=Paper=
{{Paper
|id=Vol-2518/paper-FOMI1
|storemode=property
|title=Applied Ontologies for Assembly System
Design and Management within the Aerospace Industry
|pdfUrl=https://ceur-ws.org/Vol-2518/paper-FOMI1.pdf
|volume=Vol-2518
|authors=Rebeca Arista,Fernando Mas,Manuel Oliva,Domingo Morales-Palma
|dblpUrl=https://dblp.org/rec/conf/jowo/AristaMOM19
}}
==Applied Ontologies for Assembly System
Design and Management within the Aerospace Industry==
https://ceur-ws.org/Vol-2518/paper-FOMI1.pdf
Applied Ontologies for Assembly System
Design and Management within the
Aerospace Industry
Rebeca ARISTA a,1, Fernando MAS b, Manuel OLIVA b and
Domingo MORALES-PALMA c
a
Airbus, France
b
Airbus, Spain
c
Universidad de Sevilla, Spain
Abstract. The aerospace industry is characterized by a low-scale manufacturing rate,
producing mid to high level customized products, with high level of complexity.
Factories, and the complete assembly system in charge to manufacture these
products, are not efficiently flexible to new manufacturing scenarios or new product
developments. This work shows a preliminary review of the literature and proposes
the application of ontologies for the assembly system definition and management
within the aerospace industry. The resources addressed are not only human or tools
at factory level, but the complete capital set conforming the aerospace assembly
system. This work will enable trade-off scenarios and re-configure the system to a
new manufacturing scenario or product design.
Keywords. Aerospace Manufacturing Ontologies, Assembly Systems Ontology,
Assembly Line Design, Knowledge-based systems, Models for Manufacturing
1. Introduction
The assembly system of an aerospace product, comprising all the resources that
constitute it, represent 70% of the cost of a product development [1]. During the design
process, a unique deliverable should be made including the product functional and
industrial design, as well as the assembly system design, within a collaborative
engineering process. This key deliverable is named industrial Digital Mock-Up (iDMU)
[2,3]. Industrial requirements are gaining weight on the design conceptual phase, mainly
driven by cost reduction targets on new product developments.
The complexity of the assembly system design lies in different factors: the resources
that form an aerospace assembly system have a complex design process; the industrial
setup is driven by contractual workshare agreements between the manufacturer and the
customers or governments; the product size constrains impact logistic plans and means;
product functional requirements force most of times the use of ad-hoc tools and means
to support the assembly process; among other constrains.
1
Corresponding Author, Rebeca ARISTA, Industrial System Engineer, Airbus, Allée Pierre Nadot,
31700 Blagnac, France; E-mail: rebeca.arista@airbus.com. Copyright © 2019 for this paper by its authors. Use
permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
Moreover, assembly lines in the aerospace industry have particular characteristics
that differ in other industries. They have a low-scale manufacturing rate, produce mid to
high level customized products with high level of complexity, and are dedicated to only
one product family. Other industries (eg. automotive) produce multiple standard products
in the same assembly line, with a medium or high-scale manufacturing rate.
Figure 1. AIRBUS product lifecycle and development milestones.
As shown in Figure 1, Airbus product lifecycle is made of three stages: development,
production and in-service support. The development process starts with a product idea,
launching a feasibility phase both for the product and assembly system. Further
development maturity gates are reached in a collaborative engineering process, creating
a unique final deliverable, the iDMU. After development, the system has to be prepared
for drastic ramp-up increases in the production phase. The same development process
can be launched at different points of the product lifecycle when a new product version
is created, or due to changes on the assembly system requirements.
In order to achieve the maturity gates, decisions on the assembly system are done
sometimes with no detailed consideration of key assembly process constrains. This
problem is due to a lack of tools that can support the assembly system conceptual design,
and do it collaboratively with the product functional and industrial design.
This work shows a preliminary review of the literature on this field, and proposes
the application of ontologies for the assembly system definition and management. The
objective is to set the foundation to support this process considering the distinctive
features of the assembly lines in the aerospace industry, making a first proposal of the
product, process and resources structure relationship that would support this process.
2. Research work conducted on this field
This section describes a preliminary review of the state of the art on the assembly systems
design process, and the way resources are described and managed inside this process.
This is followed by a preliminary literature review on resource modeling, and a
consideration on the evolution from Knowledge-based Engineering methods to Models
and Ontologies applications.
2.1. Resources and their management in the assembly system design process
Following a collaborative engineering approach for the product and assembly system
design process, once having a preliminary product design in the conceptual phase, the
first step towards a preliminary assembly system design is to generate a product
manufacturing breakdown. To support this step, a product structure management tool is
needed to support the different product views generated.
The product design is split into components to be assembled, generating a product
manufacturing breakdown structure made up of components and sub-assemblies. The
work of Janardanan [4] proposes a web-based product structure manager based on STEP
standard PDM Schema, to support designers in this phase.
This manufacturing breakdown defines the set of parts to be assembled, and the
characteristics of this joint. Considering the assembly system requirements defined, a
high level assembly process is sketched to carry out the joints of the components and
sub-assemblies, considering technical precedence of the tasks to be performed, and
initiating the assembly line or assembly system design process.
Mas [5] presents how to generate a product industrial breakdown (called “as-planned
view”) from a product functional breakdown (called “as-designed view”), and link it to
a process structure, that includes process diagrams to consider process time, and a
preliminary resource structure (both structures are called “as-prepared view”). All done
with the purpose of a knowledge-based application to define aircraft final assembly lines
at the conceptual design phase. Based on the previous research, Escalona [6] applies
model-driven engineering in CALIPSOneo project to build an iDMU in practice.
Whitney [7] defines the assembly system or assembly line design process following
these basic factors: capacity planning (available time and required number of units per
year); assembly resource choice; assignment of resources to operations; floor layout;
workstation design; material handling and work transport; part feeding and presentation;
quality (assurance, prevention, and detection); economic analysis; documentation and
information flow; personnel training and participation; and intangibles.
One of the most difficult steps in the design process is to choose among different
resources for each task so that the work is done within the cycle time and the whole
assembly system has minimum cost. Due to product functional requirements, mechanical
equipment may have to be designed specifically for some steps. Often, a company
outsources the design of its assembly lines and is at the mercy of the vendor regarding
types of equipment.
Assembly process planning is the term used to describe this activity, in which the
part assembly sequence and resource usage is determined in an iterative process, to
minimize assembly costs and time [8]. Bukchin [9] work focus on station paralleling and
equipment selection, minimizing the number of stations, and minimizing the total cost.
Complementing the work of Mas [5], Gomez [10] developed a methodology for
assembly process design at the conceptual phase for aerospace products, including multi-
criteria evaluation of possible alternatives using Ant Colony Optimization (ACO)
metaheuristic method, and fuzzy logic theory for solutions evaluation.
DARPA iFAB program work [11] centered on two functions: providing
manufacturability feedback to the designer, and configuring what they call a foundry (or
assembly system) of networked manufacturing capabilities tailored to the final verified
design. This configuration includes supply chain considerations, assembly planning, and
automatically generated computer-numerically-controlled (CNC) and human work
instructions.
2.2. Resource modeling in the literature
There is extensive work in the literature regarding assembly process modeling, to define
assembly requirements, key characteristics, assembly variation, assembly parametrical
models, mathematical or feature models, assembly or manufacturing planning, among
others [12-14]. In these works, resource models are poorly considered, and mainly done
in terms of workforce or means needed within the process.
Whitney [7] highlights the following points to be considered on resources within
process modeling: what resources are applicable or available to a given task; time for
transport from station to station; reuse of resource for several tasks; and reuse of tools at
one station. He also distinguishes three basic types of assembly resources: people, fixed
automation, and flexible automation.
Mas [5] define three different resource levels (line, station and basic), and within the
basic level three types of resources: tools (ad-hoc mechanical equipment), industrial
means (standard means or easily configurable that can be procured), and human
resources (with defined set of skills).
Research on manufacturing resource modeling conducted by Chengying [15],
proposes an architecture of the general model of manufacturing resource as a 3D solid
model composing three aspects: organization structure (divided in 5 levels each
aggregating the lower level manufacturing behavior), capability status (properties of the
manufacturing resource), and development activity (relationships between the product
development stages and the manufacturing resource involved).
The ontological approach for modeling manufacturing resources presented by
Sanfilippo [16] intends to lay down a conceptual framework with a representation of
manufacturing resources, based in the idea that manufacturing resource relates to a
manufacturing process plan as far as it is relevant for some goal specified in the plan. It
proposes as well a high level classification of manufacturing resources based on three
principles: agentivity, mode of deployment and control.
2.3. From Knowledge-based Engineering to Models and Ontologies applications
Capture and knowledge representation using Knowledge-based Engineering (KBE) has
been a paradigm during the last decades. Huge efforts have been made by researches and
industrials through different projects and initiatives like CommonKADS, MOKA
(Methodology and tools Oriented to KBE Application) and other [17,18]. An interesting
application using KBE to design and industrialize tools for High Speed Milling machines
(HSM) was developed by the authors [19].
Lemaignan [20] propose a preliminary upper ontology for manufacturing named
MASON (MAnufacturing’s Semantics ONtology), aimed to draft a common semantic
net in manufacturing domain, and exposing two applications of this ontology: automatic
cost estimation and semantic-aware multi-agent system for manufacturing.
Traband [21] describes the work conducted in DARPA iFAB project on
manufacturability, generating detailed formal models which represent the capabilities of
various manufacturing machines and processes. By mapping these models into the same
semantic domain as the product design, an automatic constrain on the design trade space
can be made, such that designs that are not manufacturable in a given assembly system
configuration are automatically discarded.
Models for Manufacturing (MfM) [22] is a recently proposed methodology to define
manufacturing or industrial ontologies, by generating a set of interconnected models:
Scope models, that defines the limits where the model works; Data models, that include
the information managed in the selected scope; Behavior models, for the inherent
behavior of the system within the given scope; and Semantics models, that considers
generic objects for connection of data model instances to data location, and/or between
models inside ontologies among the models lifecycle.
As described by Mas [23], the product lifecycle management infrastructure that
would support the described type of model-based methodology for manufacturing, called
“PLM generation 3”, would have data models format based on international standards.
3. Open points on resource modeling and management
From the findings on the preliminary research review, the authors describe in this section
the open points on resource modeling and management within the conceptual phase of
the assembly system design process.
First point is the resources consideration. Most of the research is focused on the
product structure, process structure and process optimization in terms of planning and
scheduling. Resources are considered mainly as assembly or manufacturing operations
enablers (eg. workers, tools, etc.), being limiting parameters inside a process
optimization. The question of the resources structure, classes, key parameters, and
assembly system overall optimization is slightly approached.
Another point is the resources scope within the assembly system. Resources should
not be limited to the scope of an operation, but should include all production processes
at the different scales. This means primary resources (e.g. power supply, water supply,
raw materials), logistic resources (e.g. worldwide logistic means, logistics means within
a facility), and all which conforms the product manufacturing and industrial footprint.
One point to be addressed is the assembly system baseline definition and flexibility,
which needs to be correctly modeled as starting point for a new conceptual design. Even
when a completely new product development process is launched, the assembly system
design starts with the baseline information of an existing assembly system setup for an
aircraft product under production phase or other predecessor.
The existing assembly system setup should be considered both, inside the
manufacturer and the extended enterprise. A potential reuse of this setup might be an
industrial requirement or can reduce development costs, relocating available human
resources skills, coping with existing customer workshare agreements, and getting easier
access to needed elementary resources.
Last point is the relationship between the product, process and assembly system
structure (including process time relationship), as well as the application of industrial
ontologies for knowledge-based decision making during the conceptual design phase.
4. Proposed Ontology for product, process and assembly system relationship
Based on the previous research [5], [10] a proposed schema to present a complete
framework for ontology of product, process and resources is presented in Figure 2.
Product structures “as-design” and “as-planned” include the common layer of
parts/subassemblies and joints. For each joint an assembly process plan is developed
giving a collection of assembly plans. Modelling product and assembly processes and
use ontologies in manufacturing has been developed by the previously mentioned
research and by the MfM methodology [22].
When the manufacturing engineers develop the assembly plan, a set of resources are
involved and modelled as part of the industrial solution. Resources are picked-up from a
pool of resources where they could be seen as separate products by others. Initially the
assembly plan makes use of infinite capacity of the resources and the pool of resources
provides models to build the iDMU [3]. Now every assembly plan fits a joint with a
complete industrial solution.
Figure 2. iDMU model with Product, Processes and Resources.
Despite the definition of the assembly processes fulfilling the joints defined in the
Functional definition of the product, assembly processes are arranged as a net linked by
precedence (start-to-start, finish-to-start, etc.) or left as free processes. The net with
precedence defines the complete assembly line. By assigning precedence to every
assembly process and balancing the assembly line against criteria [24], the
manufacturing engineers define the quantity of the resources based on the real resources
constrains and the use of shared resources in the net.
5. Conclusions and further work
This work aims to set the framework for ontologies applications in the aerospace industry,
to support the assembly system design process and management in the conceptual phase:
in particular, the framework to define resources modelling and the relationships with
product and process.
The paper shows a preliminary review of the literature of the assembly system
design process, and how resources are managed within this process. It describes the open
points on resource modeling and management within the conceptual phase of design, and
proposes the application of ontologies including an early stage proposal of product,
process and resources structure relationship to support this process.
Further work will address the ontology detailed definition to support the assembly
system design process in the conceptual phase, with the aim of supporting trade-off
scenarios, decision making and flexibility of the assembly system to re-configure in new
manufacturing scenarios (rate, resources available, structure, etc.) or new product design.
6. Acknowledgements
The authors would like to thank Sevilla University colleagues and AIRBUS colleagues
in France and Spain, for their support and contribution during the development of this
work.
References
[1] I. Marinescu, G. Boothroyd, Product design for manufacture and assembly, Marcel Dekker, Inc, New York,
2002. ISBN: 0-8247-0584-X
[2] F. Mas, J. L. Menéndez, M. Oliva, A. Gómez, J. Ríos, Collaborative engineering paradigm applied to the
aerospace industry, IFIP International Conference on Product Lifecycle Management Proceedings
(2013), 675-684.
[3] F. Mas, J.L. Menendez, M. Oliva, J. Rios, A. Gomez, V. Olmo, iDMU as the Collaborative Engineering
engine: Research experiences in Airbus, Proceedings of IEEE Engineering, Technology and Innovation
International Conference (2014). DOI: 10.1109/ICE.2014.6871594.
[4] V. K. Janardanan, M. Adithan, P. Radhakrishnan, Collaborative product structure management for
assembly modeling. Computers in Industry 59 (8) (2008), 820–832. DOI:
10.1016/J.COMPIND.2008.06.005.
[5] F. Mas, J. Rios, J.L. Menendez, A. Gomez, A process-oriented approach to modeling the conceptual design
of aircraft assembly lines, The International Journal of Advanced Manufacturing Technology 67(1-4)
(2013), 771-784. DOI: 10.1007/s00170-012-4521-5.
[6] M.J. Escalona, J.A. García-García, F. Mas, M. Oliva, C. del Valle, Applying model-driven paradigm:
CALIPSOneo experience, CAiSE Industrial Track (2013), http://ceur-ws.org/Vol-
1017/Paper4CAiSE_IT2013.pdf
[7] D. E. Whitney, Mechanical assemblies: their design, manufacture, and role in product development,
Oxford University Press, Inc., New York, 2004. ISBN: 0-19-515782-6.
[8] B. Q. Su, S. S. Smith, An Integrated Framework for Assembly-Oriented Product Design and Optimization,
Journal of Industrial Technology 19(2) (2003), 1–9.
[9] J. Bukchin, J. Rubinovitz, A weighted approach for assembly line design with station paralleling and
equipment selection. IIE Transactions, 35(1) (2003), 73–85. DOI: 10.1080/07408170304429.
[10] A. Gomez, J. Ríos, F. Mas, A. Vizán, Method and software application to assist in the conceptual design
of aircraft final assembly lines, Journal of Manufacturing Systems 40 (2016), 37-53. DOI:
10.1016/j.jmsy.2016.04.002.
[11] DARPA iFAB Program, https://cps-vo.org/group/avm/ifab (last access May 2019)
[12] N. W. Bowland, J. X. Gao, R. Sharma, A PDM- and CAD-integrated assembly modelling environment
for manufacturing planning, Journal of Materials Processing Technology 138 (1–3) (2003), 82–88. DOI:
10.1016/S0924-0136(03)00053-0.
[13] E. J. Schut, M. Van Tooren, A Knowledge Based Engineering approach to automation of conceptual
design option selection, 45th AIAA Aerospace Sciences Meeting and Exhibit (2013). DOI:
10.2514/6.2007-968.
[14] H. Falgarone, N. Chevassus, An innovative design method and tool for structural and functional analysis,
Proceedings of 14th CIRP Design Seminar (2004).
[15] L. Chengying, W. Xiankui, H. Yuchen, Research on manufacturing resource modeling based on the O-
O method, Journal of Materials Processing Technology 139 (2003), 40-43. DOI: 10.1016/S0924-
0136(03)00179-1.
[16] E. Sanfilippo, S. Benavent, S. Borgo, N. Guarino, N. Troquard, F. Romero, P. Rosado, L. Solano, F.
Belkadi, A. Bernard,. Modeling Manufacturing Resources: An Ontological Approach, Proceedings of the
15th IFIP WG 5.1 International Conference PLM 2018 540 (2018). DOI: 10.1007/978-3-030-01614-
2_28.
[17] CommonKADS, https://commonkads.org/ (last access June 2019).
[18] M. Callot, S. Kneebone, K. Oldham, MOKA: A user-driven Project to Develop Methodology and Tools
oriented to Knowledge Based Engineering Applications, Proceedings of the European Conference
Product Data Technology Days (1998), 19-24.
[19] J. Rios, J.V. Jimenez, J. Pérez, A. Vizán, J.L. Menéndez, F. Mas, KBE Application for the Design and
Manufacture of HSM Fixtures, Journal, Acta Polytechnica 45 (3) (2005).
[20] S. Lemaignan, A. Siadat, J.Y. Dantan, A. Semenenko, MASON: A Proposal For An Ontology Of
Manufacturing Domain, Proceedings of the IEEE Workshop on Distributed Intelligent Systems:
Collective Intelligence and Its Applications (2006). DOI: 10.1109/DIS.2006.48.
[21] M. Traband, Using the iFAB Architecture to Execute Rapid Response Manufacturing, MBE Summit
(2013). Retrieved from http://www.nist.gov/el/msid/upload/2Traband_iFAB.pdf
[22] F. Mas, M. Oliva, J. Racero, D. Morales-Palma, A Preliminary Methodological Approach to Models for
Manufacturing (MfM), Product Lifecycle Management to Support Industry 4.0. PLM 2018. IFIP
Advances in Information and Communication Technology 540 (2018). DOI: 10.1007/978-3-030-01614-
2_25.
[23] F. Mas, R. Arista, M. Oliva, B. Hiebert, I. Gilkerson, J. Rios, A Review of PLM Impact on US and EU
Aerospace Industry, Proceedings of The Manufacturing Engineering Society International Conference
132 (2015). DOI: 10.1016/j.proeng.2015.12.595.
[24] J. Rios, F. Mas, J.L. Menendez, A review of the A400m final assembly line balancing methodology, AIP
Conference Proceedings 1431, 601 (2012). DOI: 10.1063/1.4707614.