2025 SMSI Bannerklein

6.4 Machine Learning Assistant for Aircraft Acceptance

Event
ettc2020 - European Test and Telemetry Conference
2020-06-23 - 2020-06-25
Virtual Conference
Chapter
6. Big Data
Author(s)
R. López Parra - Airbus Defence & Space, Madrid-Getafe (Spain)
Pages
185 - 190
DOI
10.5162/ettc2020/6.4
ISBN
978-3-9819376-3-3
Price
free

Abstract

The task of testing an aircraft to be considered “ready for delivery” is a task that implies a lot of work, involvement of pilots, flight engineers and it demands significant costs. Optimize the time of ac-ceptance flights is critical to save a lot of time and money. Currently the flight test engineers are re-sponsible for supervising and planning all the maneuvers that are required to verify that all systems work properly. The objective is to create a machine learning system which will analyze all the actions carried out in the aircraft during acceptance flights, and it will learn when to perform them, taking into account the maximum information available.This machine learning system must be able to perform the task cur-rently performed by the flight engineer and lead the pilots in the maneuvers to be performed, also checking if they have been performed correctly. Moreover, controlling thousands of variables that exist in an aircraft is only suitable for an artificial intelligence entity, which might detect possible problems and warn them with enough time to avoid major problems.This system could become an intelligent assistant essential in any environment.

Download