Access to this full-text is provided by MDPI.
Content available from Systems
This content is subject to copyright.
systems
Article
Designing and Integrating a Digital Thread System
for Customized Additive Manufacturing in
Multi-Partner Kayak Production
Euan Bonham 1, Kerr McMaster 1, Emma Thomson 1, Massimo Panarotto 2, * ,
Jakob Ramon Müller 2, Ola Isaksson 2and Emil Johansson 3
1Department of Mechanical and Aerospace Engineering, University of Strathclyde, 75 Montrose St,
Glasgow G1 1XJ, UK; euan.bonham.2015@uni.strath.ac.uk (E.B.);
kerr.mcmaster.2015@uni.strath.ac.uk (K.M.); emma.thomson.2015@uni.strath.ac.uk (E.T.)
2Department of Industrial and Materials Science, Chalmers University of Technology,
Chalmersplatsen 4, 412 96 Gothenburg, Sweden; jakob.muller@chalmers.se (J.R.M.);
ola.isaksson@chalmers.se (O.I.)
3RISE IVF, Argongatan 30, 431 22 Mölndal, Sweden; emil.johansson@ri.se
*Correspondence: massimo.panarotto@chalmers.se
Received: 27 September 2020; Accepted: 5 November 2020; Published: 10 November 2020
Abstract:
Additive manufacturing (AM) opens the vision of decentralised and individualised
manufacturing, as a tailored product can be manufactured in proximity to the customers with minimal
physical infrastructure required. Consequently, the digital infrastructure and systems solution
becomes substantially more complex. There is always a need to design the entire digital system so
that different partners (or stakeholders) access correct and relevant information and even support
design iterations despite the heterogenous digital environments involved. This paper describes how
the design and integration of a digital thread for AM can be approached. A system supporting a
digital thread for AM kayak production has been designed and integrated in collaboration with a
kayak manufacturer and a professional collaborative product lifecycle management (PLM) software
and service provider. From the demonstrated system functionality, three key lessons learnt are
clarified: (1) The need for developing a process model of the physical and digital flow in the early
stages, (2) the separation between the data to be shared and the processing of data to perform
each parties’ task, and (3) the development of an ad-hoc digital application for the involvement of
new stakeholders in the AM digital flow, such as final users. The application of the digital thread
system was demonstrated through a test of the overall concept by manufacturing a functional and
individually customised kayak, printed remotely using AM (composed of a biocomposite containing
20% wood-based fibre).
Keywords: additive manufacturing; digital thread; design automation
1. Introduction
Compared to conventional subtractive manufacturing, additive manufacturing (AM)
has numerous benefits, including minimising material waste and an increased capability to make quick
and highly individualised customisations to a product [
1
]. Additionally, AM provides the opportunity
for decentralised production, as the product can be printed wherever an AM machine is located.
Manufacturing the product in close vicinity to the customer has the potential to minimise delivery
time and cost, as well as reduce associated carbon emissions [
2
]. Such a decentralised approach to
manufacturing can be enabled by digital technologies, which allow different parties to contribute to
the manufacturing process. For example, a vision in the kayak manufacturing industry [
3
] is to allow
Systems 2020,8, 43; doi:10.3390/systems8040043 www.mdpi.com/journal/systems
Systems 2020,8, 43 2 of 17
users to customise their kayaks through a User Interface (UI) (Figure 1) and communicate through
a central database. Such information can be used by the original equipment manufacturer (OEM)
to communicate with local AM providers in the proximity of the user, who can then manufacture and
deliver the kayak to the user.
Systems 2020, 8, x FOR PEER REVIEW 2 of 17
[3] is to allow users to customise their kayaks through a User Interface (UI) (Figure 1) and
communicate through a central database. Such information can be used by the original equipment
manufacturer (OEM) to communicate with local AM providers in the proximity of the user, who can
then manufacture and deliver the kayak to the user.
Figure 1. Vision for distributed and multi-party kayak production. The locations on the map are
demonstrative only.
This approach has been covered in a previous research paper [4]; however, this previous
research required significantly improved methods and multi-party integration techniques to bring it
closer to commercialisation standards. These improvements are presented in this paper.
Despite the benefits of such a vision, one key challenge brought by customisation is that every
individual product is traced along the supply chain due to quality assurance and warranty issues. It
is important to know where and how each product has been designed. Additionally, in such a context,
it is important to catalogue and trace all parties’ contributions through the design and manufacturing
process. For example, the transfer of production based on individual specifications from the OEM to
local providers poses serious concerns for manufacturing repeatability, and it is important that faulty
components can quickly be traced to source [5]. Rarely are the conditions the same for different
printing locations. Another challenge is to ensure that all the partners use the most current data and
can react quickly as changes are made. As in this case, the customisation procedure is an actual design
step, requiring several steps to assure the validity of the specific instance being designed.
Due to the novelty of AM, there is a need to design how different parties (or stakeholders) access
information and models, such that they can access appropriate information at the right time.
Additionally, there is a need to design such a system in a streamlined and ideally automated way,
such that it is not labour intensive. In this context, the notion of digital thread [6] is becoming
increasingly important among AM manufacturers. Digital thread refers to the communication
framework that allows a connected data flow and integrated view of the asset’s data throughout its
lifecycle across traditionally siloed perspectives from each party or stakeholder [7]. This thread also
is a prerequisite to trace back the settings and conditions used throughout the design and build phase.
Designing the digital thread for AM production is important when considering how to distinguish
Figure 1.
Vision for distributed and multi-party kayak production. The locations on the map are
demonstrative only.
This approach has been covered in a previous research paper [
4
]; however, this previous research
required significantly improved methods and multi-party integration techniques to bring it closer to
commercialisation standards. These improvements are presented in this paper.
Despite the benefits of such a vision, one key challenge brought by customisation is that every
individual product is traced along the supply chain due to quality assurance and warranty issues. It is
important to know where and how each product has been designed. Additionally, in such a context,
it is important to catalogue and trace all parties’ contributions through the design and manufacturing
process. For example, the transfer of production based on individual specifications from the OEM
to local providers poses serious concerns for manufacturing repeatability, and it is important that
faulty components can quickly be traced to source [
5
]. Rarely are the conditions the same for different
printing locations. Another challenge is to ensure that all the partners use the most current data and
can react quickly as changes are made. As in this case, the customisation procedure is an actual design
step, requiring several steps to assure the validity of the specific instance being designed.
Due to the novelty of AM, there is a need to design how different parties (or stakeholders)
access information and models, such that they can access appropriate information at the right time.
Additionally, there is a need to design such a system in a streamlined and ideally automated way, such
that it is not labour intensive. In this context, the notion of digital thread [
6
] is becoming increasingly
important among AM manufacturers. Digital thread refers to the communication framework that
allows a connected data flow and integrated view of the asset’s data throughout its lifecycle across
traditionally siloed perspectives from each party or stakeholder [7]. This thread also is a prerequisite
to trace back the settings and conditions used throughout the design and build phase. Designing
Systems 2020,8, 43 3 of 17
the digital thread for AM production is important when considering how to distinguish the different
parties involved in the AM industrial system along the digital thread. This raises the following research
question: how can a digital thread for AM kayak production be designed and integrated?
Therefore, this paper describes how the design and integration of a digital thread for AM can be
defined in a system that allows enterprises to anticipate and communicate bi-directionally and up and
downstream where the individual product is in the lifecycle. By doing this, this paper presents a digital
thread system, designed and integrated for kayak AM production. The next Sections (Section 1.1,
Section 1.2, and Section 1.3) provide a literature review focusing on the benefits and challenges related
to AM customisation, as well as on current research in digital threads. Following this, the results of the
research are shown in Section 2, detailing the main building blocks of the digital thread system for
kayak production. These results are discussed in Section 3, where the implications of the presented
digital thread system are discussed. Finally, Section 4presents the research methodology that was
adapted to design and integrate the digital thread system.
1.1. Customisation in Additive Manufacturing: Potential and Challenges
Customisation in manufacturing is motivated by the opportunity to provide users with a
personalised product that more closely matches the demands of a customer compared to mass-produced
products [
8
]. Mass customisation, as opposed to mass production, is defined by Bhrigu
Ahuja et al.
[
9
]
as “the manufacture of similar but not identical products, thus enabling a unique customizable
feature ideally personalized”. Mass customisation originally gained popularity in the 1980s as an
opportunity for companies to gain a competitive advantage by matching demand for increasing
product diversity [
10
]. Similarly, AM is a relatively novel field, which originally emerged to produce
functional and aesthetic prototypes known as rapid prototypes [
9
]. Manufacturing can begin instantly
once a new printable computer aided design (CAD) file is received, meaning AM offers an additional
degree of geometric freedom [
11
], which makes it highly suitable to make quick design alterations and
offer personalised products [
5
]. A good example is where AM is utilised to offer personalised hip and
knee replacements that match a patient’s geometric measurements [11].
Despite this technology’s high potential and growth in the last three decades [
1
], the commercial
uptake has been somewhat stunted for several reasons. Particularly, as the technology is evolving
rapidly, the information exchange associated with the manufacturing chain has become increasingly
digitalised and data-intensive [
12
]. To ensure the quality of the printing phase, extensive monitoring
and control measurements are logged as a part of tuning and optimising the print processes. The success
of the AM build process is further dependent on the design parameters and configurations made
up-stream. The concept of a digital thread is a recurring theme that digitally connects all involved
parties in the AM process. This is especially important when considering manufacturing and material
traceability, but poses additional challenges regarding:
•Conformance with existing standards and lack of existing digital standards [13].
•
The fact that manual processing of digital information can be highly labour intensive. This is
made worse by a disjointed digital thread.
•Remaining adaptable to allow for new parties in the AM process (e.g., new suppliers) [14].
The research in this field identifies that to maximise the potential of AM pertaining to Mass
Customisation, several digital issues must be addressed. As noted by Bonnard et al. [
12
,
13
], a key
difficulty is interacting with and exchanging models, in which the AM digital supply chain convention
for 3D CAD models is an STL representation. This limits the useful information that can be extracted
from the model. Such digital aspects are critical when considering AM being used for customisation,
as this requires local AM providers, end-users, OEMs, and designers to facilitate the digital exchange
of information in a streamlined and effective way. Kim et al. [
14
], developed a methodology for
integrating and streamlining the flow of digital information in the AM product realisation method,
noting there is a need to improve how individuals can interact with data.
Systems 2020,8, 43 4 of 17
As far as the authors of this paper can tell, the consideration of digital thread for AM specific to
mass customisation in literature is very sparse, meaning a potential research gap exists here.
1.2. Multi-Party Industrial Systems in Additive Manufacturing: Potential and Challenges
Traditionally, large manufacturing businesses, such as those in aerospace, have had to deal with
complex and remotely operated industrial systems. This is due to the involvement of many partners in
the supply chain of an aircraft or engine. In basic terms, the supply chain can be viewed as the exchange
of information, models, and materials between multiple parties in a system. The sustainability of
the supply chain relies on these links being secure and traceable [
15
]. As the aerospace industry
has grown and developed in practise to develop complex, individualised products with multi-site
production, they have used their resources to create a more streamlined supply chain that relies heavily
on outsourcing and collaboration. The key players in the industrial system are:
•The OEMs who are the outward-facing organisation providing the product to the customer.
•First tier suppliers who directly supply the OEM.
•Second level suppliers who are the main suppliers of the first tier.
•Third level suppliers who are the suppliers of special components or a high-skilled process.
These complex systems also have access to research and government help [
16
]. This has created
more complex work packages and an element of risk; however, it can be designed to reduce costs and
delivery times [16].
It has been said that you cannot think of a small medium enterprise (SME) as a scaled-down large
business [
17
]. It operates differently, and an SME adopting AM can lead to a more complex industrial
system than it is used to without the resources to handle it. It has been established that even in existing
supply chains, information sharing is important, and when SMEs introduce AM, it will result in further
multi-party collaboration where the transparency and traceability of data are essential to the success of
the business and product [
17
]. This is an area that SMEs struggle to develop and integrate [
17
]. It is
important that a clear solution to this problem is found to make multi-party industrial systems in AM
successful. The authors have identified a gap in the literature regarding this problem in the kayak
production industry and believe a solution can be found.
1.3. Digital Threads
To successfully upscale AM methods from basic prototyping to fully-fledged manufacturing
lines, the method’s core reliance on digital technologies must be addressed [
18
]. Whether researching
new designs and experimenting with new materials or manufacturing customised components, huge
quantities of data are captured throughout a product’s lifecycle [
19
]. This data is required to optimise
the manufacturing process; however, not all information should be freely available to all parties in
the supply chain. Sensitive customer data should not be shared with the local manufacturers and IP
must remain protected. Alongside this, the data must be stored and handled in a way that allows fault
tracking back down the supply chain. This process is described through the phrase “digital thread”.
This is a relatively new term, and as such is not fully understood by industry yet. This provides the
opportunity for research to be completed and the technology gap to be filled.
Traditionally, collection of data and commercial traceability would be achieved manually through
paperwork [
20
]. With large projects, this method can quickly become overwhelming with strict, large
data filing systems being required. In modern-day manufacturing companies, electronic software-based
data recording and storage is utilised. The key advantages of AM allow the rapid creation of complex
and high-value products at a low cost [
21
]. This process and material based optimisation generates
a lot of data that requires analysing, as such this data management is one of the key issues for the
commercial adoption of this technology [
22
]. Currently, the solution to this issue is outdated and relies
on methods developed in the 1980s such as G-code and STL file type combinations [
23
,
24
], requiring
several modifications and file type combinations to produce a final product. These methods require
Systems 2020,8, 43 5 of 17
modernisation and a standard framework for data exchange is key to the overall contextualisation
of the digital thread in this sector [
25
]. This standardisation will ensure key parameters will remain
accessible from design to testing.
One solution could be the additive manufacturing file (AMF) that has been recently developed,
which has native support for colours and materials. Even though standardising this file type would
streamline the data exchange, there is still a requirement to control the movement and access rights to
potentially sensitive data. However, there is little literature available on how to achieve this within this
context, and as such, the technology gap remains. The standardisation of an effective digital thread
remains as the primary method to close this gap.
2. Results
This section describes how a digital thread system for kayak production has been designed and
integrated in collaboration with an established kayak manufacturer—Hereafter named as original
equipment manufacturer (OEM).
2.1. Designing a Digital Thread for Customised Additive Manufacturing in a Multi-Party Industrial System for
Custom Kayaks
The OEM has an interest in AM for its benefits of providing large scale customisable products
that can be printed in proximity to the customer by local AM manufacturers. Compared to traditional
manufacturing methods, AM in principle requires no pre-existing tooling; usually, only a CAD file
needs to be provided for the product to be manufactured. Therefore, AM requires minimal physical
infrastructure (for example, in terms of costly tooling equipment) compared to other manufacturing
methods. However, one of the key challenges for the OEM in this context is that while the physical
infrastructure (machinery required) appears less complex than a conventional machining process,
the digital infrastructure (management of data flow) increases substantially. For distributed AM,
this means sharing richer information (design and production parameters are quite dependent)
effectively. The fact that the dialogue between several partners is a part of a design process means
that the requirement on traceability and effectivity (actuality of data accessed) becomes decisive.
For example, it is crucial that each local AM manufacturer only receives access to the required geometry
files for their specific orders, and that the same information is associated with the actual state of the
version communicated. This is important not just from the perspective of IP rights from the OEM,
but also to avoid confusion and customer dissatisfaction through the accidental production of the
wrong product. As issues with products will inevitably occur, data traceability serves an important
role to isolate and track the source of error (i.e., which party supplied the failing part, etc.). To facilitate
this within the digital thread, it is important to allow multi-directional data flow, but still conserve the
primary objective of selective data sharing. Therefore, the first step of designing the digital thread
was to create a process model of the digital thread itself, in which the physical process and the digital
process are mapped alongside.
Figure 2visualises the process model of the digital thread. This step requires the OEM to interact
with the other stakeholders, as the AM physical and digital processes are substantially different
compared to a process based on traditional manufacturing. Figure 2shows the results of these
interactions, using a modelling method similar to service blueprint [
26
]. How these interactions were
conducted are described in the methodology Section 4.3.
Systems 2020,8, 43 6 of 17
Systems 2020, 8, x FOR PEER REVIEW 6 of 17
Figure 2. Process model of the digital thread for kayak AM production.
In this model, the new stakeholders that are involved in the AM kayak production are identified
and mapped. For example, the OEM needs to provide the CAD model selected by the users to the
local AM manufacturers. At the same time, the manufacturers must provide traceability of data such
as material consumption and source, and power details for use by a CO
2
consultancy firm to conduct
CO
2
analysis. Many manufacturers may choose to complete any post-processing required on their
products inhouse; however, this may be outsourced, dependent on the specialisation required in the
process (e.g., specific powder coatings or heat treating), and this will require the transfer of data
regarding specification and tolerance data from the OEM together with existing measurements from
the manufacturer.
2.2. Integrating the Digital Thread
The data generated throughout the physical process is required by various parties along with
the manufacturing process flow. However, this reveals an issue with such a multi-party industrial
system. Not all data should be accessed by all parties; for example, user address and card payment
data require strict restrictions for both user security and data protection act compliance. Likewise,
the courier does not require access to the manufacturing options such as material data that are
selected at the same stage. To solve this issue, the approach taken in collaboration with a professional
product lifecycle management (PLM) company was to design a digital infrastructure in which each
party (or stakeholder) uses dedicated digital applications to interact bi-directionally with a single
collaborative PLM platform through application programming interfaces (APIs). The motivation for
this approach is to separate the information contained in the PLM platform from the data processing
that each party performs to fulfil their tasks. In this way, each party can extract only the allowed
information, process this data through its dedicated applications, and input this data in the PLM
database. This ensures the use of the most current data at each step of the physical process, as parties
working with downstream processes cannot modify the data produced upstream. Additionally, this
approach allows each party to select the access rights to be allowed to other parties (e.g., only read,
read/write etc.). Figure 2 visualises this separation between the collaborative PLM database and the
digital applications used by each party. While some of these applications are commercial software
used already by the parties (e.g., a commercial robot simulation software), some other applications
needed to be developed ad-hoc to demonstrate the digital thread.
The following sections will describe in more detail these applications, in particular:
• A web-based user interface (UI) allowing kayak customers to configure their desired kayaks,
• a CAD design automation framework, to pre-generate desired kayak configurations,
• a cost and CO
2
model to visualise the cost and CO
2
impacts of the selected kayaks.
Figure 2. Process model of the digital thread for kayak AM production.
In this model, the new stakeholders that are involved in the AM kayak production are identified
and mapped. For example, the OEM needs to provide the CAD model selected by the users to the
local AM manufacturers. At the same time, the manufacturers must provide traceability of data such
as material consumption and source, and power details for use by a CO
2
consultancy firm to conduct
CO
2
analysis. Many manufacturers may choose to complete any post-processing required on their
products inhouse; however, this may be outsourced, dependent on the specialisation required in the
process (e.g., specific powder coatings or heat treating), and this will require the transfer of data
regarding specification and tolerance data from the OEM together with existing measurements from
the manufacturer.
2.2. Integrating the Digital Thread
The data generated throughout the physical process is required by various parties along with the
manufacturing process flow. However, this reveals an issue with such a multi-party industrial system.
Not all data should be accessed by all parties; for example, user address and card payment data require
strict restrictions for both user security and data protection act compliance. Likewise, the courier does
not require access to the manufacturing options such as material data that are selected at the same
stage. To solve this issue, the approach taken in collaboration with a professional product lifecycle
management (PLM) company was to design a digital infrastructure in which each party (or stakeholder)
uses dedicated digital applications to interact bi-directionally with a single collaborative PLM platform
through application programming interfaces (APIs). The motivation for this approach is to separate
the information contained in the PLM platform from the data processing that each party performs to
fulfil their tasks. In this way, each party can extract only the allowed information, process this data
through its dedicated applications, and input this data in the PLM database. This ensures the use
of the most current data at each step of the physical process, as parties working with downstream
processes cannot modify the data produced upstream. Additionally, this approach allows each party
to select the access rights to be allowed to other parties (e.g., only read, read/write etc.). Figure 2
visualises this separation between the collaborative PLM database and the digital applications used
by each party. While some of these applications are commercial software used already by the parties
(e.g., a commercial
robot simulation software), some other applications needed to be developed ad-hoc
to demonstrate the digital thread.
The following sections will describe in more detail these applications, in particular:
•A web-based user interface (UI) allowing kayak customers to configure their desired kayaks,
•a CAD design automation framework, to pre-generate desired kayak configurations,
Systems 2020,8, 43 7 of 17
•a cost and CO2model to visualise the cost and CO2impacts of the selected kayaks.
The following sections will describe in more detail these applications, and how each step of the
AM physical process is supported by the digital thread.
2.3. Web-Based User Interface (UI) to Support the Customer Perspective
The customer perspective encompasses the user and purchaser (these may be separate parties,
e.g., when purchasing a product as a gift). Customers and purchasers use a web-based UI to input
their user data. User data is a broad term that encompasses shipping and payment details such as user
address, card numbers, delivery type, and customisation and manufacturing specific detail. Each user
order is also assigned a unique order number. This is used at various stages in the process to connect
all order-specific data. In this case study, a UI was created using Vue.js and Firebase. A snapshot of the
UI can be seen in Figure 3. Thus, a digital application was created to effectively harvest the appropriate
data from the user to create a kayak that met all their needs. Using the UI, the customer would go step
by step to customise the kayak by selecting options from dropdown menus.
Systems 2020, 8, x FOR PEER REVIEW 7 of 17
The following sections will describe in more detail these applications, and how each step of the
AM physical process is supported by the digital thread.
2.3. Web-Based User Interface (UI) to Support the Customer Perspective
The customer perspective encompasses the user and purchaser (these may be separate parties,
e.g., when purchasing a product as a gift). Customers and purchasers use a web-based UI to input
their user data. User data is a broad term that encompasses shipping and payment details such as
user address, card numbers, delivery type, and customisation and manufacturing specific detail. Each
user order is also assigned a unique order number. This is used at various stages in the process to
connect all order-specific data. In this case study, a UI was created using Vue.js and Firebase. A
snapshot of the UI can be seen in Figure 3. Thus, a digital application was created to effectively
harvest the appropriate data from the user to create a kayak that met all their needs. Using the UI,
the customer would go step by step to customise the kayak by selecting options from dropdown
menus.
Figure 3. Web-based user interface.
These options include:
• Kayak base model,
• user anthropometrics (height, weight, and waist size),
• skeg,
• rudder,
• hatch and Storage areas,
• material selection,
These options were identified after a workshop with the kayak OEM, focused around an exercise
using a Kano model [27], to identify “attractive”, “one-dimensional”, and “must-be” features for a
kayak. How the Kano-model was used in this research is explained in the methodology Sections 4.1
Figure 3. Web-based user interface.
These options include:
•Kayak base model,
•user anthropometrics (height, weight, and waist size),
•skeg,
•rudder,
•hatch and Storage areas,
•material selection.
Systems 2020,8, 43 8 of 17
These options were identified after a workshop with the kayak OEM, focused around an exercise
using a Kano model [
27
], to identify “attractive”, “one-dimensional”, and “must-be” features for a
kayak. How the Kano-model was used in this research is explained in the methodology Sections 4.1
and 4.2. Once the options have been selected, the customer would be able to see a preview of their
custom kayak, along with the cost and the CO
2
output associated with the manufacture of their
unique product.
If the customer is satisfied with the kayak, the order would be submitted to a database, where it is
assigned a randomly generated order number, meaning that all data from the order can be traced to
that unique key. This is important, because some of the customer ’s data will be shared, but ensuring
the data is connected even though certain elements will be hidden from the other parties. Traditionally,
this step is followed by the generation of a CAD model by a designer, who will customise the kayak
based on customer preferences. As this step would be particularly labour intensive, the approach taken
was to exploit design automation (DA) techniques to pre-generate all the possible kayak configurations.
2.4. CAD-Based Design Automation Framework
The design step in the physical process flow (Figure 2) takes the user’s data, which is then fed to a
Python-based journal code (Figure 4). This is a DA technique, which is based in a commercial CAD
software package. A master CAD file is required to automatically be edited by this code, producing a
personalised model. This will be created and operated by the OEM (i.e., the local AM manufacturers
only receive a STEP file generated automatically by the OEM).
Systems 2020, 8, x FOR PEER REVIEW 8 of 17
and 4.2. Once the options have been selected, the customer would be able to see a preview of their
custom kayak, along with the cost and the CO2 output associated with the manufacture of their
unique product.
If the customer is satisfied with the kayak, the order would be submitted to a database, where it
is assigned a randomly generated order number, meaning that all data from the order can be traced
to that unique key. This is important, because some of the customer’s data will be shared, but
ensuring the data is connected even though certain elements will be hidden from the other parties.
Traditionally, this step is followed by the generation of a CAD model by a designer, who will
customise the kayak based on customer preferences. As this step would be particularly labour
intensive, the approach taken was to exploit design automation (DA) techniques to pre-generate all
the possible kayak configurations.
2.4. CAD-Based Design Automation Framework
The design step in the physical process flow (Figure 2) takes the user’s data, which is then fed to
a Python-based journal code (Figure 4). This is a DA technique, which is based in a commercial CAD
software package. A master CAD file is required to automatically be edited by this code, producing
a personalised model. This will be created and operated by the OEM (i.e., the local AM manufacturers
only receive a STEP file generated automatically by the OEM).
The master CAD file is key to allowing a DA approach to CAD model generation. This required
specific ratios and formula-based measurements to be input to the dimensioning and sketching phase
of the CAD generation. This, combined with correct constraining of the model, allowed specific
dimensions (e.g., length and width) to be input as a single dimension, and the model would update
to these new specifications but maintain its overall structure and predefined shape.
Figure 4. CAD Model Generation Sequence.
Figure 4. CAD Model Generation Sequence.
Systems 2020,8, 43 9 of 17
The master CAD file is key to allowing a DA approach to CAD model generation. This required
specific ratios and formula-based measurements to be input to the dimensioning and sketching phase
of the CAD generation. This, combined with correct constraining of the model, allowed specific
dimensions (e.g., length and width) to be input as a single dimension, and the model would update to
these new specifications but maintain its overall structure and predefined shape.
This type of CAD modelling is referred to as parameterising, and the model produced is known
as a master file, as when the python code is running, it selects the dimensions in the model that
require changing (e.g., the cockpit width must adjust in coordination with the user’s waist width)
and automatically cycles each through a predefined list; examples of this can be seen in Table 1.
Table 1. CAD model variable parameters example.
Dimensional Parameters Accessories
ID Length (mm) Width (mm) Cockpit Width (mm) Storage Hatch Layout Rudder
0 4500 480 400 1 Yes
1 4500 480 400 1 No
2 4500 480 400 2 Yes
3 4500 480 400 2 No
. . . . . . . . . . . . .. . . . .
1500 4500 480 400 7 Yes
1501 4500 480 400 7 No
1502 4500 500 430 1 Yes
1503 4500 500 430 1 No
1504 4500 500 430 2 Yes
1505 4500 500 430 2 No
. . . . . . . . . . . . .. . . . .
3195 5500 600 490 7 Yes
3196 5500 600 490 7 No
The code was created in a way that allowed each possible combination of both dimensional
changes and the addition or absence of any optional features to be created and saved with a specified
naming mechanism. This allowed thousands of CAD models to be created and stored logically without
any additional human interaction or additional hours being spent. The file type can also be specified,
which is important for the consistency of the digital thread and utilisation of other parties further
down the process flow. Pre-generating the individual versions also allow the designer the ability to
eliminate undesirable or unwanted instances, but this was not further explored in this study.
2.5. Pre-Manufacturing Processes and CO2& Cost Model Generation
The design of this master CAD file must be verified through an iterative collaboration with
pre-manufacturing simulations, most likely outsourced to a specialised consultancy company
(for complex processes, multiple external consultancies may be used). This requires geometry data
output as a stereolithography (STL) file, alongside other specific requirement data, to be exchanged
between the OEM and consultancy parties. From a data management perspective, this enforces the
need to capture the changes and additions made in each step against the versions initiated. This stage of
the process also requires a CO
2
model to provide meaningful feedback for increasingly environmentally
aware customers. Most likely, this model is developed by an external consultancy firm. In this case
study, this was developed in house for demonstration purposes, and integrated with the UI. A life cycle
assessment (LCA) approach was developed using open-source software, open LCA [28], considering
only the manufacturing elements of development (i.e., no waste considerations). Figure 5illustrates
the overall logic of the developed LCA and cost models.
Systems 2020,8, 43 10 of 17
Systems 2020, 8, x FOR PEER REVIEW 10 of 17
Figure 5. CO
2
model visualisation.
This model pieced together the various stages of manufacture that contributed to CO
2
emissions.
This includes material granulate production, transportation, and electricity consumption. Figure 5
displays the role of a CO
2
consultancy in this case, and how the data generated fits back into the
overall process. To generate a linear prediction expression across the entire range of kayak model
volumes, statistical analysis software was used to develop a surrogate linear regression model. This
was achieved by sampling the volumes of several kayaks in the LCA study and then populating the
CO
2
emission values for intermittent kayak volumes through the model. Then the output of this
model is fed back in real-time to the UI to display to the user. This information about the model and
user are utilised in JavaScript Object Notation (JSON) format by the OEM for use in populating a
database which feeds the web-based UI. At this stage, the purchaser will confirm their order and the
STL file will be sent to the manufacturer.
Depending on the material options selected by the user and ranging kayak volumes available,
the CO
2
emissions varied between 41.60 and 128.51 kg CO
2
equivalent according to the global
warming potential (GWP) as determined within the LCA software.
Similarly, a cost analysis was also important to allow the customer to be able to assess the trade-
off between CO
2
emissions and cost of the kayak (i.e., the customer may be willing to choose a more
sustainable material at greater personal expense). The cost model considers the manufacturing costs
associated with the three critical stages of the AM process, which include pre-processing, printing,
and post-processing. These costs are based on the time and operator costs for each phase, as well as
material costs for printing. The pre- and post-processing costs remain fixed for all kayak volumes,
while the printing cost varies depending on the volume of kayak and material choice as selected by
the user. At this point, the live digital input is required in a loop, as print time varies depending on
Figure 5. CO2model visualisation.
This model pieced together the various stages of manufacture that contributed to CO
2
emissions.
This includes material granulate production, transportation, and electricity consumption. Figure 5
displays the role of a CO
2
consultancy in this case, and how the data generated fits back into the overall
process. To generate a linear prediction expression across the entire range of kayak model volumes,
statistical analysis software was used to develop a surrogate linear regression model. This was achieved
by sampling the volumes of several kayaks in the LCA study and then populating the CO
2
emission
values for intermittent kayak volumes through the model. Then the output of this model is fed back in
real-time to the UI to display to the user. This information about the model and user are utilised in
JavaScript Object Notation (JSON) format by the OEM for use in populating a database which feeds
the web-based UI. At this stage, the purchaser will confirm their order and the STL file will be sent to
the manufacturer.
Depending on the material options selected by the user and ranging kayak volumes available,
the CO
2
emissions varied between 41.60 and 128.51 kg CO
2
equivalent according to the global warming
potential (GWP) as determined within the LCA software.
Similarly, a cost analysis was also important to allow the customer to be able to assess the trade-off
between CO
2
emissions and cost of the kayak (i.e., the customer may be willing to choose a more
sustainable material at greater personal expense). The cost model considers the manufacturing costs
associated with the three critical stages of the AM process, which include pre-processing, printing,
and post-processing. These costs are based on the time and operator costs for each phase, as well as
material costs for printing. The pre- and post-processing costs remain fixed for all kayak volumes,
while the printing cost varies depending on the volume of kayak and material choice as selected by
the user. At this point, the live digital input is required in a loop, as print time varies depending
Systems 2020,8, 43 11 of 17
on the kayak volume. The AM provider can perform simulations to determine the print time for a
corresponding geometry. For this case, this information was determined manually for a select few
kayak volumes in collaboration with the AM provider, but further work is required to fully integrate
such aspects digitally.
Depending on the material options, ranging kayak volumes available, and delivery location,
the cost varied between €746.76 and €793.61.
2.6. Manufacturing, Post-Manufacturing, and Delivery
Another important step during pre-manufacturing is to conduct a robot simulation to define the
most effective robot path. In the demonstration of the digital thread, robot simulations have been
performed by accessing the CAD files through the collaborative PLM (Figure 6a).
Systems 2020, 8, x FOR PEER REVIEW 11 of 17
the kayak volume. The AM provider can perform simulations to determine the print time for a
corresponding geometry. For this case, this information was determined manually for a select few
kayak volumes in collaboration with the AM provider, but further work is required to fully integrate
such aspects digitally.
Depending on the material options, ranging kayak volumes available, and delivery location, the
cost varied between €746.76 and €793.61.
2.6. Manufacturing, Post-Manufacturing, and Delivery
Another important step during pre-manufacturing is to conduct a robot simulation to define the
most effective robot path. In the demonstration of the digital thread, robot simulations have been
performed by accessing the CAD files through the collaborative PLM (Figure 6a).
(a) (b)
Figure 6. Remote pre-manufacturing and manufacturing of kayaks: (a) Simulation of manufacturing
using ABB Robot Studio
®
; (b) manufacturing of kayak (made of biocomposite containing 20% wood-
based fibre).
After simulations are performed (checking the feasibility of production), the manufacturing
phase is executed. To demonstrate the feasibility of the digital thread system for remote production,
the manufacturing was performed by a research institute that possesses dedicated equipment for
large scale AM (Figure 6b). The institute remotely accessed the STL files of the kayak through the
collaborative PLM interface. A functional AM kayak (composed of a biocomposite containing 20%
wood-based fibre) was then manufactured.
In a real-life scenario, the completed product will require delivery to the user. Although the
concept of a local manufacturer facilitates user collection of products, this may not always be feasible,
and as such, a courier will be required. This party requires no information about the manufacturing
process, only requiring specific user data such as address and name.
3. Discussion
The concept of the digital thread for AM is still evolving. Hence, the practices for designing and
integrating a digital thread for AM needs to be investigated through multiple use cases with
industrial partners. One such example, presented by Mandolla et al. [29], uses blockchain to support
a digital twin for AM in the aircraft industry, showing great potential concerning manufacturing
infrastructure and product component history. Similarly, Lui et al. [30], develops a data management
system for metal AM. This effectively tackles the challenges associated with metal AM (e.g., process
repeatability etc.) to manage the various product lifecycle stages and support a more efficient
application of metal AM. Given the range of challenges and novel proposals in the field as well as the
rapidly changing nature of the technology, it is clear there is a lack of a widely accepted framework
for the digital thread for AM. There may indeed be no ideal one size fits all approach.
Compared to previous studies, this research has put a stronger emphasis on the design aspects
of a digital thread. One of the key insights for this design focus is that while with AM the physical
infrastructure decreases compared to traditional manufacturing methods (e.g., less tooling
Figure 6.
Remote pre-manufacturing and manufacturing of kayaks: (
a
) Simulation of manufacturing
using ABB Robot Studio
®
; (
b
) manufacturing of kayak (made of biocomposite containing 20%
wood-based fibre).
After simulations are performed (checking the feasibility of production), the manufacturing
phase is executed. To demonstrate the feasibility of the digital thread system for remote production,
the manufacturing was performed by a research institute that possesses dedicated equipment for
large scale AM (Figure 6b). The institute remotely accessed the STL files of the kayak through the
collaborative PLM interface. A functional AM kayak (composed of a biocomposite containing 20%
wood-based fibre) was then manufactured.
In a real-life scenario, the completed product will require delivery to the user. Although the
concept of a local manufacturer facilitates user collection of products, this may not always be feasible,
and as such, a courier will be required. This party requires no information about the manufacturing
process, only requiring specific user data such as address and name.
3. Discussion
The concept of the digital thread for AM is still evolving. Hence, the practices for designing and
integrating a digital thread for AM needs to be investigated through multiple use cases with industrial
partners. One such example, presented by Mandolla et al. [
29
], uses blockchain to support a digital
twin for AM in the aircraft industry, showing great potential concerning manufacturing infrastructure
and product component history. Similarly, Lui et al. [
30
], develops a data management system for
metal AM. This effectively tackles the challenges associated with metal AM (e.g., process repeatability
etc.) to manage the various product lifecycle stages and support a more efficient application of metal
AM. Given the range of challenges and novel proposals in the field as well as the rapidly changing
nature of the technology, it is clear there is a lack of a widely accepted framework for the digital thread
for AM. There may indeed be no ideal one size fits all approach.
Systems 2020,8, 43 12 of 17
Compared to previous studies, this research has put a stronger emphasis on the design aspects
of a digital thread. One of the key insights for this design focus is that while with AM the physical
infrastructure decreases compared to traditional manufacturing methods (e.g., less tooling equipment),
the digital infrastructure increases substantially in complexity. In this regard, the model of the physical
and digital flow developed in the early stages (Figure 2) shows this increase in digital complexity.
Additionally, the digital applications that needed to be developed in this case study (e.g., the web-based
UI and the DA framework) show how new stakeholders need to be involved in the AM digital flow in
a different way than before. The results of these studies suggest to decision- and policymakers how
the design of the digital interactions between the partners is decisive for the uptake of decentralised,
customised, and near net shape manufacturing.
This research underlines the importance of careful management of information within AM for
it to successfully integrate into the manufacturing industry [
31
]. This research has looked at these
aspects in collaboration with a professional PLM platform provider. The industrial implementation of
PLM platforms indicates its usefulness. However, examples of integrating a streamlined AM process
with a corresponding digital thread using PLM in literature are scarce [
32
]. Indeed, Hedberg et al. [
33
]
explore a novel PLM concept, observing that a key challenge of PLM relates to the setting in which
data is required (i.e., different users along the digital thread interact with the same data in different
ways, requiring different data formats, access rights, etc.). The accessibility and adaptability of such
platforms is a key step towards the industrialisation of AM. One key insight that emerges from the
study presented in this paper is the separation between the data to be shared and the processing of data
to perform each party’s task. In this way, each party can extract only the allowed information, process
this data through its dedicated applications, and input this data in the PLM database. This ensures the
use of the most current data at each step of the physical process.
There are limitations in this approach that need further attention in the future. Although the
development of a UI to support customisable AM products through DA is not novel, it is important to
remark that this was developed to link not only between the designer, manufacturer, and customer,
but also to placeholder consultancies that developed and integrated cost and CO
2
models associated
with the AM process, enabling live feedback to be provided to the customer. The database behind
the UI stored all kayak design variants, while also providing appropriate movement and access to
information for the multiple parties associated with the process.
It is important to note that the digital thread that has been designed serves as a demonstration
of certain capabilities, with not all parts yet fully complete. One key limitation of this approach is
the requirement to redesign the digital thread for different AM applications with changing parties,
with this approach being very specific to the kayak manufacture case study. Future research work
will focus on the streamlined integration of the different digital tools with the collaborative platform.
This would allow better traceability and data security, as it follows industry standards.
4. Materials and Methods
The framework for Action Research [
34
], was used in the development of the digital thread
for the custom kayak case study. Action research describes a methodology in which researchers
implement a theory or solution within a real-life situation, while also iteratively receiving feedback and
improving performance. The research was conducted in close collaboration with an established kayak
manufacturer, who has already invested in pre-production of AM kayaks. An important stage in the
research was to identify design requirements and features for a kayak. This was achieved by conducting
a Kano-model-based [27], survey in conjunction with a focus group with the industrial partners.
4.1. Use of a Kano-Model-Based Survey to Identify Requirements and Features for Kayak Design
The Kano model distinguishes between three different categories of product requirements:
•
Must-be: Customers will expect these requirements and they are essentially taken for granted,
resulting in extreme dissatisfaction if they are absent.
Systems 2020,8, 43 13 of 17
•
One-Dimensional: These result in satisfaction if present and dissatisfaction if they are not, with
the level of customer satisfaction proportional to the amount to which the requirement is fulfilled.
•
Attractive: These attributes are not expected by the customer, so when they are present, they are
said to delight the customer, but when they are not, there is no dissatisfaction.
An online Kano survey was distributed to experienced sea touring kayakers. The survey was
designed around a five-point scale, in which opinions about customisable features were investigated.
A range of appropriate features was identified through a literature survey as well as consultation
with experienced kayakers. Additionally, there was an open question at the end in which the user
could enter any other kayak features that were not addressed within the survey. The total number
of responses for the survey was 35 participants, and Figure 7shows the results of the survey after
applying the satisfaction and dissatisfaction indexes as described in [
27
]. The customer satisfaction
(CS) coefficient is a useful analysis tool that identifies if satisfaction will be improved by fulfilling a
product requirement, or if meeting this product requirement only avoids dissatisfaction.
Systems 2020, 8, x FOR PEER REVIEW 13 of 17
• Attractive: These attributes are not expected by the customer, so when they are present, they are
said to delight the customer, but when they are not, there is no dissatisfaction.
An online Kano survey was distributed to experienced sea touring kayakers. The survey was
designed around a five-point scale, in which opinions about customisable features were investigated.
A range of appropriate features was identified through a literature survey as well as consultation
with experienced kayakers. Additionally, there was an open question at the end in which the user
could enter any other kayak features that were not addressed within the survey. The total number of
responses for the survey was 35 participants, and Figure 7 shows the results of the survey after
applying the satisfaction and dissatisfaction indexes as described in [27]. The customer satisfaction
(CS) coefficient is a useful analysis tool that identifies if satisfaction will be improved by fulfilling a
product requirement, or if meeting this product requirement only avoids dissatisfaction.
Figure 7. Results of the Kano-model-based survey.
Thigh braces have a dissatisfaction coefficient of 0.79, which leads to greater dissatisfaction if
absent than satisfaction if they are present. This means that thigh braces are a product feature for
kayakers that most users will be expecting (must-be). The satisfaction coefficient for the presence of
a skeg was higher than the dissatisfaction coefficient, meaning this feature has more of a tendency to
be an exciting or surprising feature. For internal storage, the satisfaction and dissatisfaction
coefficients are similar, which is inherently characteristic of a one-dimensional feature. These results
reinforce the range of importance that various features have to kayakers, indicating the suitability of
kayak manufacture to providing customisable products through an AM platform.
4.2. Focus Group to Refine the Requirements and Features for Kayak Design
The Kano-based survey was a solid foundation for the suitability of certain product features for
a customisable kayak. However, certain product features, such as internal storage, were very vague,
and it was going to be a challenge to translate this directly to the modelling phase. Therefore, a focus
group with the kayak manufacturer was conducted. In this focus group, the basis of the Kano model
was explained, and then an exercise was proposed in which the kayak experts were given post-it
notes. They could write down features and put them into the three categories following the Kano
model. The result of this workshop is shown in Table 2.
Figure 7. Results of the Kano-model-based survey.
Thigh braces have a dissatisfaction coefficient of 0.79, which leads to greater dissatisfaction if
absent than satisfaction if they are present. This means that thigh braces are a product feature for
kayakers that most users will be expecting (must-be). The satisfaction coefficient for the presence of a
skeg was higher than the dissatisfaction coefficient, meaning this feature has more of a tendency to be
an exciting or surprising feature. For internal storage, the satisfaction and dissatisfaction coefficients
are similar, which is inherently characteristic of a one-dimensional feature. These results reinforce
the range of importance that various features have to kayakers, indicating the suitability of kayak
manufacture to providing customisable products through an AM platform.
4.2. Focus Group to Refine the Requirements and Features for Kayak Design
The Kano-based survey was a solid foundation for the suitability of certain product features for a
customisable kayak. However, certain product features, such as internal storage, were very vague,
and it was going to be a challenge to translate this directly to the modelling phase. Therefore, a focus
group with the kayak manufacturer was conducted. In this focus group, the basis of the Kano model
Systems 2020,8, 43 14 of 17
was explained, and then an exercise was proposed in which the kayak experts were given post-it notes.
They could write down features and put them into the three categories following the Kano model.
The result of this workshop is shown in Table 2.
Table 2. Kayak Features.
Feature Category Configurable? Printed?
Deck lines Must-be No No
Footrest One-dimensional No No
Skeg One-dimensional No Yes
Rudder One-dimensional No Yes
Bulkheads One-dimensional No Yes
Handling lugs One-dimensional No No
Hatches Attractive Yes Yes
Deck recesses Attractive Yes Yes
Material choice Attractive Yes Yes
Kayak volume Attractive Yes Yes
Cockpit size Attractive Yes Yes
Thigh braces Attractive Yes Yes
Footrest position Attractive Yes No
Personalised backrest Attractive Yes No
One main result from the workshop was that any configurable features, that is, any features of the
kayak that could be tailored to suit the anthropometric measurements or personal preference of the
customer, would be attractive for the user. One-dimensional features were determined to be ones that
are simply there or not, and finally, must-be features are those that are essential. Additionally, it was
made apparent from the session which of the features could be 3D printed and which would be added
during the post-processing phase of production.
Given that it would only be appropriate for printable features to be modelled through CAD, the
session assisted with prioritising features for the kayak model. These features are shown with ‘Yes’
under the column ‘Printed’ in Table 2. Thereafter, an appropriate and verified customisation strategy
had been identified to take forward into the modelling phase.
4.3. Workshops for Designing the Digital Thread System
The digital thread system was designed after several workshops with industrial partners. At the
beginning of the study, a prototypical version of a digital thread [
4
], was presented to the industrial
partners in training sessions led by a professional company that offers a collaborative PLM software.
These training sessions defined the architecture of the digital thread and its associated digital
infrastructure (and are presented in Figure 2).
Once the final design had been created, a proof of concept trial was completed. This phase of
the project was to demonstrate that the company—Customer interaction tool that was developed
would be a commercially appropriate option. The application of the digital thread was demonstrated
through a test of the overall concept, by manufacturing a functional AM kayak (composed of a
biocomposite containing 20% wood-based fibre) remotely. This proof of concept displays the feasibility
of the approach developed, with multiple parties contributing to an AM process in a streamlined and
effective way.
5. Conclusions
This paper has presented a case study aiming at designing and integrating a digital thread system
for AM kayak production in which kayaks are customised and manufactured in the proximity of the
customer. In this context, the tracing of sources and decisions made are key to realising a sufficient
degree of quality in a customisation system (digital thread).
Systems 2020,8, 43 15 of 17
Regarding the research question: how can a digital thread for AM kayak production be designed and
integrated? This study has shown that designing a digital thread system is a matter of implementing
a solution within a real-life situation as a proof-of-concept, in order to iteratively receive feedback
from a number of relevant stakeholders. For this reason, working with design tools such as a service
blueprint and the Kano model to perform a careful design and integration of the digital thread is a
decisive factor to promote customisation and near net shape manufacturing. Additionally, developing
a demonstration of a digital thread system on a real case was considered a crucial factor for iteratively
refining the digital thread itself. While the designed digital thread system is still at a demonstration
stage and can be considered as a proof-of-concept (as there is more elaborate knowledge needed within
all steps), the key take away is that building a customisation system similar to the one presented
represents a source of both benefits and risks for the manufacturer. The degree of complexity of the
digital infrastructure is unprecedented for small-scale manufacturers such as the kayak manufacturer
that participated in this research. For this reason, there is a need to capture and manage information
more carefully.
Author Contributions:
Conceptualisation: E.B., K.M., and E.T.; methodology: E.B., K.M., and E.T.; formal analysis:
E.B., K.M., and E.T.; writing—Original draft preparation: E.B., K.M., and E.T.; writing—Review and editing: E.B.,
K.M., E.T., and M.P.; supervision: M.P., E.J., and J.R.M.; project administration: M.P. and E.J.; funding acquisition:
O.I. All authors have read and agreed to the published version of the manuscript.
Funding:
This research was funded by VINNOVA, the Swedish Energy Agency, and Formas through the strategic
innovation program Produktion2030, Reference number 2017-04776. The research was performed in the DISAM
project (Digitalization of Supply Chain in Swedish Additive Manufacturing).
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Holmström, J.; Partanen, J.; Tuomi, J.; Walter, M. Rapid manufacturing in the spare parts supply chain:
Alternative approaches to capacity deployment. J. Manuf. Technol. Manag. 2010,21, 687–697. [CrossRef]
2.
Peels, J. Decentralized Production with 3D Printing. Available online: https://3dprint.com/174375/
decentralize-production-3d-print/#:~{}:text=Instead%20of%20one%20central%20factory,demand%
20using%203D%20printing%20locally.&text=Decentralized%20production%20close%20to%20customers,
make%20their%20manufacturing%20more%20agile (accessed on 10 August 2020).
3.
Tanner Hartmann Incredible Prints: Can You 3D Print a Functional Kayak? Available online: https:
//all3dp.com/2/incredible-prints-can-you-3d-print-a-functional-kayak/(accessed on 25 September 2020).
4.
Lithgow, D.; Morrison, C.; Pexton, G.; Panarotto, M.; Müller, J.R.; Almefelt, L.; McLaren, A. Design
Automation for Customised and Large-Scale Additive Manufacturing: A Case Study on Custom Kayaks.
Proc. Des. Soc. Int. Conf. Eng. Des. 2019,1, 699–708. [CrossRef]
5.
Almefelt, L. Digitalized Large Scale Additive Manufacturing (DiLAM). Available online: https://www.
chalmers.se/en/projects/Pages/Digitalized-Large-Scale- Additive-Manufacturing- QDiLAMQ.aspx (accessed
on 10 August 2020).
6.
Digital Thread for Additive Manufacturing (DTAM). Available online: https://www2.deloitte.com/us/en/
pages/public-sector/articles/digital-thread.html (accessed on 10 August 2020).
7.
Siedlak, D.J.L.; Pinon, O.J.; Schlais, P.R.; Schmidt, T.M.; Mavris, D.N. A digital thread approach to support
manufacturing-influenced conceptual aircraft design. Res. Eng. Des. 2018,29, 285–308. [CrossRef]
8. Pollard, D.; Chuo, S.; Lee, B. Strategies for Mass Customization. J. Bus. Econ. Res. 2008,6, 7.
9.
Ahuja, B.; Karg, M.; Schmidt, M. Additive Manufacturing in Production: Challenges and Opportunities;
Helvajian, H., Piqu
é
, A., Wegener, M., Gu, B., Eds.; International Society for Optics and Photonics:
San Francisco, CA, USA, 2015; p. 935304.
10.
Mourtzis, D.; Doukas, M. Design and Planning of Manufacturing Networks for Mass Customisation and
Personalisation: Challenges and Outlook. Procedia CIRP 2014,19, 1–13. [CrossRef]
Systems 2020,8, 43 16 of 17
11.
Guo, N.; Leu, M.C. Additive manufacturing: Technology, applications and research needs. Front. Mech. Eng.
2013,8, 215–243. [CrossRef]
12.
Bonnard, R.; Hascoët, J.-Y.; Mognol, P.; Zancul, E.; Alvares, A.J. Hierarchical object-oriented model (HOOM)
for additive manufacturing digital thread. J. Manuf. Syst. 2019,50, 36–52. [CrossRef]
13.
Bonnard, R.; Hascoët, J.-Y.; Mognol, P.; Stroud, I. STEP-NC digital thread for additive manufacturing:
Data model, implementation and validation. Int. J. Comput. Integr. Manuf. 2018,31, 1141–1160. [CrossRef]
14.
Kim, D.B.; Witherell, P.; Lipman, R.; Feng, S.C. Streamlining the additive manufacturing digital spectrum:
A systems approach. Addit. Manuf. 2015,5, 20–30. [CrossRef]
15.
Bales, R.R.; Maull, R.S.; Radnor, Z. The development of supply chain management within the aerospace
manufacturing sector. Supply Chain Manag. Int. J. 2004,9, 250–255. [CrossRef]
16.
Moncenco, D. Supply chain features of the aerospace industry particular case airbus and boeing. Sci. Bull.
-Econ. Sci. 2015,14, 17–25.
17.
Fulton, M.; Hon, B. Managing advanced manufacturing technology (AMT) implementation in manufacturing
SMEs. Int. J. Product. Perform. Manag. 2010,59, 351–371. [CrossRef]
18. Singh, V.; Willcox, K.E. Engineering Design with Digital Thread. AIAA J. 2018,56, 4515–4528. [CrossRef]
19.
Mies, D.; Marsden, W.; Warde, S. Overview of Additive Manufacturing Informatics: “A Digital Thread”.
Integr. Mater. Manuf. Innov. 2016,5, 114–142. [CrossRef]
20.
Collecting Production and Process Data. Available online: http://www.leanmanufacture.net/operations/
productiondata.aspx (accessed on 8 August 2020).
21.
Attaran, M. The rise of 3-D printing: The advantages of additive manufacturing over traditional
manufacturing. Bus. Horiz. 2017,60, 677–688. [CrossRef]
22.
Bonnard, R.; Hascoët, J.-Y.; Mognol, P. Data model for additive manufacturing digital thread: State of the art
and perspectives. Int. J. Comput. Integr. Manuf. 2019,32, 1170–1191. [CrossRef]
23.
Kwon, S.; Monnier, L.V.; Barbau, R.; Bernstein, W.Z. Enriching standards-based digital thread by fusing
as-designed and as-inspected data using knowledge graphs. Adv. Eng. Inform.
2020
,46, 101102. [CrossRef]
24.
Sturm, L.D.; Williams, C.B.; Camelio, J.A.; White, J.; Parker, R. Cyber-physical vulnerabilities in additive
manufacturing systems: A case study attack on the.STL file with human subjects. J. Manuf. Syst.
2017
,
44, 154–164. [CrossRef]
25.
Xiao, J.; Anwer, N.; Durupt, A.; Duigou, J.L.; Eynard, B. Correction to: Information exchange standards for
design, tolerancing and Additive Manufacturing: A research review. Int. J. Interact. Des. Manuf. IJIDeM
2018,12, 767–768. [CrossRef]
26. Lynn Shostack, G. How to Design a Service. Eur. J. Mark. 1982,16, 49–63. [CrossRef]
27. Sauerwein, E.; Bailom, F.; Matzler, K.; Hinterhuber, H.H. The Kano model: How to delight your customers.
In Proceedings of the Nineth International Working Seminar on Production Economics, Igls/Innsbruck,
Austria, 19–23 February 1996; pp. 313–327.
28. GreenDelta, openLCA. Available online: https://www.openlca.org/(accessed on 25 September 2020).
29.
Mandolla, C.; Petruzzelli, A.M.; Percoco, G.; Urbinati, A. Building a digital twin for additive manufacturing
through the exploitation of blockchain: A case analysis of the aircraft industry. Comput. Ind.
2019
,
109, 134–152. [CrossRef]
30.
Liu, C.; Le Roux, L.; Körner, C.; Tabaste, O.; Lacan, F.; Bigot, S. Digital Twin-enabled Collaborative Data
Management for Metal Additive Manufacturing Systems. J. Manuf. Syst.
2020
, S0278612520300741. [CrossRef]
31.
Demartini, M.; Galluccio, F.; Mattis, P.; Abusohyon, I.; Lepratti, R.; Tonelli, F. Closed-Loop Manufacturing
for Aerospace Industry: An Integrated PLM-MOM Solution to Support the Wing Box Assembly Process.
In Advances in Production Management Systems towards Smart Production Management Systems; Ameri, F.,
Stecke, K.E., von Cieminski, G., Kiritsis, D., Eds.; IFIP Advances in Information and Communication
Technology; Springer International Publishing: Cham, Switzerland, 2019; Volume 567, pp. 423–430.
ISBN 978-3-030-29995-8.
32.
Lu, Y.; Choi, S.; Witherell, P. Towards an Integrated Data Schema Design for Additive Manufacturing:
Conceptual Modeling. In Proceedings of the Volume 1A: 35th Computers and Information in Engineering
Conference; American Society of Mechanical Engineers: Boston, MA, USA, 2015; p. V01AT02A032.
Systems 2020,8, 43 17 of 17
33.
Hedberg, T.; Feeney, A.B.; Helu, M.; Camelio, J.A. Toward a Lifecycle Information Framework and Technology
in Manufacturing. J. Comput. Inf. Sci. Eng. 2017,17, 021010. [CrossRef] [PubMed]
34.
Avison, D.E.; Lau, F.; Myers, M.D.; Nielsen, P.A. Action research. Commun. ACM
1999
,42, 94–97. [CrossRef]
Publisher’s Note:
MDPI stays neutral with regard to jurisdictional claims in published maps and institutional
affiliations.
©
2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
Content uploaded by Massimo Panarotto
Author content
All content in this area was uploaded by Massimo Panarotto on Dec 12, 2020
Content may be subject to copyright.
