See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/336109478

Transitioning from technical 2D drawings to 3D models: a case study at

defense systems

Article in INCOSE International Symposium · July 2019

DOI: 10.1002/j.2334-5837.2019.00641.x

CITATIONS READS

2 1,451

3 authors, including:

Satyanarayana Kokkula G. Muller

University of South-Eastern Norway University of South-Eastern Norway
34 PUBLICATIONS 101 CITATIONS 253 PUBLICATIONS 1,943 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Satyanarayana Kokkula on 11 February 2020.

The user has requested enhancement of the downloaded file.

Transitioning from technical 2D drawings to 3D models:
a case study at defense systems
Martin Sandberg Satyanarayana Kokkula
Kongsberg Defence & Aerospace University of South-Eastern Norway
[email protected] [email protected]

Gerrit Muller
University of South-Eastern Norway
[email protected]
Copyright © 2019 by Martin Sandberg. Permission granted to INCOSE to publish and use.

Abstract. Companies in the defense and aerospace industry are experiencing long time-to-market
and high costs related to the development of new systems. Companies use traditional design elements
such as technical 2D drawings and manual work along the development chain. Owing to today’s
software capabilities, a change to more automated processes is possible, where the use of 3D models
as governing documents through the entire development process holds a natural part.
This paper studies a 3D model-based definition (MBD) approach to the design development work at
Kongsberg Defence & Aerospace. Using a pilot project, we researched the current design processes
and the effect an MBD approach would have. We found that an MBD approach would reduce the
time-to-market and cost, since it eliminates the need for a technical 2D drawing and streamlines the
development work. In addition, it reduces project risk by enabling early validation of the design.

Introduction
Domain. This paper focuses on the defense and aerospace industry. Several system classes exist
within this industry, ranging from small volume and costly systems such as submarines and fighter
jets that require multiple years of development and manufacturing, to the comparatively less com-
plex, cheaper, and larger volume systems such as remote weapon stations (RWS). This study will
focus on the design development processes related to RWS systems. An RWS is a platform-mounted
system that can be installed on vehicles, ships, and man-portable stand-alone towers, see Figure 1. It
allows the operator to use and control the weapon from a safe environment. The system can operate
light, medium, and heavy machine guns. Customers ordering RWS systems use a variety of vehicles
and weapons, which requires flexibility in the design. The industry seeks design processes that re-
duce time-to-market and cost, and to be more agile and effective. The industry sees automated design
and manufacturing as a necessary future direction to achieve this.
Company. Kongsberg Defence & Aerospace (KDA) is a subdivision of the technology group
KONGSBERG. KDA is a well-known supplier within the defense domain. KDA develops systems
for command and control, surveillance, communication, remote weapon stations, missiles, and
composite structures. The product portfolio is comprehensive, where the area of use ranges from
underwater to surface, land, and air to space. The division that is involved with remote weapon sta-
tions is the focus of this study. KDA has delivered more than 18,000 RWS systems to 18 different
nations and is a world-leading supplier of such systems. At the end of 2017, KDA had 2421 em-
ployees and the annual revenue was MNOK 6333 (KONGSBERG, 2017).
 Figure 1. A remote weapon station alone (I) and mounted on top of a vehicle (II).
Problem. The competition to win contracts within the defense industry is fierce. Several companies
compete in the same domain, and the decisive factors are often time-to-market and price. As of today,
KDA is designing RWSs using SolidWorks 3D modelling software. Engineers perform the design
process in two steps. They first create a 3D model. Then they translate the 3D model into technical
2D drawings. The process of handing over the technical 2D drawings is ineffective in terms of both
time-to-market and cost, according to the management and engineers at KDA. The time to produce
and approve technical 2D drawings can range from one day to one week, depending on the com-
plexity of the part. The 2D drawings must completely represent the 3D physical model in order to
provide the manufacturer complete information on the component. Overall, this requires multiple
views, sections, and details.
This translation from 3D to 2D becomes even more costly as concurrent or subsequent changes to the
3D model require manual updates to the 2D drawings as well. If one should use a fully digital design
tool, where the 3D model is the governing element, this extra manual step does not exist. The focus
would be solely on the 3D model. This should also eliminate the errors that occur when the 3D and
2D models are not properly synchronized.
The existing unified CAD format, standard for exchange of product model data (STEP), is too simple
and cannot transfer design properties that are essential for the manufacturing process from the
original 3D model. These properties are required in the workflow and a richer format of the 3D model
is therefore necessary.
Goal. KDA wants to be more competitive by reducing the time and cost of developing new systems.
This study aims to formulate an MBD methodology. An effective agile design development process
would reduce the time-to-market and cost of their systems, hence improving their overall competi-
tiveness. To realize this, we propose a system that will streamline the overall design process, elim-
inate manual duplication of work, and enable automated design validation.
Solution. The proposed solution uses an MBD approach as its foundation, where the technical
documentation handover resides in the 3D model. This is sensible since the 3D model is the artifact
that the engineers produce first. We study the current situation and examine what material the me-
chanical engineers should include in the handover.
Research questions. In this study, we ask the following questions:
- What competences and tools do KDA and its mechanical manufacturing suppliers need to use
the MBD technique?
- What investments does KDA need to make for the transition to an MBD approach?
- What cost and time benefits may KDA expect from an MBD approach?
We address the stated questions by researching the current design processes and developing a new
outline of the design process in accordance with MBD. We interview internal and external stake-
holders to analyze their design-process needs. We study drawbacks with the current processes and
identify time and cost drivers. By using system modelling and user interviews, we evaluate the new
proposed process.
Case. The RWS sub-assembly that we use for this study is the azimuth manual release (Figure 2). Its
function is to enable the operator to manually rotate the RWS along the azimuth axis when the servo
system is disabled.

Figure 2. The sub-assembly related to the case.

Research method
Systems engineering domain. We used the methodologies of systems thinking to analyze the im-
plementation process. Systems thinking is a preferable method to use when performing deep analysis
and focuses on several aspects of the system; the big picture view, dynamics, soft factors, patterns,
and life-cycle (Walden et al., 2015). We included several stakeholders in the study to analyze the
majority of the value chain. This helped us to keep the big picture view and to have several aspects
and scenarios in focus during the project.

Figure 3. The research method of the project.

Initial discussions. We held baseline conversations to understand the context and to present our
approach to the employees at KDA. The conversation was of a door-to-door nature and considered as
informal.
Interviews. We performed interviews and presentations with both internal and external stakeholders
to understand the problem, current processes, and needs. This included field visits to mechanical
manufacturers and RWS assembly facilities. We conducted interviews during the entire time span of
the project to target specific elements and to obtain information when the engineers were both new to
and accustomed to the MBD approach.
Discussions and reviews. We included process reviews with the employees to enable feedback on
the conducted work. The process reviews functioned as a means of evaluation and validation of the
theoretical processes we formulated.
MBD preparation. We organized an MBD preparation phase to ensure that all necessary software
programs and licenses were present. The preparation made the execution phase effective and enabled
us to focus solely on the pilot project.
Pilot project. To narrow the scope and to be able to fully understand and analyze the MBD approach,
we selected, together with KDA, a sub-assembly of an RWS to function as a pilot project. The me-
chanical design process is the same for all components and assemblies, so our findings should be
relevant for designing a full RWS system.
Analysis. We analyzed the pilot project progress and result. We based the analysis of the project
work on feedback from stakeholders.
Questionnaire. The engineers contributed feedback by answering a questionnaire. The respondents
attended a presentation dealing with the result of the pilot project before answering the questionnaire.
The presentation highlighted the new process with fully defined 3D models using the MBD ap-
proach. The questionnaire included 10 questions, with both multiple-choice answers (using the Likert
scale) and open questions. The Likert scale is an ordinal level of measurement where the response
types have a ranking order. It is not suitable to analyze in terms of mean and standard deviation
(Jamieson, 2004). This is because we cannot consider the intervals between the response type groups
as equal. Therefore, we use net promoter score (NPS) and overall impression when analyzing the
result. NPS is a strict yet effective way to measure loyalty, future actions, and behavior of a group
(Reichheld, 2003). An NPS > 0 is per definition good. An NPS ≤ 0 indicates that the overall satis-
faction level is low. In these cases, the System Engineer should dig deeper to understand the un-
derlying factors. The calculation includes subtraction of the neutral and disagreeing responses (de-
tractors) from the positive, strongly agreeing responses (promoters).
The distribution between multiple choice and open questions allows for a broader collection of data,
ranging from an informal, free format to a formal, standardized format (Muller, 2013).
Evaluation. We used interviews and questionnaires as a mean to evaluate the project. The evaluation
incorporates all parts of the project including the project work, academia and research related theo-
ries, and the stakeholders’ impressions. We did this to obtain a bird’s eye perspective to cover as
much as possible and to be able to pinpoint specific critical elements.

State-of-the-art
The digital era and automation. Industry 4.0 is a German initiative aimed at implementing auto-
mation and data transfer to the manufacturing industry; to create smart factories. The term Industry
4.0 is a hot topic nowadays, and a framework that deals with the shift that the industry must make to
be able to meet future demands (Lasi et al., 2014). The framework involves increasing flexibility,
reducing the time spent on development stages (time-to-market), and including the ecological and
environmental aspects of the processes (not solely focusing on economics).
Why is the design model important? The first aspect to consider is cost. The new generation of design
models allows full automation, and one can assume that this shift implies large investment costs. The
implementation process does indeed cost money, but the management and decision makers in sev-
eral companies that have performed it witness that it is money well spent (Widman et al., 2016). This
is because digitization enables strategic change. A digital model gives users the possibility of en-
gaging in projects and competing for orders that previously were out of bounds. New business op-
portunities can arise following such an implementation because the company becomes more flexible
and dynamic.
Lundqvist et al. (2016) states that digital tools can reduce time to market significantly since they
could incorporate the validation and review process as an integrated step linked to the concept
modelling process. All employees would have access to the updated models and all related infor-
mation at any given time. In addition, the simulations and calculations could take place at an early
phase since realistic and fully defined 3D models are present early in the process. There is also room
for extra simulation iterations, ensuring quality and performance. Digital design processes enable
quick and easy sharing of information and technology platforms between different departments
within companies. Marion et al. (2015) claims that this would, in theory, reduce both cost and time
spent; a statement indicating that there is much to gain if one can achieve interfaces between the
design tools and the IT system.
Value of systems engineering. The development of systems has become far more complicated than
it previously was. Increasing changes and complexity that today hold a natural presence in system
development, demand structured work and a holistic mindset from the systems engineer. Systems
engineering helps to reduce risk related to the development of new complex systems. This is im-
portant since risk is often associated with cost. Walden et al. (2015) claims that there is a big dif-
ference between life cycle cost and committed costs (Figure 4). The figure illustrates that when the
engineers have completed the design phase, which averages 15% of the total life cycle costs, 85% of
the costs are already committed. This highlights the importance of a robust and comprehensive ap-
proach when developing new systems. Figure 4 also highlights that early verification and validation
of designs are key success factors when it comes to managing cost. To detect and adjust errors early
in the design process is crucial since the cost to extract defects later in the process increases signif-
icantly.

Figure 4. Committed life cycle cost against time (Walden et al., 2015).
Implementation of design tools. To increase performance related to product quality, reduced cost,
and shorter time-to-market, research suggests that companies should focus on and improve the de-
sign-related processes positioned early in the development process (Lehto et al., 2011). In general,
there is a need for a full implementation of a design tool if it shall contribute in a positive manner
(Booker, 2012). However, partial and small-scale implementation provides insight of the concept,
and whether one should continue with a big-scale implementation process or not. This induces the
urgency of a comprehensive analysis and preparation for the implementation to receive a positive
result.
To enable automated processing, steps such as feeding data to machining and measurement machines
using the technology and standards available today, the format must be rich. In these cases, the ex-
isting standard format, STEP files, would not be enough. To utilize the power of automation, these
applications are dependent on the active use of fully defined 3D models in the receiver domain.
 System context: current state in the company
Stakeholders. As a first effort to understand the system of interest and its current processes, we
formulated a stakeholder diagram (Figure 5). The stakeholder diagram contains both internal and
external stakeholders and a rationale of how they are affected and/or how they affect the system. The
focus of our study was on the mechanical engineering processes, but to reach the top objective of our
research, improving KDAs competitive ability, it was necessary to include key actors and their needs.
The MBD approach is a comprehensive method and requires all involved departments to work to-
gether to be effective.
X Influence
MIL-STD
Engineering Regulators
ISO
Influence
Influence
Design processes
-How to work (2D vs.3D) Management Cost and time-to-
Quality market reduction
Assurance
X
Influence
Influence
Component properties Customers
Documentation
Operational
-Material Purchasers requirements
-Tolerances
X
-Dimensions & geometry Documentation Influence
-Surface treatment Mechanical Personnel Material for
manufacturers procedure creation

Assembly
X
personnel
Goods
External Internal
reception Stakeholders Stakeholders

Figure 5. Stakeholders of the design and development process and their influence with the system.
Generate
project Material is selected Rewiev Rewiev
execution plan of 3D of 2D
Production model drawing
method is
Outlines Perform Develop 3D- determined
Systems Engineering Perform high Facilitates
system component model Purchasing
department department level design Generate 2D supplier
specification design specification department
drawing competition

Technical design
Choose supplier
spesifications
Geometrical tolerances uploaded to database
are detemined

Existing
Technical
Recives 3D- Sends 3D
Preliminary Engineering Makes price discussions Purchasing
PDR design
estimate between supplier
model and 2D Supplier model and 2D
department drawing drawing department
and KDA

New
Purchasing
department Engineering
Performs formal Perform
Purchasing
supplier audit quality
department review inspection
Recives price
Quality
estimate
Assurance

Issue PO

Test release
Approves order Generates Perfoms
of engineering Supplier Produces parts measurement engineering
units reports tests

Engineering
Approves serie
Purchasing DVT/ev PVT CDR / Design TDP Inspects
production Supplier Produces parts
@ customer freeze product quality
FAI
department sample
Quality
Assurance
Supplier

Product is
Assembles the Pefoms receiving Goods
shipped to
product
Assembly Store parts
Inspection
Ships parts
customer reception

Figure 6. Process flow diagram describing the design and development for mechanical components.
Process flow. We mapped the current design and development of mechanical components processes
at KDA (Figure 6). This increased our understanding of the routing and what steps are dependent
upon each other as it stands today. As seen in the figure, the design process from order to delivery is
comprehensive and includes several actors and elements. The process flow diagram was an iterative
and living model, and we performed regular reviews of it together with employees at KDA. For the
purpose of this study, our focus was on the actions taking place inside the dotted rectangle on the top
of the figure.
Technical information flow. We created a technical information flow diagram (Figure 7) to analyze
and understand what kind of technical documentation the different stakeholders use today. The
technical information flow illustrates that the technical 2D drawings have a limited area of use; en-
gineers create them, and the operational purchasers include them in the purchase orders to mechan-
ical manufacturers. The mechanical manufacturers are dependent on receiving detailed design
properties, and the only component able to illustrate this in the current workflow is the 2D drawing.
For the rest of the workflow, less rich formats, such as assembly drawings and the current capacity of
the 3D models are adequate.

Assembly Assembled
3D model system
Work instructions
Bill of materials
System Assy drawing
specification Quality
Assy drawing Assurance
Material specification, test result &
Engineering procedure information
Technical design discussions
Mechanical
manufacturers
Bill of materials
3D model
Assy drawing
3D model
2D drawing
Assy drawing 2D drawing
Operational
purchasers
Purchasing Order

Figure 7. KDA’s current technical information flow in industrial practice.

Work instructions. Today KDA uses photographs in combination with text to illustrate the assem-
bly sequence to their operators. According to KDA, this can cause misinterpretation since people
tend to understand text in different ways across the globe. An assembly operator in Norway could
understand the work instruction differently from an assembly operator stationed in the United States.
It is difficult to formulate text that is specific enough and unable to be misinterpreted in a local
language or in the English language.
Re-iteration of the problem. KDA sees challenges and drawbacks with the current design process.
The engineers are actively using 3D models when performing mechanical design. However, the
handover artifact is a technical 2D drawing. We have formulated a new MBD process that would
reduce non-value adding and time-consuming work.

Process review
The digital trend that we see today is here to stay, and if KDA wants to achieve a competitive position
in the future, it is necessary to evolve with the trends. Automation is where development processes
are now moving. One step on KDA’s road toward automation and a digital work environment is the
implementation of MBD. The objective of MBD is to reduce the time spent on design processes by
shifting from traditional 2D drawings and going completely digital instead. The idea is that the 3D
model is leading in every step of the design and manufacturing process (Cicconi et al., 2017). The
engineers detail the design by adding annotations in the form of dimensions, tolerances and other
remarks to the 3D model. This allows the 3D model to structure all the data regarding the properties
of the components and store them in one location. This is a big step forward compared to the current
STEP format, since STEP files suffers from missing information.
Using MBD, the engineers can export the annotated 3D model as a native file, a 3D PDF. A portable
document format (PDF) is a file format used for presenting documents that include text and images.
The format is independent of software, hardware, and operating systems (Adobe, 2005). SolidWorks,
which is the 3D modeling tool that KDA uses, has its own version of this as well, called an eDraw-
ings file. Engineers can choose to generate either the SolidWorks standard eDrawings file or a 3D
PDF. The low level of richness of the two is the same, i.e. receivers cannot change the content, as
would be the case for source formats. Receivers can add comments in the form of text and lines that
works as feedback to the constructors. In general, rich files belong in the engineering domain (de-
signers, constructors) to ensure the quality of the design process and to control traceability and
changes. Frozen files belong in the receiver domain (suppliers, mechanical manufacturers) where it is
not possible to make actual changes to the models.
A native 3D PDF offers an insurance for the future compatibility of the information. PDFs are
standard and not directly linked to a modeling software, which makes the users soft-
ware-independent. One can argue that an eDrawings file adds value if SolidWorks is the comput-
er-aided design software used by a company. On the other hand, it can disable the incorporation of
other design tools that could serve as useful, since other tools could be unable to communicate with
or open the eDrawings file.
These files can be accessed using desktop computers, iPads, or tablets, providing great flexibility
(SolidWorks, 2017). Engineers and manufacturers can examine the model without the use of the 3D
modeling software, since the eDrawings and 3D PDF files enable the possibility to rotate, show/hide
specific components, and split views. The engineers can link other technical information that is re-
quired for manufacturing, support, and life-cycle management to these files (Alemanni et al., 2011).
Alemanni et al (2011) states that, if all customers and suppliers work with these files, then it would
eliminate the need for translations between 3D and 2D, and the usual paperwork. This lets engineers
fully focus on the design and development of new systems.
Released 3D PDF
Discussions
uploaded to
Rewiev of Using with
database
3D Model stakeholders
& 3D PDF
Using

Early validation
Assign material Using
of the design
Fully defined
Perform Perform Develop a
MBD 3D model Released MBD 3D
high level component fully defined Purchasing order to
& 3D PDF model uploaded to
design design 3D-model Using manufacturing
Determine production database
supplier
method

Using
Assign datums, geometrical
tolerances, fittings, and surface Creation of
treatment work
instructions

Figure 8. The envisioned mechanical development process.

Engineers can attach file notations and comments directly to a visualization of the model by using
SolidWorks eDrawings or Acrobat Reader, This emphasizes one of the benefits that was exclusively
provided by the 2D drawings in the past (SolidWorks, 2017). Engineers can perform design reviews
and validation at an earlier stage with the 3D PDF file as a basis, since this format allows the ex-
amination of views, sections, and needed components separately. This is an important aspect since
errors that occur early in the design process are easier to fix, and the cost associated with this is
significantly lower in comparison to fixing them in later phases of the development process (Walden
et al., 2015).
The MBD approach is beneficial since it solves several of the drawbacks that standard 2D drawing
process is struggling with. When creating technical 2D drawings for detailed and complex parts,
engineers spend most of their time making sure every feature is fully dimensioned and every view fit
on the sheet. MBD solves this automatically since this approach allows engineers to assign all tech-
nical information straight to the 3D model. The engineers’ functional idea of the product can be
difficult to express accurately on a piece of paper. This can be problematic when the product is in the
workshop of the manufacturer. The inclusion of MBD files in the purchase order would help the
mechanical manufacturers in their work. KDA and their mechanical manufacturers would be able to
hold technical discussions related to the component and the detailed design at an earlier stage. The
mechanical manufacturers are experts within their field and would be a great help to the engineers at
KDA with their insights on different design solutions that are connected to lower manufacturing
costs.
Active use of the 3D models when creating work instructions could reduce the misinterpretation that
text brings. Figures and models function as a universal language that is easy to understand. Coloring,
highlighting, and hiding specific components of the assembly is simple to do in 3D modeling soft-
ware. This would make the work instruction robust and understandable by almost anyone.
The proposed process enables early review and validation of the design and includes a fully defined
3D model in purchase orders. Further, it allows technical discussions with stakeholders without the
use of a 3D modeling software and emphasizing automation. These actions are taking place at an
early stage of the design process, which according to Letho (2011), is where companies should put
their improvement focus to increase the performance connected to quality, cost, and time-to-market.

Results
Technical documentation. We created new 3D models according to the MBD approach. Figure 9
illustrates the models related to the pilot project. The figure contains the azimuth manual release full
assembly (III), an exploded view of the full assembly (V), and the base plate with tolerances, datums,
and other technical information important to the design, assigned to the model (IV).

FULL MODEL

III EXPLODED VIEW

MBD MODEL

IV V

Figure 9. Illustration of the pilot project parts.

We created a template and generated a 3D PDF (Figure 10) of the base plate. The purpose of the 3D
PDF is so that stakeholders who do not possess a SolidWorks license can examine the model. The 3D
PDF includes the complete model with the important tolerances, technical manufacturing require-
ments, and a section where the users can submit comments. We created two pre-defined views of the
component (top view and side view) that illustrate the most critical tolerances related to the model
separately. The viewers can also rotate the model as desired to perform detailed examination. We
generated the model and 3D PDF from the SolidWorks 2015 edition. Newer versions of the software
include additional functionality in terms of traceability and revisions. This would make the solution
more robust in terms of quality.

Figure 10. The 3D PDF generated by SolidWorks.

Mechanical manufacturers. We found that the exclusion of the technical 2D drawings would not
generate major changes in the development process. During our interviews with the mechanical
manufacturers, we discovered that they have limited use for the technical 2D drawings. They are
ready to use the 3D model actively when programming their CNC machines. When the mechanical
manufacturers later perform the component measuring process, the measuring machine is ready to
use the MBD 3D models as a reference and does not require any manual input of reference data. The
only occasion where they see a need for a technical 2D drawing today is when the component has low
complexity and they manufacture it using a manual lathe. This finding points out that the mechanical
manufacturers are, in general, ready for a transition to the MBD approach, and they possess the
needed competencies. The mechanical manufacturers also stated that a fully defined 3D model would
increase their understanding of the components.
When performing the research, we found that the STEP file is unable to fully represent tolerances. A
STEP file is only able to understand and represent nominal tolerances. This creates issues when the
mechanical manufacturers are to use the STEP file for the measurement process. In addition, it is
impractical since the conversion of model formats between different CAD software is incapable of
capturing specific model properties. If KDA and its mechanical manufacturers apply the automated
MBD approach, then there is a need of a full SolidWorks version of the 3D model. This requires that
the mechanical manufacturers use and have competence in SolidWorks as their CAD tool.
Work instructions. During our post-interviews with the assembly personnel, we found that they
were positive toward a solution that emphasizes a digital format. Using the MBD approach, they
would get a head start since MBD would give them fully defined 3D models at an earlier stage. MBD
makes it possible to use these 3D models when creating documentation with the help of SolidWorks
Composer, instead of photographs in combination with text that is the standard today. SolidWorks
Composer automatically synchronizes the documentation so that it corresponds with the latest revi-
sion in the case of a design change. Furthermore, the assembly personnel stated that work instructions
completely built up by the 3D model would be a step forward in terms of providing an understanding
of the system. The responsible personnel can assign different colors to the subcomponents of the
assembly; clearly illustrating which part is the subject in the different assembly steps. This coloring
maneuver, in combination with hiding irrelevant parts of the assembly, would increase the effec-
tiveness of the work instructions and reduce the amount of text required. The fact that MBD makes
the work instructions easier to understand reduces the risk that failures occur in production.
Engineering evaluation. Figure 11 illustrates the result of the questionnaire that the engineers at
KDA responded to. We used a 5-point Likert scale to extract promoters and detractors, which ena-
bled us to analyze the result using NPS. To consider a statement as good, the NPS must be > 0.

Figure 11. Questionnaire related to the MBD process.

The outcome of the questionnaire demonstrates that the engineers have mixed feelings and opinions
regarding the MBD method. A positive aspect is the low portion of respondents answering “disagree”
and “strongly disagree”. This indicates that the engineers are not strongly and fundamentally against
the implementation of the MBD method. However, such a top-level analysis of the result is not ad-
equate on its own. The NPS tells another story. There are only three of the six questions that have
NPS > 0. This result points toward the need for additional efforts from KDA’s side to be able to
implement the MBD method in an effective and structured manner. NPS suggests that the engineers
only fully agree with two of the statements:

 MBD will reduce the time spent on a component.

 MBD would facilitate early validation of the components.
The statement relating to early detections of errors has an NPS=0, which is neither good nor bad.
However, the respondents did only just select “strongly agree”, “agree”, and “neutral” as responses,
which indicates that none of them have disagreeing opinions related to the statement. Overall, the
proportion of “neutral” opinions is rather high. We expect that the reason behind this is that the en-
gineers do not feel comfortable enough to have a strong opinion towards a method they have not seen
in full-scale operation.
When asked to express their opinions regarding the benefits (Table 1a) and concerns (Table 1b) of
using the method, the engineers highlighted several important comments.
Table 1a. The engineers’ perspective regarding benefits of the MBD method.

Aspect Benefits
- Illustrates components in an orderly and understandable manner to manu-
facturers and customers.
Illustration and - Provides better overview and understanding of the assemblies in the review
overview process using 3D models. This should reduce the sources of errors related to
design.
- Provides a better understanding of the design among the engineering team. A
3D model can tell a greater story related to design decisions, which could
boost the control of the functionality of the components and assemblies.
- Facilitates for early validation, which should imply the discovery of possible
errors at an earlier stage.
Quality - Engineers can spend more of their time assuring a good design, instead of
creating documentation.
- Engineers can start conversations with the mechanical manufacturers at an
earlier stage, which can increase the manufacturability of the components.
- Reduces time related to create technical documentation. This is especially
good when dealing with complex designs.
Cost and time - Enables the operations and production departments to be involved early in the
process. They can use fully defined 3D models when creating work instruc-
tions.
Table 1b. The engineers’ concerns related to the MBD method.

Aspect Concerns
- There can emerge issues related to revision traceability when mechanical
Traceability manufacturers use fully defined 3D models.
and usability
- The exchange of 3D model content with manufacturers due to software de-
pendent models.
- How well deals the method with changes on the models, and how to visualize
them?
- Training and adapting to the new process could be time consuming.
Process
- There is a need for a test in a larger scale for it to validate the method.
- Fear of that the implementation is too big to handle in one sequence, and that
the use of the one step at a time principle is necessary.
- The method must be able to address relations between tolerances in assem-
Quality blies to avoid component collision.

A big benefit the engineers mentioned related to the method was that it brings a great overview and
understanding to the review process. This would reduce the sources of errors linked to the design.
This is a very important benefit, since Walden et al. (2015) highlight the big difference regarding cost
to extract defects early versus late in the design process. The engineers also claimed that the MBD
method would reduce time related to generate technical documentation, since it eliminates the need
for a technical 2D drawing. This would allow them to focus more of their time on the functionality of
the design. This is one of the top objectives of this study, and a cornerstone in the MBD philosophy.
Finally, they mention that the method would allow them to discuss the design with mechanical
manufacturers at an earlier time. As stated earlier, the mechanical manufacturers are experts within
their field of manufacturability and producibility. If the engineers could hear their opinions regarding
design decisions earlier and adjust parts of the design accordingly, it would have a great impact on
the time needed for manufacturing and the related cost.
The engineers expressed a need for a larger-scale test to validate and properly understand the method.
This statement goes hand in hand with Booker’s (2012) claim that a full implementation is necessary
to validate the method regarding benefit and value. Further, they mentioned concerns related to the
revision management of 3D models. To mitigate this, the engineers must use the revision feature
included in SolidWorks. This would help them in terms of keeping track and assuring the use of
correct models. Finally, they mentioned an opinion that often arises when dealing with the imple-
mentation of new processes: fear of the new processes being too comprehensive and effort con-
suming to implement. KDA can perform actions to reduce this concern. A thorough presentation and
illustration of the new process applied to a full RWS system would serve as useful to increase the
engineers’ understanding of the concept further.
When asked about alternative solutions, the engineers mentioned STEP models combined with
technical 2D drawings and the current process. We discuss the use of STEP models earlier in this
paper. The main drawback with this solution is that the format is not rich enough. A STEP model is
unable to represent tolerances and other design properties, thus eliminating the possibility for au-
tomated processes and independence from technical 2D drawings. We performed this study because
the current process is ineffective, hence ruling out that option as a step forward.
Cost estimate. KDA needs several SolidWorks MBD licenses to start using the MBD approach. This
is a cost they must incur to be MBD compatible. MBD is a plug-in to SolidWorks and a license costs
NOK 19k plus an annual subscription fee of NOK 5k/year. In addition to the direct investment cost,
there is cost related to the organizational implementation process that KDA must perform. KDA
would reduce development time by eliminating the detailed 2D drawings; this would reduce cost as
well. The time that MBD saves is worth more than the direct costs it incurs since engineers could
spend their time working on other value-adding activities related to the design development.
Therefore, MBD would boost the effectiveness of their work.
Cost versus value. A question that is common when the implementation of a new process is up for
discussion is cost versus value. To be justifiable, a new process should add more value than it takes
effort to implement and maintain. Value is hard to measure since it often consists of both hard and
soft elements. One way to quantify value is to analyze in what degree the new process contributes to
reduced risks. The most prominent risks associated with the current design process at KDA are late
discovery of errors, mismatch between assemblies due to late validation, and mismatch of the tech-
nical 2D drawings and 3D models due to revision changes. We consider the occurrence of these
aspects as critical since they impact KDAs overall goal to reduce time-to-market for new systems.
Since the new proposed process is based on the same competence as is current today, the need for
courses and on-the-job training for the engineers is minimal. We expect that the cost related to the
new process, most of it consisting of SolidWorks MBD licenses, is minor compared to the added
value the new process will bring to KDA. This is based on the assertion that the MBD method will
reduce the engineering hours required to design new systems due to a more effective work process,
and that early validation of designs will reduce the late discovery of errors. With new working pro-
cesses comes a time of transition. However, the change to including dimensions, geometrical toler-
ances, and datums to the 3D model instead of to technical 2D drawings, would not be problematic to
perform.
Given this, we suggest that the value added is greater than the cost. We expect a clear benefit fol-
lowing an implementation of MBD.
Validity of the results. The MBD method is a comprehensive method to analyze and implement. A
complicating fact is that it touches most of the design chain, which means that it is dependent on
multiple stakeholders.
Alemanni et al. (2011) verifies our claim that an adoption of MBD by all parties in the design chain
would save the time-consuming work of translating 3D models into technical 2D drawings. This
reduces the workload of the engineers and let them solely focus on the design and development of
new systems; an aspect that the engineers requested. We confirm this by our process mapping and
evaluation with the engineers and other stakeholders to the design process.

Recommendations
In general, MBD complies with the predominant industry standards ISO, ASME, MIL, and DIN. We
recommend that KDA perform further studies with the aim of identifying an effective manner of
designing their technical documentation and the traceability of it to fulfill the requirements related to
the standards.
Our findings imply that the value the MBD method would bring to the company is greater than the
required effort to implement it. Widman et al. (2016) claim that an implementation of a digital design
tool such as the MBD method is an investment well spent. Nevertheless, without a full implementa-
tion of the method and an analysis of it, we cannot validate such claims. Therefore, we see a need for
a large-scale study, performed on a whole RWS system, to validate our results. An implementation of
a process with this magnitude is complex with shifting elements, and as Booker’s (2012) claims,
there is a need for a full implementation for proper validation. Factors that do not affect small-scale
projects can build momentum and increase regarding influence when transferred into a larger con-
text.
Finally, we recommend KDA to create an architecture for its tool chain and processes. This archi-
tecture should establish a basis regarding which formats KDA should use for the information trans-
fer.

Conclusion
The goal of this study was to analyze KDA and their suppliers regarding an implementation of MBD
in terms of what competences and tools they need, what investments they must make, and what time
and cost benefits they may expect. We did this by applying the MBD method to a sub-assembly of an
RWS. We identified stakeholders, required technical documentation, and studied the current design
and development for the mechanical components process to analyze cost and time-to-market drivers.
Our research indicates that the MBD method could serve usefully in terms of cost and time-to-market
reduction. The MBD method enables early validation and eliminates the need for technical 2D
drawings. Due to this, MBD adds additional value since it reduces project risk. The effective way of
work should reduce the engineering hours, as well; an aspect highly linked to cost. KDA would need
to invest in new MBD licenses to utilize the method.
The process enables partial automation since mechanical manufacturers can use the rich and fully
defined 3D model when programming CNC and measurement machines. However, working effi-
ciently requires a rich 3D model. This entails that the mechanical manufacturers possess a Solid-
Works license and competence in using the Software. The engineering domain required compe-
tences related to the MBD method are the same as the current process; the difference lies in when the
engineer applies the data to the 3D model. Our proposed solution will reduce time-to-market and cost
by streamlining KDAs overall design process, eliminating manual duplication of work, and enabling
automated manufacturing.
 Future research
There is a need for further analysis of what formats companies should include in the tool chain. The
MBD method makes it possible to fill the 3D models with data in a unified and reusable format;
elements that are of utmost importance of the design processes regarding time-to-market. The ex-
isting unified 3D model standard (STEP) does not incorporate rich data since the software transforms
the content to a solid, or surface, model losing the feature tree. The feature tree is a necessity for
creating a solid foundation for the implementation of MBD.
It is necessary to perform more research on how much time and cost savings engineering companies
could expect from the MBD method. Given the short timeframe of this project, we could not obtain
any hard data related to this. A full-scale case study performed over a longer period when data is
available should be able to identify these aspects.

Acknowledgements
We would like to thank KDA, especially Ola Jacob Iversen for giving us the opportunity to perform
the master thesis project with them.
Finally, we would like to thank Aleksander Kristoffersen, the industry Supervisor at KDA. Ale-
ksander served as an excellent help when it came to identify the current way of work and to discuss
the practical aspects of the case.

References
Adobe. 2005. PDF reference : Adobe portable document format, version 1.6. Adobe portable
document format, version 1.6. Peachpit Press.
Alemanni, M., Destefanis, F. & Vezzetti, E. 2011. Model-based definition design in the product
lifecycle management scenario. The International Journal of Advanced Manufacturing
Technology, 52, 1-14.
Booker, J. 2012. A survey-based methodology for prioritising the industrial implementation qualities
of design tools. Journal of Engineering Design, 23, 507-525.
Cicconi, P., Raffaeli, R. & Germani, M. 2017. An approach to support model based definition by
PMI annotations. Computer-Aided Design and Applications, 14, 526-534.
Jamieson, S. 2004. Likert scales: how to (ab)use them. Medical Education, 38, 1217-1218.
KONGSBERG. 2017. Annual report and sustainability report [Online]. www.kongsberg.com.
Available:
https://kongsberg.com/~/media/KOG/Files/Investor/Annual%20reports/2017/KOG-A%CC
%8Arsrapport-GB-interaktiv-web_ny.ashx?la=en [Accessed 16.04.2018].
Lasi, H., Fettke, P., Kemper, H.-G., Feld, T. & Hoffmann, M. 2014. Industry 4.0. The International
Journal of Wirtschaftsinformatik, 6, 239-242.
Lehto, J., Harkonen, J., Haapasalo, H., Belt, P., Mottonen, M. & Kuvaja, P. 2011. Benefits of DfX in
Requirements Engineering. Technology and Investment, Vol. 2 No. 1, 27-37.
Muller, G. 2013. Systems Engineering Research Methods. Procedia Computer Science, 16,
1092-1101.
Reichheld, F. F. 2003. The one number you need to grow. (the influence of customer
word-of-mouth)(Cover Story). Harvard Business Review, 81, 46.
Solidworks, D. S. 2017. SOLIDWORKS Technical Communications - Adding Innovation to your
technical communication deliverables [Online]. Available:
http://www.solidworks.com/sw/docs/3DS_2017_SWK_TechnicalCommunications_Datashe
et.pdf [Accessed 01.12.2017].
Walden, D. D., Roedler, G. J., Forsberg, K., Hamelin, R. D., Shortell, T. M. & INCOSE 2015.
Systems engineering handbook : a guide for system life cycle processes and activities,
Hoboken, N.J, Wiley.
 Widman, R., Hemphälä, J. & Ritzén, S. 2016. Investigating Model-Based Definition
Implementation: A Case Study on Cross-Functionality through Product Lifecycle
Management. KTH Royal Institute of Technology.

Biography
Martin Sandberg, born in Sweden, received his Master of Science in Systems
Engineering with Industrial Economics from the University of South-Eastern
Norway in 2018. He is currently employed as a Systems Engineer for Kongsberg
Defence & Aerospace in Kongsberg, Norway.

Gerrit Muller, originally from the Netherlands, received his Master’s degree in
physics from the University of Amsterdam in 1979. He worked from 1980 until
1997 at Philips Medical Systems as system architect, followed by two years at
ASML as manager systems engineering, returning to Philips (Research) in 1999.
Since 2003, he has worked as senior research fellow at the Embedded Systems
Institute in Eindhoven, focusing on developing system architecture methods and the
education of new system architects, receiving his doctorate in 2004. In January
2008, he became full professor of systems engineering at University of
South-Eastern Norway in Kongsberg, Norway. He continues to work as senior research fellow at the
Embedded Systems Innovations by TNO in Eindhoven in a part-time position.
All information (System Architecture articles, course material, curriculum vitae) can be found at:
Gaudí systems architecting http://www.gaudisite.nl/

Satyanarayana (Satya) Kokkula received his Master’s degree in Applied

Mechanics from IIT Delhi (Indian Institute of Technology, Delhi) in 2000. For one
year (2001-2002), he worked as an Assistant Systems Engineer at TATA Consul-
tancy Services Pune, India. In 2005, he received his PhD from the Norwegian
University of Science and Technology (NTNU), Trondheim, Norway. After fin-
ishing PhD, he joined FMC Kongsberg Subsea AS as a Specialist Engineer in
Structural Analysis from 2006 to 2016. In August 2017, he joined the University of
South-Eastern Norway as an Associate Professor of Systems Engineering.

View publication stats