44 Weeds: The Hidden Value of Non-Hazard Data
Convincing an airfield operator that classic FOD matters is not hard.
A piece of metal, a chunk of concrete, or a strip of tire on a runway is an immediate safety event. The industry has spent decades putting numbers on that risk. Boeing’s long-cited figure puts the global annual cost of FOD at about $4 billion. A later bottom-up analysis by Insight SRI put direct airline maintenance costs at $1.1 billion a year considering only the top 300 airports, or roughly $263,000 per 10,000 movements, and estimated that delays, aircraft changes, and related disruption can multiply that cost by as much as ten times. FAA-sponsored work has put the fully loaded global figure as high as $22.7 billion when indirect costs are included. Those are not marketing numbers. They are part of the reason Part 139 inspections exist, the reason FAA AC 150/5210-24A exists, and the reason nearly every airport employee has at some point participated in one or more FOD walks.
That case sells itself.
The harder case, and the one I want to make here, is for everything the industry has historically treated as a footnote at best.
Five months. Eight airports. Two ledgers.
Over the last five months, Illuminex AI deployments of FODᴬᴵ at eight airports produced two very different piles of detections.
1,692 were classic FOD. Metal. Chunks of concrete and asphalt. Pieces of tire. The objects every inspector is trained to pick up.
1,418 were soft FOD.
That second number is almost as large as the first. It is also the number that historically has rarely been kept.
Soft FOD, in our taxonomy, is the material that sits on or in the pavement but is not treated as an immediate aircraft hazard: crack sealant proud of the surface, vegetation in joints and along edges, and a long tail of other items that range from dragonflies to earthworms. Individually, none of these is a strike risk. Collectively, they are a map of how the airfield is aging.
Historically, they have not been logged. An inspector driving a taxiway does not have the time to write up an earthworm, a bead of seal standing a quarter inch above a contraction joint, or a tuft of a weed peeking through a crack.
That is a missed asset.
One taxiway, 44 plants, a winter coming
Take one of our airports.
In a single operating period, we recorded 44 vegetation-based soft FOD logs. In isolation, each one is benign. A plant in a joint, nothing that would pose a safety risk.
Viewed on a map, in a time series, the picture changes.
The vast majority sit on the east side of a taxiway that runs north/south. The pattern is not random. It is failing joint seal, with vegetation intrusion, across roughly half of a major taxiway concentrating at joints and fillets. Vegetation does not appear in sound sealant. It appears where the seal has lost bond or gone missing, and where water and fines have already found a path into the reservoir.
This is classic joint seal damage, and any engineer trained in PCI scoring would know how to score it. But a human inspector sees one joint at a time, on one morning; perhaps a different inspector sees another in the afternoon, and has no practical way to accumulate 44 observations into a single picture. A vehicle-mounted system doing the same inspection route every day does.
Winter is the deadline for that picture. Water in an open joint through a freeze-thaw cycle is how crack seal turns into a slab replacement. The remediation at this stage is unglamorous: vegetation control and crack seal. The alternative is waiting until the same joints start producing the chunks of runway that fall on the FOD side of the ledger.
FOD does not spontaneously generate. It has a source. In our experience, the most common culprit is airfield surface material breaking off from defects. Soft FOD is often the defect announcing itself before the surface does.
There is a structural reason soft FOD is invisible.
Safety programs are built around the object that can damage an engine or a tire. Maintenance programs are built around PCI scores, usually months old, and around work orders that start after something has already failed. Soft FOD sits in the gap between those two systems. It is not dangerous enough for the technical inspectors to log, and so it disappears.
The value only appears when you do three things the daily inspection was never designed to do:
Detect it consistently, including the objects inspectors are trained to ignore.
Geo-reference every detection, so 44 weeds become a line on a taxiway instead of 44 forgotten and disparate observations.
Use time-series data mapping, so a pattern that develops over weeks is visible before winter, before the next construction season, before the next capital plan.
That is not a new sensor problem. It is a logging and aggregation problem. The inspection already happens. The vehicle is already on the pavement. What has been missing is a durable record of the things that were not considered worth writing down.
Two returns, one inspection
Classic FOD pays for itself in avoided strikes, avoided closures, and avoided delay. The Boeing and Insight SRI figures are the language operations and airlines already speak.
Soft FOD pays for itself on a different ledger: pavement life.
Airfield pavement is often the single largest infrastructure investment an airport will make, so a year of additional service life on a major movement area is not a rounding error. Proactive joint resealing and targeted vegetation control are inexpensive compared with slab replacement, but they are also the kind of work that only gets prioritized when someone can point to a map and say, with a date range and a count, that this side of this taxiway is where the seal is failing.
That is the unappreciated product of a FOD program that refuses to throw away the “non-hazard” detections. The 1,692 objects protect today’s operation, while the 1,418 objects protect the surface that operation runs on. The same vehicle, the same inspector, and the same inspection are producing value in two different ways.
The real magic of data is not necessarily a new detector. It is the second story you were already driving past.
Some observations will always be FOD and need to be removed. Others may not require anyone to do anything today. But when those observations are consistently collected, connected to location and time, and analyzed through InspectEx, they can become part of a much larger picture of the airfield. Patterns begin to emerge that are difficult to see from a single inspection or a single work order: where deterioration is recurring, where vegetation is encroaching, where joints are beginning to fail, and where small observations may be pointing toward a larger maintenance issue.
The real magic of data is not necessarily a new detector. It is the second story you were already driving past.
