NATO and Allied Airfield Pavement Condition by Drone

Who runs into this

Allied nations agreed the method long ago. STANAG 7181 , published as AEP-56, is the NATO standard method for airfield Pavement Condition Index surveys. UFM 3-260-16 points to it for NATO operations. On paper there is one way to rate an allied runway.

In practice each nation walks its own sample units on its own cycle. The result is a picture of the allied network that is incomplete, because only a sample of each pavement was rated, and not comparable, because the cycles, the raters and the sampling differ. A walked survey also closes the runway while the raters work, and it needs trained raters on site, which is hard to arrange at a deployed or dispersed operating location.

It shows up in four questions.

• Can every sample unit of a main operating base be rated, and not only the units a sampling rule picked?
• Can two nations’ airfields be compared, knowing that the same rule set rated both?
• How long does the survey close the runway, and how few people need to be on the pavement?
• Where does the data go, and can it stay with the nation that owns the airfield?

The pavement condition assessment is the service behind this page. It is in development and offered as early access. This page sets out what the software does today, what it does not do, and where it is going.

Where this applies, and where it does not

This applies to paved runways, taxiways and aprons of asphalt or jointed concrete, at main operating bases, at deployed and dispersed operating locations, and at civil airports in host-nation use. It works the same on a familiar airfield and on one the team has never seen, on any asphalt or concrete surface not covered by snow, sand or water.

It follows the same PCI method as STANAG 7181 and the ASTM D5340 definitions. It sees nothing below the surface, so it gives no structural capacity and no friction. And it is not a substitute where a requirement names a walked survey signed by a certified inspector, until a side-by-side comparison with such an inspector exists.

Every sample unit rated, one rule set

The drone images the pavement at 1 to 5 mm per pixel. The software then finds, measures and rates every distress by the D5340 definitions and computes the PCI of every sample unit and of each section. Cracks are rated automatically with a severity taken from the measured width. On asphalt that covers alligator, block, longitudinal and transverse cracking. On concrete it covers linear cracking and joint and corner spalling. A classifier labels each tile with the other D5340 types and already recognises all 33 of them, and the report lists which types were rated and which were not, because an unrated type biases the PCI upward.

The ratings are the same whoever flies the survey. A national team, a host-nation team or a partner operator gives the same pavement the same figure, because the figure comes from the imagery and one rule set and not from the judgement of the rater of the day. That is what lets two nations compare their airfields.

Checked against the worked examples in UFM 3-260-16, the PCI of both the asphalt and the concrete example lands within 1 point. A section surveyed again on the same runway repeats within 1 point.

Review on a concrete runway at 1 mm per pixel. The analyst sees a hairline crack next to the 22 detections that look like it and marks them in one save.

Minutes of runway time, not a closure

Capture is the only step on the airfield. For 1 km of a 45 m wide runway, one aircraft needs 4 minutes to find cracks from 10 mm, 13 minutes to find cracks from 5 mm and 45 minutes for the full survey at 1 mm per pixel, with cracks from 2 mm. More aircraft cut these times. The aircraft launches from the shoulder and nothing touches the pavement, so the runway reopens as soon as the drone lands. A walked survey keeps the runway closed for as long as the raters walk it.

Processing on site, data with the nation

About 20 minutes pass from the end of capture to the PCI report for 1 km of runway at 5 mm per pixel. On a MacBook processing takes about 30 minutes per 100 m of runway at 1 mm per pixel and about 2 minutes at 5 mm. The software runs on a MacBook or on a PC with a graphics card, on site, and needs no cloud. Imagery and results stay on the machine the nation provides. The orthomosaic is built from the drone positions, 6 times faster than photogrammetry, with a PCI within 1 point of the photogrammetric result.

Analyst review

The report never waits for review. Each detection is shown next to the cracks that look like it, and the analyst marks a sealed crack, joint, vegetation, paint or stain in one save and sees the PCI of the unit change at once. Sealant and paint edges read as cracks, and clearing them takes minutes. Every verdict becomes a training label for the next version of the models.

What a national GIS receives

Every distress keeps its map position, its length in metres and its width in millimetres, and on concrete the distress sits on the slab grid. A repeat survey therefore shows how each crack grows. The inventory is a GeoJSON file, so any national GIS opens it. The PDF condition sheet for each sample unit shows the deduct calculation line by line, which is the document an engineer checks the PCI against.

Why walked surveys do not add up across nations

A walked survey rates the units a sampling rule selected and infers the section from them. That is a sound statistical method, and it is the reason a PCI of a large pavement was practical to walk at all. The weakness is in what happens around it. The cycle differs from nation to nation. The raters differ. The units picked differ. When two PCI values from two airfields are put in one table, the table hides all of this.

Rating every sample unit removes the sampling error from the capture. Using one rule set removes the difference between raters. What remains is the pavement itself, and that is the quantity a network planner wants to compare.

Operating from a host-nation airfield

An allied survey often happens on ground that belongs to someone else. The data rules are those of the owning nation, the airspace belongs to its air traffic control, and the operator may have to be a vetted one.

TarmacView j. s. a., Bratislava, holds NATO NCAGE code 0181M, the identifier defence buyers use to find a supplier. Our partner operators keep a fleet of Blue UAS cleared aircraft, the aircraft the US Department of Defense vets for cybersecurity and supply-chain requirements. TarmacView itself is not on that list. The mission planner is drone agnostic, so a nation that flies its own aircraft, DJI, MAVLink on PX4 or ArduPilot, or UgCS, uses the same planner and the same processing. The data stays with the nation because the processing runs on its own laptop.

Where it is going

Everything above runs on our own flights today. The direction, with no dates, is this. Every distress type rated automatically with severity. Debris flagged in the same pass. Results loading into PAVER and into GIS. Crack growth forecasts. Crack depth from angled photos. Drone docks at the base. Processing on the customer’s own secure cloud.

For a damaged airfield the direction is crater and spall depth and volume, measured from the drone imagery, to support a rapid damage assessment after an attack. This is a direction and not a feature. Airfield damage repair has its own NATO standard, STANAG 2929 and AATMP-03, and a damage assessment would feed that process. Today’s survey stops before that process.

Evidence and the honest limits

The strong claims can be checked in a pilot. Every sample unit is rated. The runway time is the flight. The PCI lands within 1 point of the UFM 3-260-16 worked examples, and the processing runs on a laptop with no connection. The PCI explainer sets out how a PCI is built, and the ASTM D5340 standard page sets out what the method requires.

Then the limits. Wind, temperature and rain limits are the aircraft maker’s. Flying is by daylight and preferably over a dry surface. Sealant and paint edges read as cracks and are cleared in review in minutes. The survey sees nothing below the surface, so it gives no structural capacity and no friction. And no certified PCI inspector has yet rated our sections side by side with the software. The way to test it on an allied airfield is to walk a few sections in the same season as a drone survey and compare unit by unit.

Objections worth raising before a pilot

“The standard is the walked survey.” Keep it where a requirement names it. The drone survey rates the whole surface in minutes of runway time and gives the walked team a map of where to look.

“Our data cannot leave the base.” It does not. Processing runs on a laptop on site with no cloud, and the files stay on the nation’s machine.

“We cannot accept a supplier we do not know.” TarmacView j. s. a. holds NATO NCAGE code 0181M, and the aircraft are operated by partners that keep a fleet of Blue UAS cleared aircraft. Neither is a NATO endorsement.

“Will two nations’ figures really compare?” The same rule set rates both, and the comparison is checkable on a few sections against walked ratings.

Related tasks

If the buyer is a national airport or road authority, government airport pavement condition takes that angle. If the question is the US military installation and UFM 3-260-16, US military airfield pavement PCI covers it. The service itself is described on the pavement condition assessment page. To talk about early access on your own sections, contact TarmacView .