coverage map the installation was signed off against.
Then count the traffic honestly. Task cards are small; photographs, borescope video, drone image sets and avionics software loads are not, and it is those four that decide the sizing. A hangar carrying two heavy checks concurrently is an industrial site with industrial bandwidth requirements, and it should be specified like one rather than as an extension of the office network.