Predictive Airplane Maintenance


Overview

Roles
User Researcher and Designer

Date
January 2018

Summary
Researched and designed a system to predict airplane maintenance issues in real time with artificial intelligence, directing mechanics to possible problem areas, and providing proper instruction documentation. This system speeds up ground repairs to ultimately keep airplanes on time and in the air.


Problem

Plane Pains
The more an airplane flies the higher likely hood it will be at risk for mechanical failure. This is especially true if a plane is subjected to unfavorable weather conditions which puts the plane’s components under more stress and wears them out quicker.

Airlines run on very tight schedules and discovering a part that needs replacement is very time consuming. Airplanes are huge vessels and have thousands of intricate parts which makes it impossible to check every little one before every flight.

If a part is in fact broken the entire plane has to be grounded until the issue is found and fixed. If a part has not broken yet but is on the verge of failure, the only way its condition can be determined is in a painstaking visual check which is not feasible between every flight.

Even when proper diagnoses of failed parts occurs, there is no guarantee the part will be on hand when needed or that the plane will be able to be repaired in time without delaying or cancelling all subsequent flights.

All of these difficult repair scenarios involve time consuming repairs that are difficult to diagnose and keep airplanes on the ground for extended periods of time which spells higher costs for the airlines, delayed/cancelled flights and of course unhappy passengers.

Predicting the Future
What if there was a way to know when airplane component failures were going to occur before they happened?

While it would be nice to be able to actually see into the future, I set out to do the next best thing by designing a system that will make predictive airplane maintenance before it is necessary.

To predict issues before they happen the system will utilizes AI, networked sensors on individual components (IOT), weather data, engage the supply chain earlier and enhance communications between all parties involved. With this system the airlines will not only spot issues before they turn into problems and save money by keeping their planes off the ground due to avoidable maintenance but ultimately they will make their passengers happier by keeping them on-time, flying to their next destinations.

GOALS

  • Utilize real time data
  • Avoid downtime
  • Reduce costs
  • Maximize equipment up-time
  • Enhance supply chain
  • Increase safety

NO DELAYS OR CANCELLATIONS = HAPPY CUSTOMERS


Research

Interviews
In order to better understand the perspectives of the people at the scene of the repairs, I interviewed my Uncle who has been a pilot for over 35 years and my friend who just recently graduated from avionics school.

Private Airline Pilot

“The crew conducts a mandatory visual inspection before every flight.”

“Flying a plane is kind of like driving a car, you can feel when things are going wrong.” (referring to parts failing)

Pilots play a big part in relaying information to the mechanics, as they can feel when things are going wrong in the plane and have to check off a comprehensive list required by the FAA before every flight. Bar digital and analog sensors that check the parts themselves pilots have the most insight into any issues arising on an aircraft.

Avionics Tech

“Parts are known to give out at certain predetermined flight hours.”

“The paper documentation for a single aircraft would fill up a large room.”

Mechanics know that all parts have a certain threshold of flight hours before they are prone to breaking and have to be replaced. Logging and keeping track of these numbers, especially when they are put under heavier stresses caused by weather, would be crucial to knowing when a part might need replacement.

The paper documentation on all of the repairs for an aircraft is incredibly vast and cumbersome. Being able to digitize these manuals would be critical to saving time and making repairs go smoothly.

Understanding the Users

Empathy Map and Persona


Solution

Storyboard

1. Mike eagerly starts his day by drinking a cup of coffee and checking his schedule.

2. Mike sees on his schedule that a plane will be arriving at 8:05 AM at gate A5 and the system suggests when he needs to leave in order to arrive on time.

3. Checking the plane’s diagnostic information, the system suggests there might be an issue occurring in the engine on the right wing.

4. After the plane lands, Mike checks the problem location and finds an issue at the suggested location and confirms the problem with the system.

 

5. Mike confirms and submits a work order on the broken part and which automatically brings up a digital manual on how to fix the part. Mike is relieved because he has not fixed on of these parts in quite some time and could use the refresher. While all of this occurs the backend system simultaneously sends for the new part.

6. New part is taken from warehouse and is delivered to Mike just as he removes the broken part, which is then discarded.

7. Mike finishes the repair and makes the appropriate required notes, documenting the process. Once the report is submitted, the flight is cleared for take off and continues on its way, right on time.

8. The passengers load onto the plane happily on their way, never even knowing their plane was in jeopardy of being delayed.

Journey Map

A lot of unseen technical ‘wizardry’ went on in the background of the previous storyboard scenario that ensured everything went smoothly for Mike.

To be able to see everything happening throughout the repair, a Journey Map helps highlight and analyze the different complex stages as they unfold.

Click to enlarge

Wireframes

Prototype


Previous project

Next project