Plane Reliability

Boeing 2018 • Web App • 0 → 1

Role

  • Main designer; lead foundational research, wireframing, prototyping, and specs for development.
  • Part-time design collaboration with Jessie C. (UX designer).
  • Product shaping with John T. (product manager) and John D. (product owner).

Problems to solve

  • Through our foundational research we saw that reliability program maturity varied wildly between airlines even if the overall system looked very similar.
  • One of the main problems faced by reliability engineers was managing and searching though the high volumes of data about the airplane.
  • Another big problem was about properly encoding semi-structured data like comments where much of the contexual meat is hidden.

💁 A reliability program scales from the immediate to the long term — from daily to the 20 years of a plane’s life. The lynchpin is understanding how data becomes living objects reflecting component and system lifecycles. A reliability engineer is responsible for understanding long term trends and long term reliability strategies.

Solutions

  • An easy to use web app that brings together all the imporant data necessary for a reliability engineer to identify long term trends in their fleets while being able to access detail down to a single event. 
  • We also supported the fast creation of regulatory reports which were time consuming and legally required but provided little value to airlines themselves.
  • A NLP model was implemented that codes semi-structured data from comment fields so that analysis was much quicker. A companion app was created that helped engineers take part in the coding process and improve it.
  • A starting model for alerting engineers to emerging trends was implemented.


Impact

  • 4 International Launch Customers: Virgin Atlantic Airlines, New Zealand Air, Japan Airlines, and SouthWest Airlines.
  • 80% reduction in data query times.
  • 30% reduction in basic analysis times.