1. Big Data in Airbus Flight Test and Integration Center
OOW AIRBUS
26th of october 2015
ORACLE open world 2015
Laurent PELTIERS Jean-Marc WATTECANT
Big Data in Airbus Flight Test and Integration Center

2. Some figures about AIRBUS
OOW AIRBUS
26th of october 2015
Some figures about AIRBUS
74000 employees over mainly 4 countries (more than 1000 in the US)
Revenues (2013) € 42,000 million / EBIT € 1,710 million
8121 aircrafts in service operated by more than 400 companies
Largest civil aircraft: the double deck A380
629 aircrafts delivered in 2014 2’
Ce qui est important pour les EV:
+ en + transnational
3 FAL
Nombre d’A/C livrés

3. Flight & Integration test center : our place in A/C development
OOW AIRBUS
26th of october 2015
Flight & Integration test center : our place in A/C development
A/C certified!
MG3
MG13
MG7
MG9
MG11
MG4.1
Entry into concept
Entry into service
MG5
MG6
Auth. to Offer
Entry into Definition
End of Concept
Entry into Production
Freeze of Definition
Entry into FAL
First Flight
Data for Manufacturing
Integration Tests
Flight Tests
Production
Assembly line
Ramp Up
Concept
Definition
6 Year Development lead time
~24 months
MG4.2
Test
Phase 3’
Incrémental développement + time to market => cycles de plus en plus rapides
2 parties historiquement dissociées sol / vol => intégration de plus en plus forte 3

4. Flight & Integration test center: our mission
OOW AIRBUS
26th of october 2015
Flight & Integration test center: our mission
Design office & program domains
Tests definition
Aircraft certification
Flight &Integration test center
Tests preparation
Tests reports
Ground or flight tests
Tests analysis 2’
Rôle des EV
Certification
Maturité à l’EIS
Amélioration continue des modèles

5. Evolution of data collected
OOW AIRBUS
26th of october 2015
Evolution of data collected
Parameters #
x50
parameters
450 TB archived
parameters
150 TB archived
parameters
Data archived
12,8 TB archived
x50
parameters
8,5 TB archived
AMPEX 28 tracks 8 GB
SONY AIT 2/3
50/100 GB
SCSI hard disk 300 GB
SSD hard disk 700 GB
Flight and integration center: the context
From the 80’s
Increase of the parameters collected
Sample rate of analog sensors is up to 48khz and 400khz for specific usage
AFDX buses from the A380
Switch from PCM to IENA
At the same time the recorders changed from the 28 tracks tapes to the last generation of RAID SDD hard disk
But the way of working is still the same
1987
1992
2005
2013

6. Why Big data ? - Volume Forecast
OOW AIRBUS
26th of october 2015
Why Big data ? - Volume Forecast
A lot of restore up to 10 years after
Concurrent access
High volume management
Before the A350 campaign we estimate the forecast of the data collected in the 4 next years
The beginning of the NEO campaign was before the ramp-down of the A350 with up to 7 aircrafts involved . There was a risk to slow down the analysis
An other point is about the restore of old data to compare the legacy aicrafts and the new version

7. Big data project: One year project
OOW AIRBUS
26th of october 2015
Big data project: One year project
RFI/RFP: Q3 2013
Start:feb 2014
Validation: october 2014
EIS: january 2015
Planning ambitieux pour une entreprise de cette taille avec une initiative bottom-up

8. Big data project: the scope of the first step
OOW AIRBUS
26th of october 2015
Big data project: the scope of the first step
Acquire
Raw data
Organize
Current tools compatibility Data triggering
Analyse
Machine learning…
Decide
Discovery tools
No functional impact for the user
Enhancements :
Less overhead to get the data
Full campaign on-line

9. Large time series Injection by burst Main challenges ?
OOW AIRBUS
26th of october 2015
Main challenges ?
Big data appliance
noSQL database
HADOOP cluster
Large time series
Up to 4 flights 2 times a day
Injection by burst

10. 4 aircrafts 70000 parameters / aircrafts 700h of flights
OOW AIRBUS
26th of october 2015
Current status
4 aircrafts parameters / aircrafts 700h of flights
Big Data Appliance X4-2
6 nodes noSQL
6 nodes HADOOP
5 billions of events
15% of the disk capacity
50% of the NEO processing use the appliance
Time saving for multiflights analysis

11. Next steps: processing
OOW AIRBUS
26th of october 2015
Next steps: processing
Descriptive analysis:
Complex event detection
Wide correlation between parameters
Predictive analysis:
Sensor failure
Anomalies database

12. Next steps: Vizualisation
OOW AIRBUS
26th of october 2015
Next steps: Vizualisation

13. AIRBUS Flight and Integration Test Centre : our data processing view
OOW AIRBUS
26th of october 2015
AIRBUS Flight and Integration Test Centre : our data processing view
Processing HMI
Data mining tool
Complex event detection Time correlation for diag Machine learning algo
Full campaign on-line
Results storage Flight Tests data lake

14. Big Data application : Improve « Data retrieval »
OOW AIRBUS
26th of october 2015
Big Data application : Improve « Data retrieval »
Save flight hours
Easy search will be generalized
« search  autopilot on & altitude > ft & Mn>0.8 & bank > 5° then and plot the vertical load factor »
Opportunities to bridge with other database (configuration, logs,…)
Compare versions in the frame of incremental development
If can be specific flight , ferry flight : not all of this have been gathered in DB
It may prevent from reflying if conditions are met (5% tolernace on weight increase for example)
A320 :1988
Sharklet 2012
NEO 2015

15. Big Data application : Perform « Data re-use »
OOW AIRBUS
26th of october 2015
Big Data application : Perform « Data re-use »
Increase system design maturity
use the data for design verification prior implementation in avionic software (robustness)
Continuous detection of anomalies on the whole campaign
Tuning and re-launch surveillance algorithms
Patterns recognition out of the « usual » enveloppe

16. Big Data application : Accelerate « Test Analysis »
OOW AIRBUS
26th of october 2015
Big Data application : Accelerate « Test Analysis »
Enhance the analysis through statistical tools
Correlation analysis:
eg : longitudinal oscillation A380 specific flight needed to understand the root cause
Mn=0.85->0.8
Mn=0.85
No stimulation
stimulation
« Self » learning and clustering capability
Attention correlation ‘n’est pas causalité mais cela peut aider à identifer un phénomène

17. OOW2015 - AIRBUS 26th of october 2015
© AIRBUS Operations S.A.S. All rights reserved. Confidential and proprietary document. This document and all information contained herein is the sole property of AIRBUS Operations S.A.S. No intellectual property rights are granted by the delivery of this document or the disclosure of its content. This document shall not be reproduced or disclosed to a third party without the express written consent of AIRBUS Operations S.A.S. This document and its content shall not be used for any purpose other than that for which it is supplied. The statements made herein do not constitute an offer. They are based on the mentioned assumptions and are expressed in good faith. Where the supporting grounds for these statements are not shown, AIRBUS Operations S.A.S will be pleased to explain the basis thereof. AIRBUS, its logo, A300, A310, A318, A319, A320, A321, A330, A340, A350, A380, A400M are registered trademarks.