The whole runway surface, at a resolution you can measure from

Frequently Asked Questions

What is the difference between the scan and the pavement condition assessment?

Both start from the same capture. The scan is the narrower service: the surface, or part of it, imaged at a ground resolution computed from the camera, with every image georeferenced and retained. The pavement condition assessment adds an agreed section map and is where crack detection with ASTM D5340 coding is scoped, which is in development and close to a first MVP. Neither produces a condition rating or a ranked repair list.

What resolution do you fly at?

The one the question needs. A hairline crack needs a finer ground resolution than a spall or a patch, and the resolution is set by the flight height and the camera, so we compute the height from the resolution you ask for rather than the other way round. Where you need finer, the scan is flown lower, which means more runs and more images, and the pilot scope says so before the flight.

How long does the aerodrome have to give us?

The capture window, and nothing else. The scan is flown between aircraft movements in coordination with the tower, and the drone clears the runway when asked. Budget about an hour of flying per kilometre of a 45 m wide runway, plus battery changes, with taxiways and aprons added in proportion to their area. The runs are laid out so that an interrupted run is reflown from where it stopped rather than restarted. Any analysis happens afterwards, off site.

How is coverage guaranteed?

By computing it before the flight instead of checking it after. From the flight height, the gimbal angle and the camera's field of view, the footprint of one image on the pavement is known. From that footprint and the overlap you asked for, the number of parallel runs across the surface width and the spacing between frames along each run follow, and the flight speed is checked against the camera's frame rate so that every planned frame is actually captured. The height follows the terrain, so the resolution holds where the surface falls away. The result is that the worst-covered strip of the surface is covered as well as the best.

What do you do about debris found on the surface?

Debris recognition is in development, short of training data, so the scan does not find debris for you yet. The imagery is retained, so a loose item or a spall can be seen in it by a person looking. If the crew sees loose material on a movement area during the flight, they pass it to the tower at the time. Debris is an operational hazard rather than a pavement defect, and the aerodrome will want it cleared before the next movement.

Can you scan only part of a surface?

Yes, and it is often the right choice. The along-track window can be capped or trimmed at either end and measured from either runway end, so re-checking the first six hundred metres off one threshold is one setting. The band across the surface can be narrowed to one side, which is how a shoulder or half of a wide apron gets scanned without flying the rest. Runs can be laid along the surface or across it. A repeat scan of a known problem area is a smaller flight, and it can reuse the settings of the last scan of the same area.

What do we get, in what format?

The imagery at the flown resolution, georeferenced, and a short report that says what was flown, at what resolution and over which area. The in-development crack output is a section and sample-unit inventory as GeoJSON plus a JSON and Markdown report. Retained imagery is kept so that a disputed finding can be re-examined against the frames rather than re-flown, and so that a survey can be re-run on a newer model.