Available online at www.sciencedirect.

com
Available online at www.sciencedirect.com

ScienceDirect 
ScienceDirect 
Available
Availableonline
Procedia atatwww.sciencedirect.com
Manufacturing
online 00 (2017) 000–000
www.sciencedirect.com
Procedia Manufacturing 00 (2017) 000–000 www.elsevier.com/locate/procedia
www.elsevier.com/locate/procedia
ScienceDirect
ScienceDirect 
Procedia Manufacturing 13 (2017) 1167–1174
Procedia Manufacturing 00 (2017) 000–000
www.elsevier.com/locate/procedia

Manufacturing Engineering Society International Conference 2017, MESIC 2017, 28-30

Manufacturing Engineering Society
June 2017, VigoInternational Conference
(Pontevedra), Spain 2017, MESIC 2017, 28-30
June 2017, Vigo (Pontevedra), Spain
Scheduling of an aeronautical final assembly line: a case study
Scheduling
Manufacturing of an aeronautical
Engineering final
Society International assembly
Conference 2017,line: a case
MESIC studyJune
2017, 28-30
2017, Vigo (Pontevedra),
T. Borreguero-Sanchidrian Spain
a,b
T. Borreguero-Sanchidriana,b
Costing models for capacity optimization in
SpainIndustry
Airbus. Paseo John Lennon s/n. Getafe, Spain
4.0: Trade-off
a

Airbus.
UPM., Paseo John
José Gutiérrez Lennon2,s/n.
Abascal,
b a
Getafe,
Madrid 28006, Spain

between used capacity and operational efficiency

UPM., José Gutiérrez Abascal, 2, Madrid 28006, Spain
b

Abstract
Abstract A. Santanaa, P. Afonsoa,*, A. Zaninb, R. Wernkeb
Within the aeronautical industry, there has been in the recent years a high effort towards digitalization. This has resulted in a deep
Within the aeronautical
transformation industry,
of the way there all
of working has
a been in the
University
through the recent4800-058
of Minho,
productyears a high effort
Guimarães,
lifecycle. This towards
Portugal digitalization.
digitization has included This has of
the use resulted in a deep
scheduling and
transformation of the way of working all through
b
the
Unochapecó,product lifecycle.
89809-000 This
Chapecó, digitization
SC, Brazil has included the
line balancing support tools, which have been extremely important in the case of the industrialization of new products and use of scheduling and
line balancing
existing productssupport tools, which have
re-industrialization. In this been
caseextremely
study we important
present theinre-industrialization
the case of the industrialization of new
of a final assembly line products and
from the line
existing
balancingproducts
point ofre-industrialization.
view. It completesIn other
this case study studies
previous we present
that the re-industrialization
presented of aframework
the theoretical final assembly line from thefinal
for aeronautical line
balancing point
assembly line of view. It completes other previous studies that presented the theoretical framework for aeronautical final
balancing.
Abstract
assembly lineAuthors.
balancing.
© 2017 The Published by Elsevier B.V.
© 2017 The Authors.
Peer-review under Published by
responsibility Elsevier
ofElsevier B.V. committee of the Manufacturing Engineering Society International Conference
the scientific
© 2017
Under The
Peer-review Authors.
the concept Published
under responsibility by
of "Industry
of the4.0", B.V.
production
scientific committee processes will be pushed
of the Manufacturing to beSociety
Engineering increasingly interconnected,
International Conference
2017.
Peer-review under responsibility of the scientific committee of the Manufacturing Engineering Society International Conference 2017.
information
2017. based on a real time basis and, necessarily, much more efficient. In this context, capacity optimization
goes beyond
Keywords: Casethe traditional
Study; aimIndustry;
Aeronautical of capacity maximization, contributing also for organization’s profitability and value.
Line Balancing
Keywords:lean
Indeed, Case Study; Aeronautical
management andIndustry; Line Balancing
continuous improvement approaches suggest capacity optimization instead of
maximization. The study of capacity optimization and costing models is an important research topic that deserves
1. Introduction
contributions from both the practical and theoretical perspectives. This paper presents and discusses a mathematical
1. Introduction
model for capacity management based on different costing models (ABC and TDABC). A generic model has been
Smart factories
developed and it wasare usedone of the idle
to analyze maincapacity
characteristics of industry
and to design strategies4.0. This the
towards implies a furtherofintegration
maximization and
organization’s
Smart
value. Thefactories
digitalization of all are
trade-off the one ofmaximization
business
capacity the main characteristics
processes, including ofefficiency
industry As
manufacturing.
vs operational 4.0.
is This implies
a result, and ita is
further
digitalization
highlighted integration
is becoming
shown oneand
of
that capacity
digitalization
the main topics
optimization ofinallevery
might the operational
hide business processes,
industry [1]. including manufacturing. As a result, digitalization is becoming one of
inefficiency.
©the2017
main
Thetopics in every
Authors. industry
Published [1]. B.V.
by Elsevier
Peer-review under responsibility of the scientific committee of the Manufacturing Engineering Society International Conference
2017.

Keywords: Cost Models; ABC; TDABC; Capacity Management; Idle Capacity; Operational Efficiency

* T. Borreguero-Sanchidrian. Tel.: +0-000-000-0000 ; fax: +0-000-000-0000 .

1.T.Introduction
* E-mail
Borreguero-Sanchidrian. Tel.: +0-000-000-0000Tamara.borreguero@upm.es
address: Tamara.borreguero@airbus.com, ; fax: +0-000-000-0000 .
E-mail address: Tamara.borreguero@airbus.com, Tamara.borreguero@upm.es
The cost of idle capacity is a fundamental information for companies and their management of extreme importance
in
2351-9789 ©production
modern systems.
2017 The Authors. In general,
Published it isB.V.
by Elsevier defined as unused capacity or production potential and can be measured
in several©under
2351-9789
Peer-review ways: tons of production,
2017responsibility
The Authors. Published
of available
by Elsevier
the scientific B.V.hours
committee of manufacturing,
of the Manufacturing etc.Society
Engineering The International
management of the 2017.
Conference idle capacity
Peer-review underTel.:
* Paulo Afonso. responsibility
+351 253 of the761;
510 scientific committee
fax: +351 253 604of741
the Manufacturing Engineering Society International Conference 2017.
E-mail address: psafonso@dps.uminho.pt

2351-9789 © 2017 The Authors. Published by Elsevier B.V.

Peer-review under responsibility of the scientific committee of the Manufacturing Engineering Society International Conference 2017.
2351-9789 © 2017 The Authors. Published by Elsevier B.V.
Peer-review under responsibility of the scientific committee of the Manufacturing Engineering Society International Conference 2017.
10.1016/j.promfg.2017.09.179
1168 T. Borreguero-Sanchidrian / Procedia Manufacturing 13 (2017) 1167–1174
2 T. Borreguero-Sanchidrian / Procedia Manufacturing 00 (2017) 000–000

In [2] Mas et al studied the influence of PLM (Product Lifecycle Management) tools in the two main aircraft
manufacturers. They stated that the use of PLM tools had been a strong enabler for shortening development and
industrialization periods. In the same way, the use of different digital solutions at the shop floor has provided an
increase of competitiveness, improving the processes performance. Some examples are the use of virtual Reality or
Laser projection [3], [4]. Within all these tools, scheduling and line balancing have been left aside for long.
Nevertheless, they are of great importance in the production context.
To the best of our knowledge, there are little examples on real use cases of line balancing and scheduling within
the aeronautical industry. In 1994, Scott [5] presented a work on the main issues that need to be included when
modelling an aeronautic assembly line, such as human resource availability, space constraints and precedencies
between jobs. A more recent study, published by Heike [6] different alternatives for aeronautical assembly lines with
regards to the use of constant or variable cycle times on flow shops. Ziarnetzky [7] use simulation for a similar
study.
On top of this, in [8] Menéndez et al presented a first approach to a methodology for the first industrialization of a
final assembly line in terms of line balancing and task scheduling. It addressed the balancing line activities to be
completed during a program development phase. As a result, a method was established in order to help the
industrialization engineers evaluate the impact of the design decisions (including product design, but also processes,
jigs and tools) on the assembly line performance during the whole program lifecycle. In 2015, Borreguero [9]
extended this approach to the production phase.
This methodology has been applied to a final assembly line during its re-industrialization in order to reach the
program ramp up. Until the moment of the study, the organization had been focused on the delivery of the first
prototypes. Facing the ramp up meant that a constant delivery rate must be assured, together with on quality and on
cost manufacturing.
The aim of this case study is to provide an overview of the main methodology and outcomes of this re-
industrialization.
The remainder of this paper is structured as follows: Section 2 provides a more detailed description of the case
study context. Section 3 is dedicated to the different steps of the re-industrialization. Finally, Section 4 is dedicated
to the conclusions.

2. Case study context

This case study has been developed in the final assembly line of Multi Role Transport Tanker Aircraft (MRTT).
It is a military aircraft used for in-flight refueling as well as military transport.
It is built using a finished Airbus A330 civil aircraft as a basis. The modifications to be done on the aircraft for its
transformation include most of the technologies present in other final assembly lines. In all, MRTT production is a
complex process that involves 2000 work orders, more than 9000 tasks and more than twelve months’ lead time at
the moment of the study.
Before the re-industrialization, production was made using the hangars in parallel. Each serial number was
converted in one hangar from its arrival until the delivery to the client with the same dedicated team. Due to the
production rate, there were at most three aircrafts being produced in parallel. This had been a strong asset during
prototype development, as the team working on a prototype followed its development from the product definition to
the final aircraft delivery.
However, as the industrialization was consolidated, some drawbacks of this organization where identified. To
begin with, the three parallel lines had a negative impact in knowledge sharing between different serial numbers.
Also, due to the long lead time (several months), there was a small impact of training curves, as the same team
rarely repeated the same work and, if they did, it had been long since their previous execution. Also, there was a
need to improve intermediate quality gates so as to have a clear vision of the aircraft status and work in progress.
This is why, in the beginning of the ramp up, it was decided to turn the production flow into a pulse line. The
benefits of pulse lines have been widely discussed. For Boeing [10] it is one of their nine tactics for Lean
Manufacturing implementation. They reported that the implementation of a moving line resulted on a reduction of
the total process lead time in a 16% for the 717 and a 50% for the 777 [11].
Some of the advantages for this re-industrialization were:
 T. Borreguero-Sanchidrian / Procedia Manufacturing 13 (2017) 1167–1174 1169
T. Borreguero-Sanchidrian / Procedia Manufacturing 00 (2017) 000–000 3

 Clear Quality Gates and enhanced outstanding work visibility

 Shortening of lead time
 Enhanced demand distribution (upwards in the supply chain)
 Production managed in smaller work packages
 Improved schedule balancing
 Smoother work distribution within teams and between them
 Better risk analysis per work package

The methodology followed had three main steps:

 STEP 1: Station Definition. Define the number of stations for the pulse line & First allocation of works to each
station
 STEP 2: Detailed Scheduling. Detailed schedule per station. Task allocation per team
 STEP 3: Risk Analysis

3. Re-industrialization Steps:

3.1. STEP 1: Station definition

At this early stage, there were two main aspects to be taken into account: the investment and industrialization
constraints and the aircraft conversion process.
At the moment of the reindustrialization there existed three hangars with capacity of one aircraft per hangar.
There was no possibility of further hangar investment. Also, it was a must to have some free slots in one of the
hangars in order to deal with outstanding works from the prototype phase.
To decide the work distribution between the stations of the new build process, the former process was reviewed.
Production process had six main work packages:

 Disassembling: some items from the original A330 need to be removed. Some are scraped and others are only
disassembled in order to provide access for the installation of new elements or to prevent damages during the
aircraft conversion.
 Structural Assemblies: This step has a major constraint: its most critical tasks need to be performed with the
aircraft on stress free condition. Therefore, no aircraft movements are possible during its execution.
 Equipment Installation: Electrical installations, hydraulic and fuel pipeline modification. Tests that do not require
aircraft Power On are also included in this step.
 Indoor Power On Tests & First Refurbishing: Tests performed indoor and with the aircraft in Power On. It also
includes the refurbishing of those parts that will not be neither disassembled nor damaged during outdoor tests
 Outdoor Tests
 Final Refurbishing & Delivery: It includes one week for painting, which is done in another factory.

After a deep study of the tasks that had been allocated to each of the main processes during the prototype phase,
it was concluded that some of the tasks could be moved forward or backwards on the process. For example, some
structural non-stress free process could be done during the disassembling period with an overall lead time reduction.
However, most of the tasks had to be kept within the same process milestone, due to precedencies and technological
constraints.
These main processes had to be allocated to the different hangars. It had to be taken into account that the
expected delivery lead time was of 40 weeks and the aim was to have a process cycle as close to 13 weeks as
possible.
Two main options where studied, both of them with a cycle time of 14 weeks. Fig. 2 shows the two alternatives.
The first one was to include the disassembling in Hangar 1 followed by the structural assemblies. Hangar 2 would
host the installations and Hangar 3 the indoor tests, outdoor tests, final refurbishing and delivery. Hangar 2 would
1170 T. Borreguero-Sanchidrian / Procedia Manufacturing 13 (2017) 1167–1174
4 T. Borreguero-Sanchidrian / Procedia Manufacturing 00 (2017) 000–000

have 1 week of buffer time. As for the free hangar slots for prototypes, Hangar 3 would be free for five weeks
during outdoor tests and painting and for two weeks at the end of the cycle (seven weeks each fourteen weeks in all).

Fig.1. Main process milestones.

In option 2, the disassembling would be done in Hangar 3. There would be also one week buffer on Hangar 2 and
two weeks on Hangar 1. Moreover, Hangar 3 would be free only during the five weeks during outdoor tests and
painting. The two last weeks of the cycle time in hangar 3 would be used for the disassembling of the next aircraft.

Dis. Structural Assemblies Installations Indoor T. Outdoor T. F. Ref& Del

HANGAR 1 HANGAR 2 HANGAR 3

Dis. Structural Assemblies Installations Indoor T. Outdoor T. F. Ref& Del Dis.

H. 3 HANGAR 1 HANGAR 2 HANGAR 3 H. 3
WEEK
‐2 ‐1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Fig.2. Two new build process options.

Finally, option 1 was selected. This was due to the fact that the availability of Hangar 3 for other purposes was
prioritized. Another reason was the aim to take advantage of the synergies between disassembling and structural
assemblies.

3.2. STEP 2: Detailed schedule per station

After the hangar distribution, the detailed allocation of work orders to each of the stations was tackled. Also, the
number of worker teams in each station and their capacity was defined. In this sense, it was very important the use
of the balancing and scheduling tool presented in [9]. It gave the opportunity to smoothen the work load throughout
the cycle time and between teams.
Each task was studied in order to know its precedencies, the type of operators that could do it and the number of
operators needed. For these, multi-functional teams worked on the Value Stream Mapping of the stations. All the
data was included in the line balancing and scheduling tool. Fig. 3 shows the precedence diagram for the fuselage
tasks in Hangar 1, one of the most critical areas of the conversion.
 T. Borreguero-Sanchidrian
T. Borreguero-Sanchidrian / Procedia
/ Procedia Manufacturing
Manufacturing 13 (2017)
00 (2017) 1167–1174
000–000 11715

Fig.3. Example of precedence diagram.

At this step, the use of a scheduling tool was really important. For example, for the first station, an early start
schedule needed up to 80 operators whereas a balanced solution had enough with 50 operators, as it can be seen in
Fig. 4. Although at the end of the station the saturation decreases, this is also a risk–decreasing factor as it provides
with some time to deal with incidences without having a direct impact on the next station. Given the complexity of
the precedencies graphs and the presence of several working teams at the same time, this kind of workload
smoothing is unaffordable by hand.
The work allocation between the different teams was also reviewed at this step. Within a station, between ten and
twenty teams can be working at the same time. In most of the cases the allocation of a work to a team is due to
technological reasons. However, challenging this allocation provided an important source of improvement.
The tasks that could be done by several groups where identified. They were usually non-critical tasks. Different
team–task allocations were tested. Then, the most balanced one was chosen. In Fig. 5 there is an example about the
impact of a team re-assignment of some non-critical tasks. The number of operators needed at a station is calculated
as the peak operator demand (as no temporary contracts are envisaged). In the first solution, 14 blue operators and
12 red operators are needed, 26 in all. In the second solution, the number of red operators goes down to 10, being the
new total 24 operators (7% reduction). At the same time, in the second solution the blue team goes on working until
the end of the station and the red one has a lower peak and more extra time at the end for incidence catching up.
61172 T. Borreguero-Sanchidrian
T. Borreguero-Sanchidrian / Procedia
/ Procedia Manufacturing
Manufacturing 13 (2017)
00 (2017) 1167–1174
000–000

Fig.4. Number of operators per day in station 1 for a balanced and earliest start date solution.

By the end of this step, there was a detailed schedule of all of the stations, including all the tasks for the different
working teams.

Fig.5. Comparison of two team-task allocation solutions.

3.3. STEP 3: Risk analysis

While the first two steps looked forward to providing the best possible planned schedule, this step meant to
predict the robustness of that schedule. It is widely known that daily production faces a wide range of disruptions.
Therefore, a robust schedule was defined by Leon et al as the one whose performance remains high in the place of
disruptions [12].
Robust schedules can be obtained by means of stochastic, fuzzy or robust scheduling. Otherwise, sensitivity
analysis can be performed in other to assess the robustness of an existing schedule. [13].
In our case, we used our scheduling tool to perform a sensitivity analysis on each of the work packages.
Being the conversion of the aircraft a sum of tasks to be done all over the aircraft, most of the tasks can be
performed on their one or with little relationship with others. However, there are some areas with a high density of
works, where there is a critical path not only due to precedencies (as is calculated by a PERT diagram) but also
because of limited space and other scarce resources.
 T. Borreguero-Sanchidrian / Procedia Manufacturing 13 (2017) 1167–1174 1173
T. Borreguero-Sanchidrian / Procedia Manufacturing 00 (2017) 000–000 7

Normally, the baseline schedule seeks to reduce the number of operators. Once production has started, delivery
becomes the main objective, even if this results on an over cost. Frequently, outstanding work is reduced by working
extra hours or hiring extra personnel.
This capability to re-schedule was measured in two complementary ways. A first approach studied the impact on
the lead time that had the addition of extra workers to the work package. In a flexible work package, having more
operators available will reduce the completion lead time. However, there are other work packages (those with more
strong constraints) where adding more operators only adds cost but does not reduce proportionally (or not at all) the
completion lead time.
Another way of analyzing this is to plot the evolution of the total operator cost of a work package when
completed in its lower lead time depending on the available number of workers. The operator cost is for this purpose
calculated as the number of workers multiplied by the earliest finish time. Again, flexible work packages will have a
constant cost, no matter the number of available operators. This means that the number of parallel tasks can be
efficiently redistributed. On the other side, non-flexible work packages will suffer a cost increase when the number
of workers goes up as there will be not enough parallel tasks available to efficiently distribute the tasks among the
workers.
Identifying the non-flexible work packages is a key topic for assuring the station performance. A delay on a work
package is more likely to have a direct impact on the station completion date if it is a non-flexible work package.
Therefore, incidences must be prioritized according to their work package’s flexibility.
Furthermore, flexible work packages can be used to smoothen the workload during rescheduling. In case of delay
or scarce resource availability they can be stopped as their delay can be afterwards reduced without an over cost.
In Fig. 6 we present three examples of a Cost–Nb Operators plot for three different flexibility grades.

Fig.6. Cost – Nb Operators plot for three different flexibility grades.

4. Conclusions:

This work presents a real case study of an aircraft final assembly line re-industrialization. In it, the methodology
and tools presented in [9] have been applied.
The result of the case study has been the definition of a new pulse line, which enabled the final assembly line to
be prepared for the product ramp up. The stations have been balanced and the workload has been redistributed
between the stations and also within the teams of each station. The overall saturation coefficient (standard work
content /presence of man hour on the line) was improved in nearly 10%.
Moreover, we have analyzed the criticist of the different work packages within the stations, in order to have a
more efficient incidence management during production.
1174 T. Borreguero-Sanchidrian / Procedia Manufacturing 13 (2017) 1167–1174
8 T. Borreguero-Sanchidrian / Procedia Manufacturing 00 (2017) 000–000

Nevertheless, the more accurate Build Process has had a positive impact also in the supply chain. Part delivery
dates where previously planned in three months slots. After the pulse line definition, they were distributed along
each month of the production planned schedule. In consequence, the part demand was smoother and enabled a better
on time delivery date from customers.
As a result, the product ramp up has been achieved with success. The total man hours have been also reduced
aircraft after aircraft.

Acknowledgements

The authors want to express their most sincere gratitude to the colleagues of AIRBUS Defence & Space, who
kindly collaborated in this work.

References

[1] Recommendations for implementing the strategic initiative INDUSTRIE 4.0, April 2013. Available at: http://www.acatech.de/fileadmin/
user_upload/Baumstruktur_nach_Website/Acatech/root/de/Material_fuer_Sonderseiten/Industrie_4.0/Final_report__Industrie_4.0_accessible.
pdf .
[2] F. Mas, R. Arista, M. Oliva, B. Hiebert, I. Gilkerson, J. Ríos. Proc. Eng. 132 (2015) 1045-1052.
[3] J. Servan, F. Mas, J.L. Menéndez, J. Ríos, Using Augmented Reality in AIRBUS A400M Shopfloor Assembly Work Instructions.
Proceedings of the 4th Manufacturing Engineering Society Intl. Conf. (MESIC 2011), Cadiz (SPAIN), 2011.
[4] H. Regenbrecht, G. Baratoff, W. Wilke. IEEE Comput. Graph. Appl. 25,(2005) 6.
[5] H.A. Scott, Modelling aircraft assembly operations. Proceedings of the 1994 Winter Simulation Conference. (1994) 920-927.
[6] G. Heike, M. Ramulu, E. Sorenson, P. Shanahan, K. Moinzadeh. Int. J. of Prod. Econ. 72 (2) (2001) 103-120.
[7] T. Ziarnetzky, L. Mönch, A. Biele, Simulation of low-volume mixed model assembly lines: modeling aspects and case study. In Proceedings
of the Winter Simulation Conference 2014 IEEE. (2014) 2101-2112.
[8] J. Rios, F. Mas, J.L. Menéndez. Conf. Proc. 1431 (2011) 601-608.
[9] T. Borreguero, F. Mas, J.L. Menéndez, M.A. Barreda, Enhanced Assembly Line Balancing and Scheduling Methodology for the Aeronautical
Industry. 2015 Manufacturing Engineering Society International Conference. Barcelona. (2015).
[10] V.E. Gastelum, Application of lean manufacturing techniques for the design of the aircraft assembly line, Master’s thesis. Massachusets
Institute of Technology. (2002).
[11] http://www.boeing.com/news/frontiers/archive/2004/november/cover.html. Consulted on Mars (2017).
[12] V.J. Leon, S.D. Wu, R.H. Storer. IIE Trans. 26 (5) (1994) 32-43.
[13] W. Herroelen, R. Leus. Eur. J. of Oper. Res. 165 (2005) 289-306.