Peer-review under responsibility of the scientific committee of the The 5th International Conference on Through-life Engineering Services TESConf 2016 doi: 10.1016/j.procir.2016.08.018
Trang 12212-8271 © 2016 The Authors Published by Elsevier B.V This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/)
Peer-review under responsibility of the scientific committee of the The 5th International Conference on Through-life Engineering Services (TESConf 2016) doi: 10.1016/j.procir.2016.08.018
Procedia CIRP 59 ( 2017 ) 202 – 207
ScienceDirect
The 5th International Conference on Through-life Engineering Services (TESConf 2016)
Improving Functional Product availability: software-related measures
planned and taken
a) ProcessIT Innovations R&D Centre, Luleå University of Technology, SE_97187 Luleå, Sweden
* Corresponding author Tel.: +46 920 491528 E-mail address: john.lindstrom@ltu.se
Abstract
The paper, based on an empirical study involving five companies, concerns software-related measures that are planned and taken by providers together with their customers to improve the availability of Functional Products (FP) or similar offers The manufacturing industry is showing
an increasing interest in adding offerings based on additionally complex business models, as opposed to merely offering products and services This supports innovation and helps companies to stay competitive and profitable Considerable focus is placed on performance- or result-based business models Functional Products (FP) is one such business model, where the provider offers a function to customers at an agreed-upon level of availability, productivity or efficiency FP comprise four main constituents: hardware, software, service-support system and management of operation, which together deliver value to customers on a long-term basis The paper highlights nine software-related availability measures planned and taken by manufacturing companies and proposes additional potential software-related availability measures
© 2016 The Authors Published by Elsevier B.V
Peer-review under responsibility of the Programme Committee of the 5th International Conference on Through-life Engineering Services
(TESConf 2016)
Keywords: Availability; Functional Products (FP); products; software
1 Introduction
The paper, based on an empirical study involving five
companies, concerns software-related measures that are
planned and taken by providers together with their customers
to improve the availability of Functional Products (FP) or
similar offers A current trend for manufacturing companies is
to incorporate service offers and soft parts into their regular
product offers and also to extend the providers’ ownership of
the product throughout the entire product lifecycle This is a
business opportunity for the provider, but it is also a
requirement from customers, since it allows them to focus on
their core business and processes There are a number of
additionally complex business models, compared to products
and services, such as performance- or result-oriented business
models, which may be used to stay innovative, competitive
and profitable over time in global competition One example
of such a business model is FP, and the concept of FP [1-4],
incorporates hardware, software, service-support system and
management of operation into a combined effort providing a
function to customers with an agreed-upon level of availability
or improved productivity or efficiency Thus, the FP concept has a basis in the cyber-physical systems and industrial internet-of-things paradigms Throughout the FP lifecycle, operation of the FP must be managed, further developed and optimized, since the intent with FP is to optimize long-term value for both the customer and the provider, i.e., create a sustainable win-win situation [5-7] Further, the concept of FP has similarities with, for instance, Functional Sales (FS) [8], Extended Products [9], Total Care Product (TCP) [1], Product-Service System (PSS) and Industrial Product-Product-Service Systems (IPS2) [10, 11], Servicizing [12], Service Engineering [13] or Through-life Engineering Services (TES) [14] in the sense of increasing the focus on soft parts such as services, knowledge and know-how etc., additionally offered The FP, originating from hardware aspects, has most commonalities with
complexity development-wise FP availability can be seen as a function of reliability and maintainability [15] of the main constituents that are part of delivering the function
© 2016 The Authors Published by Elsevier B.V This is an open access article under the CC BY-NC-ND license
( http://creativecommons.org/licenses/by-nc-nd/4.0/ ).
Peer-review under responsibility of the scientifi c committee of the The 5th International Conference on Through-life Engineering Services (TESConf 2016)
Trang 2The FP lifecycle, whose contracts for customer instances
can range up to 30 years, contrasts significantly from offering
the same hardware and software as a product with services
Some of the significant differences are that the provider
retains ownership, takes on risks and responsibilities which
are transferred from the customer, and further co-creates value
together with the customer In addition, as the provider is
compensated for providing a function, the provider needs to
honour the agreed-upon level of availability or contract
parameter specified This requires that the provider can
monitor the function and foresee potential problems before
they occur and, preferably, act in a proactive manner rather
than a reactive one when a problem or breakdown is a fact
A trend is that the FP software constituent grows and
becomes additionally important as more functionality is
added, and it is commonly easier to add new functionality via
software than hardware-wise Traditionally, most
manufacturing companies have a hardware-centric approach
towards availability management, whereas the software field
offers new and interesting developments to improve
availability As the software constituent can be different than
the hardware one, by being situated on-board/locally,
distributed/federated, centrally or cloud-based, etc., some
additional consideration may be necessary in order to uphold
the desired level of availability versus the cost level accepted
Using central or cloud-based software commonly has an
economy-of-scale advantage as long as it is possible to
maintain the agreed-upon level of availability Depending on
what the software does and how it is possible to combine
on-board/local software with the other software options,
information security, connectivity, response times (i.e.,
latency), criticality as well as the coupling to other parts of the
FP limit where the software or parts of software need to
reside Thus, the use of cloud services or central software may
not be an option in many cases, although they are attractive
from an economical and scalability point of view Thus, it is
necessary to make an assessment of what software
functionality is reasonable and adequate to have
on-board/locally, distributed, centrally or in the cloud, so as to
avoid availability issues In addition, changes in, for instance,
EU and US data legislation require that management of the
overall FP must adhere and adapt to the appropriate data
legislation in cases where personal data is transferred
The current research on FP has mostly been directed
towards the hardware and service-support system by
modelling and simulation of these two constituents (and in
particular the reliability and maintainability) For instance,
Löfstrand et al [16-17] propose a simulation framework and
Reed et al [18-19] outline a language facilitating modelling of
the service-support system A high-level outline of the
challenges and need to model and simulate the whole FP has
been made by Pavasson et al [20] and FP availability,
legalities, information security, criticality and adequacy, when
planning to use cloud services in FP have been assessed by
Lindström et al [21] Currently, there is a lack of literature
addressing FP availability focusing, in particular, on the
software constituent Thus, the research question addressed in
this paper can be formulated as: which software-related
measures to improve FP availability are planned and taken by
FP providers? The purpose of the paper is to highlight a number of software-related availability measures which can be used by FP (or similar offers) providers together with their customers in the manufacturing industry
2 Related work
Many products or similar offers which include hardware are used a lot longer than initially anticipated due to economic reasons, and thus the lifecycle is sometimes extended far beyond what was originally planned for [22] This causes issues related to asset management and in particular how to manage the availability level expected or agreed upon An asset management problem is how to manage the obsolescence [23] of the hardware and, in particular, of the often rapidly evolving software (including operating systems and platforms used to run the software) Additive manufacturing may solve some of the asset and obsolescence management issues related
to hardware and spare parts, since it allows spares with the right quality and properties to be crafted on demand Further, Muñoz et al [24] posit that as software starts to become one
of the most valuable assets in the aerospace industry, the portfolio of (critical) software should be monitored performance-wise in order to minimize costs and avoid risks
It can be assumed that the same goes for many other industries
as well (including the manufacturing industry and FP)
Optimizing the level of availability versus cost is an ever-present concern, as is optimizing the FP lifecycle towards the duration of the customer contract and sustainable win-win situation between the provider and customer [5-6] Thus, the initial FP planning, design/development, realization and operation, are all subject to planning or taking measures regarding availability or improvement of it To avoid most of the expensive and long-term testing, simulations (preferably using data from actual monitoring of FP in operations or test activities) are a common tool to use either on a framework layer for the FP main constituents [16-17, 20, 25-27] or for important/vital components of FP [28-29] Up to now, most modelling and simulation efforts have targeted the hardware and service-support system constituents Guided by the simulations, the necessary actual testing can be performed and test cases mapped out based on simulation results in order to improve designs, system/component reliabilities as well as system/component maintainability Of great importance is to find or locate the potential large cost drivers (such as number
of service engineers, number of locations for spare parts, which spare parts to have in stock, driving distances to customer sites, etc.), and decide upon how to deal with them
in terms of improving reliability, maintainability or opting for re-design/re-thinking if deemed necessary [2, 25]
New ideas are needed in order to minimize availability issues originating from the software in FP An idea is a concept including morning gymnastics tests and digital envelopes proposed to minimize errors in automation software [30] This may complement the other (software-related) availability measures taken or planned for FP
In order to maintain the agreed-upon level of availability, planning and operation of FP require thorough risk management during the creation/delivery/capture of value for
Trang 3both the provider and customer sides [31] As the software
constituent of FP is expected to increase, an additional focus
on software related risks will be necessary
In addition, a number of long-term related aspects for FP
need to be considered to minimize the risk level and uphold or
improve the availability level for FP offers [32]
3 Research approach
The research approach employed in this study has been
based on in-depth qualitative studies with six respondents
representing five manufacturing companies The empirical
studies were conducted using semi-structured open-ended
interviews [33-34] with respondents working for companies
active in the Faste Laboratory at Luleå University of
Technology, Sweden, focusing on FP Innovation Further, two
additional companies, Electrolux (which sells functional offers
to customers) and Komatsu Forest (which develops forestry
equipment with availability-oriented solutions) were also part
of the empirical studies Thus, the respondents were well
aware of and knowledgeable regarding FP The respondents
were professionals responsible for marketing, services,
strategy, development and sales at the following five
international companies:
x Bosch Rexroth AB (two respondents – technical
product and service managers)
x Gestamp Hardtech AB (one respondent – manager
tool design and development)
x Volvo CE (one respondent – chief project manager)
x Komatsu Forest (one resp – Executive VP)
x Electrolux (one resp – regional category manager)
The purpose of having multiple companies with diverse
focus was to ensure an advance in the understanding of
availability-improving measures related to software taken or
planned in the context of FP, considering the similarities and
differences between the companies (cf [35]) Although the
companies have different offerings, they all face the common
challenge of how to best develop, market and sell FP and/or
similar concepts such as TES or PSS/IPS2, either as a provider
in a partner consortium or as part of their own offerings The
companies are all manufacturing companies with roots in
hardware development However, additional complimentary
components have been added to their customer offerings
What the additional components comprise and their weight or
importance differs depending on industry and customer
segments served Some of the companies aim to increase their
revenue from soft parts; i.e., services, knowledge or
know-how, etc., as well as FP sold globally Thus, the FP planned or
currently offered by the companies vary and have different
emphasis on the composition of hardware, software,
service-support system and management of operation
Initially, semi-structured interviews were used, with
open-ended questions [33-34] allowing the respondents to give
detailed answers and the possibility to add extra information
where deemed necessary [36] The duration of the interviews
was between one and two hours In order to reduce response
bias, the respondents came from various parts of the
organizations as well as different levels i.e., strategic, tactical
and operational units In order to strengthen the validity of the
study, data were continuously displayed using a projector during the interviews, allowing the respondents to immediately read and accept the collected data If immediate reading and acceptance was not possible, the interview transcript was read and accepted afterwards by the respondent After that, the collected data were displayed and analyzed using matrices (cf [37]) The analyzed data were finally summarized into a matrix comprising software-related measures taken or planned to improve availability of FP or similar offers The findings were then categorized according
to the area of concern, i.e., ‘monitoring/status/data’ and
‘management’ which were distinguishable after the analysis For reasons of confidentiality, only an aggregated view of the analysis (with relevant interview statements) is presented
4 Findings – software-related availability measures planned and taken in the context of Functional Products (or similar offers)
The findings below are categorized and sorted according to two categories: MSD (monitoring/status/data) and MGMT (management), as these were the distinguishable categories found after the analysis
Table 1 – software-related availability measures planned or taken.
1 We use the control system and a number of sensors and data extractors to extract information from the system and store the information centrally In this manner it is possible to deduct reference values for health indexes.
MSD
2 We intend to increase the monitoring of the process, and for the analytics of some parts we will use data from actual measurements and compare with our theoretical models in order to ensure that the model and production process are aligned Thus, we will have a continuous learning in between the models and production processes We plan to make even larger analyses and simulations using historical data to be able to find additional systematic flaws/issues, pareto errors, and errors based on wrong assumptions/input
Thus, the software and analytics supports our decision making regarding availability and what measures to take
We use an internal cloud for the calculations, which provides us with redundancy and high availability.
MSD
3 We will have both on-board/local and centralized software
to monitor and diagnose equipment The analytic tools will reside in the (internal) clouds, and the (historical) data we save are combined with data from other sources as well, such as service records, environmental conditions, etc A framework like this supports the back office when making, e.g., cross-searches for problems/issues and finding root causes or problem patterns.
MSD
4 The equipment is in continuous contact with the central service system and potential problems or errors and error codes are stored in the service system The central service system quickly relays response or work orders, often auto-matically, to a service/repair technician (via his/her mobile phone) Commonly, this enables the service/repair techni-cian to be on site within an hour or two In the future, if the number of equipment parts of functional offers sold is scaled up, potentially, a centrally made analysis and priori-tizetion of response/work orders may become necessary.
MSD
5 Currently, we keep a majority of the software on-board/
locally and use federated processing in order to do as much
MSD
Trang 4limit so that only flags, errors/issues or relevant operations
data are sent as well as how often they are sent The
back-office can then connect to any equipment where there is a
diagnosis need and retrieve additional data and analyze that
in greater detail Currently, depending on laws/regulations,
security and customer wishes, we only use internal clouds
and information systems for analytics of big data.
6 We have a communications connection with the machine
and measure what it is used for and how much This gives
us an indication as to what is worn and torn and which
components, oils/parts which need to be replaced/ serviced.
MSD
7 The software will be further modularized to enhance change
management, i.e., patches and upgrades.
MGMT
8 Some customers want the data storage/analytics
functionality locally, whereas other customers consider it no
problem to have a centralized or cloud solution Reasons for
local software are security requirements, which vary
between industries, or lack of communications/network
coverage/bandwidth One current issue with the data storage
is the cost associated with large amounts of data.
MGMT
9 We have not yet had any security breaches We use
satellite-links for internet access in some locations, which is
expensive but works well The communications set-up
depends on the location and what is available, and
commonly we use telecom operators’ mobile networks (or
other types of networks) Thus, we depend on the telecom
operators and in order to make the agreements stronger, we
try to make the agreements together with our customers.
MGMT
Table 1 comprises nine software-related availability
measures, and the findings point towards an increase in use of
monitoring and analysis/use of collected data combined with
meta- and stored historical data Further, monitoring of health
status, wear and tear, and simulation of availability will
increase Thus, ‘design for monitoring/diagnostics’ will likely
be part of the FP providers’ development strategies and
lifecycle planning further on In addition, management of
software, including connectivity, local/on-board versus central
versus cloud software, and security concerns, is of growing
interest In particular, security and new legislation or
regulatory frameworks will impact future FP software set-ups
The findings corroborate the literature (see related work),
however, they also indicate a further increase in software in
FP integration/automation of service measures to maintain the
desired FP availability level and prevent serious failures and
breakdowns Further, as the amount of data grows, a decision
whether to use streaming data, stored data, or a combination of
these, as well as how much data should be stored and for how
long, is needed Quantitative analytics can be applied where
suitable to find problem patterns or emerging problem areas
To sum up, the measures address both how to use software to
improve FP availability as well as how to improve the
availability of the software constituent
To complement the above findings, Table 2 presents
reflections and ideas for the future from the respondents:
Table 2 - reflections made by the respondents.
# Reflections
1 In the future we want to be better at detecting wear, which is a very slow
process We need to find out exactly where and simulate the process, and
hope that the theoretical calculations and simulations will render results
that closely approximate real conditions.
2 In the future, we want to build additional functionality into our function
in order to keep high quality on the output as well as high level of availability pertaining to the function However, there is an inherent risk in adding additional functionality in terms availability, but the increased pressure to keep costs down will lead to further addition of functionality Thus, there will be a need for improved risk awareness/management.
3 Our software is to a large extent in the (internal) cloud and more will end
up there in the future Currently, we have servers in different parts of the world sending us data as well as in between the servers In the future, we may also add more software locally in the function to enhance control of the process As of now, we do not have a lot of local/on-board software in the function – but the new ideas for measurements and enhanced monitor-ing will likely require that Further, we are also considermonitor-ing allowmonitor-ing our customers to use and access parts of our simulators/analytic tools – which may increase our level of professional services and revenue Potentially, additional software can be integrated into the hardware if our future knowledge/monitoring of functions in operation show the need for that.
4 In the future, we expect to have most software on-board/locally (in case there is no or bad connectivity) and do a majority of the pre-processing of data on-board as well This also depends on high sampling rates, and it would not be justified to send all data possible as we do not have any use for that However, for efficiency reasons, any complex and
demanding/expensive computations will be made in our internal clouds.
5 We will need to work more with distributors and encourage them to use the new monitoring/diagnosis tools, as today some of them are reluctant to
do so That will benefit the whole value-chain, including the customers.
6 The on-board software keeps the equipment in order and generates error codes in case of issues/problems, which in turn are sent to a central service system The service system can be located at our premises or at the customer Currently, we do not have any plans to use a cloud solution, partly due to connectivity issues However, we will continue to investigate possibilities like cloud solutions Future-wise, the amount of on-board and central software will likely increase due to more functionality that needs to
be added successively as new requirements arise.
7 Our on-board software is connected to a cloud service, where data is stored We will probably have more software on-board in the future than today, and we want to be able to change/adjust the software and configurations/settings as well as get more relevant data and parameters for analytic purposes Further, based on data collected, we want to be able
to optimize the operation of the machines (during run-time) We rely on other parties regarding spare parts and service personnel, and must therefore make rather advanced agreements The agreements need to clearly specify requirements and related measurements/KPIs.
Reflections 1-7 point to the fact that there will be considerably more software in the FP in the future and, further, that a lot more of the FP functionality will reside in the software, since that will be easier and more flexible to manage and change Thus, software management in terms of continuous development, maintenance, bug fixing and patching, operating system maintenance and migrations, security, digital preservation, etc., will require greater effort, additional competencies and resources Many companies will thus need to become more software-centric, compared to having been hardware-centric, and understand software-related availability
Where will the measures to improve FP availability matter most? This is a valid question to justify efforts and cost towards an increase in availability and a potential competitive edge on the market Depending on how the FP delivers the function and to which extent the main constituents are involved, an analysis should provide the answer when addressing the hardware, software, service-support system or management of operation A common approach is to increase the reliability of key components in the hardware, improve the hardware maintainability, and combine with an adequate
Trang 5service-support system set-up Further, use of
self-repair/re-configuration is an additional possibility However, extending
or adapting the service-support system is often expensive if a
large geographic area with many customers is to be covered If
the connectivity is good and the software provides a lot of
functionality, it is likely that measures applied here will render
a good effect In addition, also in scenarios with low
connectivity, e.g., local monitoring with a weekly analysis
may also provide beneficial results Regarding management of
operation, improvements in simulations and measurements of
availability can give a good long-term effect and provide
decision support based on facts (which are collected via
monitoring and from other operational data originating from
the service-support systems and other software logs)
Where is it easiest to take measures? Hardware measures
normally require a physical visit to provide service or replace
parts (in case of re-design, remanufacture or if a part is broken
or worn) In addition, if it is possible to predict problems,
repair of hardware components and service may be carried out
in a proactive manner based on facts from monitoring Thus,
the hardware commonly involves personnel-intensive
measures Changing or adapting the service-support system is
commonly expensive and may also involve negotiations with
trade unions, etc Still, such changes regarding the number of
service-support personnel, localization, number and content of
spare parts storage and their geographic distribution, etc., may
be necessary in order to uphold the agreed-upon availability
level If new knowledge needs to be added, it often demands
time and resources if the service-support system is extensive
Concerning the software, if the FP are connected and
accessible, and remote changes or updates are acceptable, it
will be efficient to do as much software changes, updates and
addition of new functionality as possible remotely (in an
automated fashion) In the future, automated service measures
may also be initiated remotely Concerning distributed, central
and cloud software, these are preferably managed from where
the best expertise is available
Further, the combination of provider/customer down-time
cost, duration of FP contracts, and how many variations of FP
are provided, will affect the design for availability The FP
customers’ total cost and revenue from using FP in production
processes depend a lot on how the FP is designed, developed,
realized and operated As the FP contract duration may span
up to 30 years, there may be considerable changes in the
software that were not foreseeable, due to the development of
society, technology and legal and regulatory frameworks
5 Discussion and conclusions
The paper highlights nine software-related availability
measures for FP, whereof six concern monitoring/status/data
and three concern management of software The measures
address how software can improve FP availability as well as
how to improve the availability of the software constituent
The most important software-related measures found are:
x Deduction of reference values for health indexes
x The monitoring of FP will cover many parameters
(including software-related ones) and the analytics will
become further advanced and predict problems
x The functionality based on software in FP will increase, and the use of cloud services will also increase where adequate connectivity is available Thus, software availability needs to be ensured early
by for instance modeling and simulations
x FP software will reside on-board/locally, centrally and
in the cloud, and be modularized to facilitate high availability, large scale management and cost effectiveness Customer preferences and the level of security will guide where the software can reside as well as where data is processed and stored
The longer the FP lifecycle, the more demanding the FP software constituent will become due to changes in society, technology and obsolescence, etc Further, security is an issue
of critical importance, as more and more FP will be connected
to the provider via the Internet or other means
Most of the measures proposed are likely applicable for other concepts, which are similar to FP, such as TES and PSS/IPS2, as the amount of software is anticipated to significantly increase in them as well
As the FP providers mainly have been hardware-centric in the past, they need to become more software-centric To quickly grow the knowledge on software, the FP providers can look at how the software providers strategize, plan and act Further, the FP providers should start to simulate the software
in the same manner as the hardware and service-support system – preferably all together in the same simulation framework If the management of operation constituent is possible to add to the simulation framework as well, the whole
FP could be simulated and optimized more efficiently
The software industry has historically used a number of strategies and methods to address availability such as, e.g., using local software, redundancy, mirroring, or having scalability/elasticity or abundant capacity During recent years developments, including virtualization of servers and various cloud services, have further enabled availability (as well as cost) improvements As large parts of the world modernize ICT-infrastructures and roll out fibre networks and high-capacity mobile networks, connectivity has improved considerably However, the new strategies and methods have weaknesses and may be fraught with, for instance, distributed denial-of-service (ddos) attacks, sophisticated viruses, and other attacks where software/file systems/databases are destroyed, made un-operable, or data are encrypted with a hostile key in order to blackmail the organization in question Recent and emerging countermeasures include new types of anti-ddos firewalls [38] as well as improved incident management by means of automating parts [39] in order to enable an adequate prioritization and quick actions/countermeasures when necessary However, new attack vectors will appear, which will create problems Finally, new recommendations to secure and improve availability of software/cloud services are being published regularly [c.f 40-42]
This initial research has been limited by involving only five companies, and additional companies and research approaches will be used in future research
Finally, the paper has highlighted nine actual and a number
of potential software-related measures to improve the
Trang 6availability of FP or similar offers As the FP software
constituent is expected to increase in the future, it is necessary
to plan, simulate and coordinate the software-related measures
with those for the hardware, service-support system and
management of operation in order to get a good overall result
A possible way to model and simulate the software
availability in FP would be to extend the FP simulation
framework in [16-17] to also include the software constituent
Acknowledgements
The research was funded by the VINNOVA Excellence
Centre the Faste Laboratory and the ProcessIT Innovations
R&D Centre, Luleå University of Technology, Sweden
References
[1] Alonso-Rasgado, T ,Thompson, G and Elfstrom, B-O (2004) The design of
functional (total care) products, Jour of Eng Design, Vol 15, No 6, pp515-540.
[2] Lindström, J., Karlsson, L., Löfstrand, M and Karlberg, M (2012) A development
process for Functional Products: hardware, software, service support system and
management of operation, International Journal of Product Development, Vol 16.,
No 3/4, pp284-303.
[3] Lindström, J., Plankina, D., Nilsson, K., Parida, V., Ylinenpää, H and Karlsson, L
(2013) Functional Products: Business Model Elements, proceedings of 5 th
CIRP Int.
Conf on Industrial Product-Service Systems, 14 th -15 th of March, Bochum, Germany.
[4] Lindström, J., Sas, D., Lideskog, H., Löfstrand, M and Karlsson, L (2015) Defining
‘Functional Products’ through their constituents, International Journal of Product
Development, Vol 20, No 1, pp1-24.
[5] Lindström, J., Dagman, A and Karlberg, M (2014) Functional Products lifecycle:
Governed by sustainable win-win situations, proceedings of The 3 rd International
Conference on Through-life Engineering Services, 4 th
-5 th
of Nov., Cranfield, UK.
[6] Lindström, J., Dagman, A and Karlberg, M (2015) The Functional Products
technical lifecycle and its four sub-lifecycles, proceedings of The 4 th
International Conference on Through-life Engineering Services, 3 rd -4 th of Nov., Cranfield, UK.
[7] Parida, V, Rönnberg-Sjödin, D, Wincent, J and Ylinenpää, H (2013) Win-Win
collaboration, functional product challenges and value-chain delivery: A case study
approach, proceedings of 2 nd International Through-life Engineering Services
Conference, 5 th
-6 th
of November, Cranfield, UK Procedia CIRP.
[8] Stahel, W (1997) The Functional Economy: Cultural and Organizational Change, in
(eds) Richards, D J., The Industrial Green Game: Implications for Environmental
Design and Management, National Academy Press, Washington DC, USA.
[9] Thoben, K-D., Eschenbächer, J and Jagdev, H (2001) Extended Products: Evolving
Traditional Product Concepts, proceedings of 7 th
International Conference on Concurrent Enterprising, 27 th
-29 th
of June, Bremen, Germany.
[10] Mont, O (2001) Introducing and developing a Product-Service System (PSS)
concept in Sweden, The International Institute for Industrial Environmental
Economics (IIIEE), Lund University, Sweden.
[11] Meier, H., Roy, R and Seliger, G (2008) Industrial Product-Service Systems –
IPS 2
, CIRP Annals Manufacturing Technology 2008, pp1-24.
[12] White, A L., Sloughton, M and Feng, L (1999) Servicizing: The Quiet Transition
to Extended Product Responsibility, Tellus Inst., Boston, USA.
[13] Tomiyama, T (2001) Service engineering to intensify service contents in product
life cycles, proceedings of EcoDesign 2001, 2 nd
Int Symposium on Environment Conscious Design and Inverse Manufacturing, 12 th
-15 th
of Dec., Tokyo, Japan.
[14] Roy, R., Shaw, A., Erkoyuncu, J A and Redding, L (2013) Through-life
Engineering Services, Measurement and Control, 46(6), pp172-175.
[15] Andrews, J D and Moss, T R (2002), Reliability and Risk Assessment Second
Edition Professional Eng Publishing Limited, London and Bury St Edmunds, UK.
[16] Löfstrand, M., Andrews, J., Karlberg, M and Karlsson, L (2011) Functional
Product system availability: simulation driven design and operation through coupled
multi-objective optimization, Int Jour of Product Dev., Vol 13, No 2, pp119-131.
[17] Löfstrand, M., Backe, B., Kyösti, P., Lindström, J and Reed, S (2012) A model for
predicting and monitoring industrial system availability, International Journal of
Product Development, Vol 16, No 2, pp140-157.
[18] Reed, S., Andrews, J., Löfstrand, M and Karlsson, L (2012) ‘The design and
modelling of maintenance during the development of functional products or product
service systems’, proceedings of 1 st
International Through-life Engineering Services Conference, 5 th -6 th of November, Cranfield, UK.
[19] Reed, S., Löfstrand, M., Karlsson, L and Andrews, J (2013) Service support system modelling language for simulation driven design of functional products, proceedings of 2 nd
International Through-life Engineering Services Conference, 5 th
-6 th of November, Cranfield, UK Procedia CIRP.
[20] Pavasson, J., Ljung, A-L., Karlberg, M., Larsson, S., Johansson, S and Lundström,
S (2014) Challenges and opportunities with simulation-driven functional product development and operation, proceedings of 3 rd CIRP International Conference on Through-life Engineering Services, 4 th
-5 th
of Nov., Cranfield, UK Procedia CIRP [21] Lindström, J., Löfstrand, M., Reed, S and Alzghoul, A (2014) Use of Cloud Services in Functional Products: Availability Implications, proceedings of the 6 th
CIRP International Conference on Industrial Product-Service Systems, 1 st -2 nd of May, Windsor, Canada Procedia CIRP, Vol 16.
[22] Norris, M (2013) Planning to extend the life of major assets, proceedings of 2nd International Through-life Engineering Services Conference, 5 th
-6 th
of Nov, Cranfield, UK Procedia CIRP.
[23] Baker, A (2013) Configurable obsolescence mitigation methodologies, proceedings
of 2 nd
International Through-life Engineering Services Conference, 5 th
-6 th
of November, Cranfield, UK Procedia CIRP.
[24] Muñoz, R G., Shehab, E., Weinitzke, M., Bence, R., Fowler, C and Baguley, P (2015) Towards Software Performance Monitoring: An Approach for the Aerospace Industry, proceedings of the 4 th
International Conference on Through-life Engineering Services, 3 rd
-4 th
of November, Cranfield, UK Procedia CIRP.
[25] Karlsson, L., Lindström, J and Löfstrand, M (2012) Functional Products -Goodbye to the industrial age, Ericsson Business Review, Vol 81, No 2, pp21-24 [26] Kyösti, P (2015) Predicting functional product availability and cost through a simulation driven approach, Doctoral thesis, Luleå University of Tech., Sweden [27] Kyösti, P and Reed, S (2015) Prediction of service support costs for functional products, Simulation Modelling Practice and Theory, Vol 59, pp52-70.
[28] Olsson, T., Källström, E., Gillblad, D., Funk, P., Lindström, J., Håkansson, L., Lundin, J., Svensson, M and Larsson, J (2014) Fault diagnosis of Heavy Duty Machines: Automatic Transmission Clutches, proceedings of International Conference on Case-Based Reasoning: Workshop on Synergies between CBR and Data Mining, 29 th
of September – 1 st
of October, Cork, Ireland.
[29] Källström, E., Lindström, J., Håkansson, L., Karlberg, M., Bellgran, D., Frenne, N., Renderstedt, R., Lundin, J and Larsson, J (2015) Analysis of automatic transmission vibration for clutch slippage detection, proceedings of 22 nd
International Congress on Sound and Vibration, 12 th
-16 th
of July, Florence, Italy [30] Garmabaki, AHS., Birk, W and Lindström, J (2016) Software fault detection in control systems, proceedigns of the European Safety and Reliability 2016 conference, 25 th
-29 th
of September, Glasgow, Scotland.
[31] Reim, W., Parida, V and Lindström, J (2013) Risks for functional products – Empirical insights from two Swedish manufacturing companies, proceedings of 2 nd
International Through-life Engineering Services Conference, 5 th -6 th of November, Cranfield, UK Procedia CIRP.
[32] Lindström, J Through-lifecycle aspects of Functional Products to consider during development and operation: A literature review In (eds) Redding, L & Roy, R Through-life Engineering Services: Motivation, Theory, and Practice Cham: Springer-Verlag London Limited, Ch 21, pp187-207 (Decision Engineering ).
[33] Patton, M Q (1990) Qualitative evaluation and research methods, Sage Publications, London, UK.
[34] Kvale, S and Brinkmann, S (2009) InterViews: learning the craft of qualitative research interviewing, Sage Publications, LA, USA.
[35] Yin, R K (2003) Case Study Research: Design and Methods, Sage Publications, Thousand Oaks, CA, USA.
[36] Fontana, A and Frey, J (1994) Interviewing, in (eds) Denzin, N and Lincoln, Y., Handbook of qualitative research, Sage Publications, Thousand Oaks, CA, USA [37] Miles, M and Huberman, M (1994) An expanded sourcebook – Qualitative Data Analysis, 2 nd
ed, Sage Publ., Thousand Oaks, CA, USA.
[38] Waverley Labs (2016), Software defined perimeter, http://www.waverleylabs.com/services/software-defined-perimeter/
[39] Resilient (2016), Resilience and automated measures, http://www.resilience.com/ [40] ISO (2016), Information Technology and Information Security standards, http://www.iso.org
[41] NIST (2016), Information Technology guidelines, http://www.nist.gov/information-technology-portal.cfm
[42] Cloud Security Alliance (2016), Cloud Security research, https://cloudsecurityalliance.org/research