Aerospace and Electronic Systems - May 2019 - 30

Feature Article:

DOI. No. 10.1109/MAES.2019.2900904

A Framework to Predict Failures for Ground Tests on
Aircrafts
~
Inigo
Monedero, Felix Biscarri, Diego Francisco Larios, Julio Barbancho,
Universidad de Sevilla, Seville, Spain

INTRODUCTION
Ground testing is an essential and expensive process for
airbus defense and space (airbus ds) [1] in the manufacturing cycle of an aircraft. all parts of the aircraft, as well as
the systems, must be meticulously checked in this type of
test, and the testing involves a large number of company
staff and resources.
To carry out the test, Airbus DS uses a test system
consisting of a software application linked to a hardware
system for interconnection and communication with the
aircraft. It is not possible to give many details of this test
system due to the confidentiality required by the Airbus
DS Company. However, this is not necessary for the
description of the developed framework.
This work was carried out as part of the project
"FSP20: Futuro Sistema de Pruebas, Vision 2020," funded
by CDTI (Centre for Industrial Technological Development) under the FEDER-INNTERCONECTA program.
When the testing phase is carried out for an aircraft, it
is common for failures to occur throughout the different
tests. These failures are known in the terminology of the
Airbus DS Company as "incidences." Thus, these incidences in the test process necessitate delays in the deliveries
of the aircraft, as well as additional costs for the company.
The delays and added costs arise because those tests that
generate incidences must be repeated (totally or partially),
which generates an increase in the number of hours for the
workers and in the time for carrying out the test.
This work addresses the incidences that have arisen in
the testing process on the A400M aircraft of the Airbus

Authors' current address: I. Monedero, F. Biscarri,
D. F. Larios, J. Barbancho, Escuela Polite´cnica Superior,
Departamento de Tecnologı´a Electro´nica, Universidad de
Sevilla, Seville 41011, Spain. E-mail: (imonedero,
fbiscarri, dlarios, jbarbancho@us.es).
Manuscript received November 27, 2018, revised
February 5, 2019, and ready for publication April 18, 2019.
Review handled by R. Sabatini
0885-8985-19/$26.00 ß 2019 IEEE
30

DS Company. The A400M aircraft is currently the most
important and largest military aircraft as well as a flagship
European collaborative procurement project. Thus, we
have found recent works about this aircraft in the research
literature [2]-[6], but none focused on the topic of our
work. Specifically, this paper describes a framework consisting of a software application based on artificial intelligence for the prediction of incidences for ground tests on
the A400M aircraft.
The framework is capable of predicting the probability that a certain test that is going to be performed on
the aircraft will generate an incidence. This prediction
allows the Airbus DS workers to focus on that prediction at the time of carrying out the test. In addition, and
this is even more interesting for the company, the
framework suggests actions to be carried out so that the
incidence does not occur either in the current aircraft or
in successive versions.
The framework described in this paper is framed
within the project: "FSP20: Futuro Sistema de Pruebas,
Visi
on 2020" in which our working group from the
Electronic Technology Department at the University of
Seville (Spain) worked for the Airbus DS Company.
The FSP20 project was a collaboration project among
Airbus DS and various companies and research organizations. The project was developed in Seville, since it
is the city where the process of assembly and testing of
the A400M is carried out.
For the development of the kernel of the framework,
a data mining process was carried out from the database
and the data registered for the test system that the Airbus
DS Company uses in the ground tests of the A400M. If we
review the research literature, we can see that although
there are some research papers in incidence prediction in
the context of aeronautics using data mining techniques
[7]-[11], it is not possible to find published research
specifically applied to ground testing on aircraft.
The developed kernel is composed of a set of models
generated from this data mining process. Additionally, an
application for Windows that integrates the kernel and

IEEE A&E SYSTEMS MAGAZINE

MAY 2019

Aerospace and Electronic Systems - May 2019

Table of Contents for the Digital Edition of Aerospace and Electronic Systems - May 2019