Airbus

Airline Oleo Maintenance App

Client
Airbus
What we did
Architecture Research Feasibility UX Development AI Machine Learning Mobile App QRCode scanner
Visit their websites
www.airbus.com

Airline Oleo Maintenance App

Airbus wanted to explore the possibility of building a smart phone app utilising the camera as a measuring device to enable rapid turn around of maintenance activities, this was before Apple's AR kit or the measure app!

An Aircraft's landing gear has an Oleo shock absorber - this is essentially a cylinder full of gas between the wheels and the aircraft body. The gas compresses and acts as a shock absorber during landing.

Maintenance engineers need to periodically measure the height of the oleo to ensure the pressure is correct and determine whether additional gas needs to be added.

Currently the oleo is measured using a ruler and then the height is looked up on a spring curve graph printed on the landing gear itself.

The big idea was to explore if it was possible to build an app that would essentially enable engineers to point the phone at the oleo (the Aircraft's landing gear shock absorber) and give instant feedback. The app would then be able to automatically perform a measurement and calculate whether it was within the acceptable tolerance. No need to get in to the space with a ruler and manually lookup information

Feasibility

In order to tackle this challenge we broke the problem up into smaller feasibility mini projects. First there was no point continuing the project if the phones cameras at the time did not have the resolution to measure accurately.

  1. Camera Measurement Accuracy: Is it possible to achieve an accurate enough measurement using a modern smartphone camera?
  2. Edge & Object Detection: Is it possible to easily identify the object to be measured?
  3. Can this be done within a reasonable time frame so its easier than the current method?

Camera Measurement Accuracy

Essentially if there was a known length in the image we could use this as a datum on which to determine the measurement. We needed to prove that there was enough resolution in the image to get an accurate enough measurement. Therefore we conducted an initial feasibility study taking into consideration the inner workings of the camera, a variable position of the user, the angle of the phone, the resolution and automatic scaling applied to the image. At what point would the measurement fail? Would the user have to put the phone right up next to the oleo? Or could they remain a comfortable distance away? What happens if the phone wobbles or the user holds the phone at an angle?

We wrote up our findings and explained the logic and calculations behind them. Essentially, yes it was feasible to get an accurate enough measurement within a comfortable distance.

We are not huge fans of large documents but we managed to cram in the essentials in 16 pages. With enough evidence to support the next phase.

☝️ A report of our findings

Edge & Object Detection Using OpenCV

So it was theoretically possible 👏.

We had proved mathematically that it was possible to measure items using the camera on iPhone 7 from a reasonable user friendly distance.

The next challenge was to determine if a smart phone had enough power to run visual recognition and select the object to be measured in real time.

We launched another mini project of only a few days (no point spending months if this part fails!) to prove if we could get some basic edge detection working.

Demo of initial iPhone edge detection

The results were very promising so we went ahead with the full project to build the app.

Testimonials

Listen to what our customers say about us

Start your future, today

Send us a message for more information about how we can help you:

We use cookies, review our privacy policy here.