Managing Change in the Delivery of Complex Projects

This research analyzes practices of managing change in complex projects, focusing on configuration management, asset information, and 'big data'. It examines leading practices in Airbus, CERN, and Crossrail, discussing the challenges of managing change in the era of 'big data' and implications for research and practice.

Managing change in the delivery of complex projects: Config-
uration management, asset information and ‘big data’
Jennifer Whyte⁎, Angelos Stasis, Carmel Lindkvist
School of Construction Management and Engineering, University of Reading, Whiteknights, Reading, RG6 6AY, United Kingdom
Received 30 September 2014; received in revised form 5 February 2015; accepted 12 February 2015
Available online 21 March 2015
Abstract
As we enter an era of ‘big data’, asset information is becoming a deliverable of complex projects. Prior research sugges
enable rapid, flexible forms of project organizing. This research analyses practices of managing change in Airbus, CERN an
desk-based review,interviews,visits and a cross-case workshop.These organizations deliver complex projects,rely on digitaltechnologies to
manage large data-sets;and use configuration management,a systems engineering approach with mid-20th century origins,to establish and
maintain integrity. In them, configuration management has become more, rather than less, important. Asset information is
managed through digital systems, using relatively hierarchical, asynchronous and sequential processes. The paper contrib
to flexibility in complex projects where integrity is important. Challenges of managing change are discussed, considering t
configuration management; potential use of analytics on complex projects; and implications for research and practice.
© 2015 The Authors.ElsevierLtd. APM and IPMA. All rights reserved.This is an open access article underthe CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords: Complex projects; Configuration management; Change; Asset information
1. Introduction
Digitaltechnologiesradically transform projectdelivery.
Twenty years ago,Morris described the evolution ofproject
managementas closely related to developmentsin systems
engineering, modern management theory, and the evolution of
the computer(Morris,1997:p.2).Today,mobile hardware,
cloud computing and integrated software are becoming used for
storage and retrieval,automated search,and prototyping and
simulation functions.As such technologiesare adopted in
project-based industries,theiruse is breaking the mould of
established approaches to project management,enabling more
rapid and agile forms of organizing (Levitt,2011; Whyte and
Levitt,2011). Up-front project planning,using multiple layers
of work breakdown structures,becameestablished by the
1960s in the managementof large complex projects (Morris,
1997: p.44). New digitally-enabled approaches are emerg
industries,such asconsumerelectronics,software develop-
ment,biotechnology and medicaldevices,thatoperatein
dynamic and less predictable situations in which plans ne
be updated and modified during projectdelivery (Whyte and
Levitt,2011).In these,data analytics and visualization using
large digital data-sets – along with rapid,informalinteraction
and exchanges ofinformation – provide the basis formore
responsive,flexibleand real-timedecision-making (Levitt,
2011).
The information used to make decisions in the manage
of complex projects is generated and stored digitally. Com
projectsare a setof projectsthatshare particulardefining
characteristics: they are high-tech, capital intensive engi
projects that are of a significant scale, relatively long dur
and require firms to work collaboratively across firm boun
aries in projectdelivery (Davies and Hobday,2006;Hobday,
1998;Miller et al., 1995).Such projectsdelivercomplex
productsystems,such as aircraft,experimentalfacilities and
⁎ Corresponding author. Tel.: + 44 118 378 7172.
E-mail address: [e-mail] (J. Whyte).
www.elsevier.com/locate/ijproman
http://dx.doi.org/10.1016/j.ijproman.2015.02.006
0263-7863/00/© 2015 The Authors.ElsevierLtd. APM and IPMA. All rights reserved.This is an open access article underthe CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Available online at www.sciencedirect.com
ScienceDirect
International Journal of Project Management 34 (2016) 339 – 351

railways.Their delivery requires systems integration capabili-
ties,as complex productsystems are designed and integrated
through a network ofcomponentand sub-system suppliers
(Davies and Mackenzie,2014; Davies et al.,2009; Hobday et
al., 2005).Within these projects information aboutcomplex
product systems is developed across multiple firms,involving
diverse professions and trades,as these organizations interact
through the digital systems.
A starting point for our work is the observation that, as we
enteran era of‘big data’,assetinformation is becoming a
projectdeliverable.Data are unprocessed,often described as
“unorganized facts” (e.g.Faucheret al., 2008:p. 55),while
information is interpreted and presented to inform in a given
context.Ownersseek to useassetinformation to achieve
sustainable and safe performance of complex systems through
the life-cycle. An asset may be an assembly, sub-assembly, or
component,butis the smallestunitmaintained by an owner.
The term ‘assetinformation’is used to describe information
aboutan asset,which may include the provenance,parttypes
and serialnumbers,design life,maintenance schedule,and
design rationale for sub-systems or components.As data gets
reused acrossthe life-cycle,setsof dataand information
become combined and can be mined,interpreted and used in
new ways. The UK government, for example, is, as a client for
built infrastructure,requiring projectteamsto deliverasset
informationthroughbuildinginformationmodelling(BIS/
Industry Working Group,2011);and seeks to aggregate and
combine data-sets, connecting them with Smart City and Smart
Grid initiatives as partof a strategy for DigitalBuilt Britain
(UK Government, 2013).
Established approaches for managing change on projects use
configuration management, a systems engineering approach with
origins in the mid-20th century. In its original form, configuration
management is characteristics of what Levitt (2011) describes as
‘project management 1.0.’ It involves hierarchical, sequential and
asynchronous processes; managing change against a baseline. Its
use focusesattentionon assetsas configurationitems:
sub-systems or components that have value to the organization,
in which changes will often have systemic consequences on the
function or layout of other items within the product structure and
hierarchy.The baseline is an agreed description ofone ora
numberof assetsat a point in time,wherethe current
configuration of a complex product system is described by the
latest baselines plus approved changes.
New practices of managing change in complex projects m
be expected as we enter an era of ‘big data’, in which inter
external data-sets become linked and asset information be
project deliverable. Morris argued that:“rigorous change co
is fundamentalto good projectmanagement” (Morris,2013:
p.126).Poor change controlis one of the issues thatlimits
managers’ ability to execute viable project plans (Pinto,2013).
Otherssee projects,themselves,as informationprocessing
systems(e.g. Winch (2010)drawingon Galbraith(1973,
1977)).As projectmanagementinformation systems (Braglia
and Frosolini, 2014) are increasingly used, altering the pac
complexity (Shenharand Dvir,2007)of projects,there are
challenges to the: “heavy formality of several of the techni
to manage large-scale, one-time, non-routine projects” (Mo
2013:p. 133).Here,Morris,like Levitt,pointsto software
projects, in particular, as rebelling, using agile forms of ma
ment,through smallprojectswith close developer-customer
relationships.
The aim ofthis research is to articulate how changes in
assets and the associated asset information are managed i
delivery of complex projects as we enter the era of ‘big dat
This is done by analysing leading practices in three organiz
tions: Airbus, CERN and Crossrail. Each of these organizatio
deliverscomplex projects;relieson digitaltechnologiesto
manage a large volume of information; and uses configura
managementto establish and maintain the integrity ofthe
complexproductsystem andassociatedinformation(see
Table 1).Airbus is an aircraftmanufacturer,operating in the
aerospace industry and engaged in production of commerc
and military aircrafts,with long-term projects to design and
develop new aircraft designs and bring them into operation
headquarters are in France but the supply-chain is global, w
the assembly of each plane involving thousands of compan
and millions of parts.CERN is the European organization for
nuclearresearch and thelargestparticlephysicsresearch
establishment in the world, with 21 member states, 6 obse
states and more than 80 collaborating countries. Its missio
provide scientists from all around the world with tools to st
the building blocks of matter and the origins of the univers
Crossrailis the largestconstruction projectin Europe,with
14.8bn funding,delivering a new 100 km railroute with 10
new stations and a tunnelthrough centralLondon connecting
40 stations. It has a complex supply-chain involved in deliv
with more than 1,300 contracts.
Table 1
Background of organizations studied, and their industries.
Airbus CERN Crossrail
Industry Aerospace design and manufacturing Nuclear research infrastructure Civil engineering and railway infrastructure
Background Leading aircraft manufacturer of commercial
and military aircrafts, with a substantial
international supply-chain.
Largest particle physics research
establishment in the world with tunnels and
particle accelerators
Design and construction of new railway
across London with tunnels and 37 stations
Relationship to projectsLong term internalprojectsto design and
manufacture new additions to the fleet, such
as the A380, integrating sub-systemsand
components and delivering to customers.
Experienced project owner, managing
supply-chain delivering accelerators such as
the Large Hadron Collider (LHC)
Delivery client for a mega-project, the
duration of which is 2008–2018, interfacing
with future operators of the railway.
Location France Switzerland UK
340 J. Whyte et al. / International Journal of Project Management 34 (2016) 339–351

The nextsection outlinesextantresearch on managing
change in delivery,before the following sections describe our
research methods and findings.It outlines the development of
configuration management techniques; characteristics of an era
of big data;and relevance configuration managementand big
data to Morris’ interest in ‘reconstructing project management.’
2. Managing Change in Project Delivery
2.1.Developmentof configuration managementtechniques to
manage change
Configuration management was developed in the 1950s by
the US military to control documentation in the manufacture of
missiles (Brouse, 2008; Burgess et al., 2005; Gonzalez, 2002).
Early documentation on engineering change control,released
by the military,clarifies the contractualobligations and role
of suppliers.It refersto a productbaseline describing the
functional,physicaland interoperabilitycharacteristicsof
components for testing and operations (DOD,1978;Military,
1988). Two example change processes, used in the commercial
arrangements ofacquisition and supply,are provided by the
Department of Defense guidance issued in 2013 (see Fig. 1 for
one of these,where the other is a variant).As well as change
control,the classicapproach to configuration management
involves the identification ofthe productstructure and con-figuration items; status accounting to determine the configura-
tion of the system at any stage of the lifecycle (Burgess et al.,
2003;Kidd and Burgess,2010) and reporton the availability
and retrievability of data; and audit to verify the consistency of
the information (Kidd and Burgess,2010).The approach has
become extensively used in the software industry (Bersoff,
1984;Estublier,2000;Williams,2009),and in safety critical
systems such as nuclear and aerospace (Burgess etal.,2005;
Williams, 2009). It became recognized as an ISO 10007 quality
managementprocess in 1995 (ISO,2003).The presentation,
with a man ata drawing board,shows the heritage ofthis
approach in the paper-based processes of the late 20th century;
though in the 21stcentury such processes are supported by
digital systems.
The US military describesconfigurationmanagement’s
overarching goalas:“to ensure there is documentation which
completely and accurately describes the intended designthe
actualproductmatchesthe documentation,and thereare
processes in place so this continues throughoutthe product’s
life” (DOD, 2013:p.10).The ambition isto addressthe
problems which occur in projects due to unchecked chan
one sub-system having widerconsequencesfor othersub-
systems of a product (Hameri,1997); and due to scope creep,
where requirementschangeduring the processof delivery
(Williams,2009);providing traceability ofproductdata to
understand where problems occur, diagnosing and contri
to recovery (Burgesset al., 2003).A configuration is:“a
generic term for anything thathas a defined structure or is
composed of some predetermined pattern” (Kidd and Bu
2010:p. 109).An authorization approach is used to control
change and there are differenthierarchy levels depending on
the use of configuration items (Billingham, 2008). A base
established: “wherever it is necessary in the product life
to define a reference for further activities.” (ISO,2003: p.4).
From an approved baseline,a configuration change authority
assessesthe recommendationof other representativesto
approve,approve with modifications,or disapprove submitted
changes based on the:“totallifecycle impactof the action to
includecost,schedule,performanceand logisticsimpact”
(DOD,2013:p. 27).Thus the processes,procedures and users
of the configuration managementsystem play an integral role
in maintaining theintegrity ofinformation throughoutthe
life-cycle by controlling changes.If users do notfollow the
process,errors can occurwhich can cause problems to the
product in production and to related information dissemin
(Hameri,1997;Hameriand Nitter,2002).Researchersof
configuration management constantly find the benefits o
a controlled process of change are notalways understood or
realized by users (Aliand Kidd,2014;Burgess etal.,2003;
Kidd, 2001; Kidd and Burgess, 2010).
The upsurge in the use of digitaltechnologies and flexible
team-working bring into question configuration managem
practices based on documents rather than information (B
et al., 2005). Characteristics of ‘configuration manageme
Submission of:
• Engineering change
proposal (ECP) for
asset
• Notice of revision
(NOR) to associated
information
Configuration
Control Board
New revision
released
Configuration
change
authority
approval
Draft
change
Engineering
release record
(ERR) process
Original
document
Configuration
officially accepted
ECP
NOR
ERRREV B REV C
123 123
Fig. 1. Standard change process (redrawn from source: DOD, 2013: p. 33, to clarify and explain acronyms, same images of documents a
341J. Whyte et al. / International Journal of Project Management 34 (2016) 339–351

notstable and fixed butare themselves evolving,to include
the life-cycle, agile approaches, and changes to strategy as well
as project.For example,aftercancelling itsconfiguration
management standard MIL-STD-973 in 2000,the US Depart-
mentof Defensefound thatit continuedto be usedin
non-standard ways. From 2010 they developed a new standard
for configuration managementto changes through life using
digital systems(Windham,2012). The interim standard
mentionsthree baselines,with a functionalbaseline giving
system level requirements; allocated baseline giving subsystem
or configuration item level requirements; and product baseline
giving detailed definition (DOD, 2013: p. 18). There are recent
attempts to develop an agile approach to configuration manage-
mentthrough asystem to accommodatesmallcontinuous
changes and manage the additional complexity across dispersed
teams (Moreira,2010);and reports suggesting an expanded
scope ofconfiguration management,to include enterprise as
well as product baselines (Wozny et al., 2014).
2.2.Reuse of information and data linkages in the era of ‘big
data’
There are rapid developments in digital technologies and an
extensive growth in the volume ofdata stored digitally that
affects projectdelivery.The term ‘big data’ indicates the use
of large heterogeneousdata-setsthat can themselvesbe
aggregated,and subjected to variousformsof analyticsto
enable patterns in the data to be visualized.While there is not
agreementon a precise definition,many writers,across both
information science and socialscience literatures,draw on
Laney (2001)to referto data volume,velocity and variety.
Volume of data is implied by the term ‘big data’ and is an issue
because there is an increasing extent of data available (Wu and
Wu, 2014), with 2.5 exabytes of data created globally each day
in 2012 and expectations thatthe rate of data production will
double every 40 months (McAfee and Brynjolfsson, 2012: 62).
Yet recentwork emphasises characteristics other than volume
(Boyd and Crawford, 2012; Kitchin, 2014). Velocity relates to
the speed ofproduction and ofaccess ofdata.Advances in
computing processing capacity enable data-sets to be engaged
with in real-time, or near real-time, rather than ‘freeze framed’
(Kitchin,2014)or processed offline.Varietyrefersto the
diverserangeof datasourcesand typesof datathatare
combined (Viitanen and Kingston,2013; Wu and Wu,2014).
On projects this variety can include hierarchically structured
data such as models, and unstructured data such as laser scans,
videos,sensordata,photosand experimentaldata.While
volume, velocity and variety are generally characteristic of the
applications described as ‘big data’,these applications vary in
the extentto which they emphasise one or the other of these
characteristics.
Big data represents a paradigm shift,where “mostof our
attitudes and behaviours still reflect hierarchical and sequen-
tial processing ofdata” (Galbraith,2014).Information is no
longer created and used for a single purpose (Constantiou and
Kallinikos, 2015). As we enter an era of ‘big data’, information
from projects becomes seen as a projectdeliverable and used
throughoutthe lifecycle.Organization scholars observe how:
“organizations are swimming in an expanding sea of data t
is either too voluminous or too unstructured to be managed
analyzed through traditionalmeans” (Davenportetal.,2012:
p. 22).Differentprofessionalusers bring theirown way of
organizing data to understanding it, so early work highlight
needfor indexingstrategiesthat were agnosticto these
structures (Laney,2001: p.1).Big data is different from ‘lots
of data’ as itis: “those data thatdisruptfundamentalnotions
of integrity and force new waysof thinking and doing to
reestablish it”(Lagoze, 2014: p.4-5). It thus raises a conund
for science (Lagoze,2014): how to utilize the benefits of ‘big
data’ while maintaining the validity of data.
Despite ongoing concerns aboutintegrity,many organiza-
tions are movingaway from asynchronousand reactive
decision-making;to the use of predictiveanalyticsfor
real-timeand proactivedecision-making.Within thesocial
science literatures on ‘big data’,analytics is described as a
differentiatorof organizationalperformance (LaValle etal.,
2011);a source of competitive advantage (Barton and Court
2012;McAfee and Brynjolfsson,2012);or new frontierof
competition (Floridi, 2012). Here, the challenge of big data
been reframed as one ofidentifying smallpatterns (Floridi,
2012) within immense databases, and the use of these to c
new value and knowledge.Emerging technologies are associ-
ated with this business analytics (Chen etal.,2012);and the
way that information is considered is different, as analytics
be used to seek insights from the flow of information as we
its content (Williams et al., 2014). Various text and web mi
as wellas socialnetwork analysis techniques are becoming
used to organizeand visualizeinformation to understand
performance in organizations (Chen etal.,2012;Williams et
al., 2014). The many-to-many non-linear data relationships
arise in large and evolving data-sets (Wu and Wu, 2014) le
challenges in ensuring reliability,which has implications for
decision making.Differentpreferences in recording data can
resultin diverse representations and relationships thatmakes
it difficultto discoverusefulpatterns (Wu and Wu,2014).
Synthesized information may thusneed to be situated in a
broaderhistoricalcontextto be used in a predictive manner
(Boyd and Crawford, 2011).
The power of flexibly linking assetinformation with other
data-sets is beginning to be realized by owners ofcomplex
productsystems such as infrastructure.Transportfor London
has, for example, made data available to customers, and a
engineers,to combinedatain new waysto develop new
applications.1 Williams et al. (2014) point to the potential for a
census ratherthan sampling approach to organizationaldata
to be used to assessorganizationalmaturity in the use of
project-management. Recent industry reports envision a fu
in which data-sets are linked and analytics are used predict
ly, for exampleto precomputescenariosand inform
decision-making (e.g.BIM 2050 Group,2014).There may be
latent applications in complex projects, for example in the
1 See https://www.tfl.gov.uk/info-for/open-data-users/(Anotherexample is
http://data.london.gov.uk/).
342 J. Whyte et al. / International Journal of Project Management 34 (2016) 339–351

of dashboards as an interface to whole data-sets to evaluate
supply-chain performance.
2.3. ‘Reconstructing’ project management: configuration
management in an era of big data
Morris (2013)has an ambition to ‘reconstruct’project
managementby shiftingattentionfrom the executionof
projects to the managementof projects – which includes the
definition of the project, the role of the owner or sponsor, and
the project context. This wider view is necessary to investigate
how change is managed in an era of‘big data’where asset
information has becomes a projectdeliverable.The complex
projectdelivers an assetthatis used in operations.As asset
information becomes a deliverable, Fig. 2 shows how data-sets
become increasingly aggregated and linked in the delivery of
the project,with links to externalsources ofdata (e.g.from
suppliers,manufacturersand maintainers),and the complex
projectresponsiblefor managing changein an increasing
volume, velocity and variety of data in project digital systems.
This anticipatespotentialnew connections,for example
between information used in the project and reference schedule
and cost data; and between asset information and the owner’s
enterprise resource planning system.
Digitalsystemsare nothomogeneous,but combine net-
works,servers,and computerswith softwarefor different
purposes,and from differentvendors.They are used in the
storage,retrieval,management and manipulation of data; with
data management software used to upload, integrate, structure,
index and search data,and to manage remote access, security,
versions and workflows. Stored data includes files, folders and
meta-data;held in a range ofspecialistsoftware.In projects
most of this data is classified and structured,though it can be
combined with a variety ofsources ofunstructured data in
projects,which havepreviously been ignored (Boyd and
Crawford,2011).Where organized fora particularpurpose
we describe itas information,hence assetinformation as a
deliverable is managed digitally, and reuses the same data-sets
that have been used to inform decision-making in delivery.
A major question,raised by the era of ‘big data’ relates to
control and the management of change as digital data-sets cross
organizationalboundaries;particularly asinformation about
assetsgets re-usedin operations.In manufacturingand
construction industriesthe use of information through the
life-cycle is often referred to as Product Life-cycle Manag
(PLM) and Building Information Modeling (BIM) respective
The practical concerns of project managers about contro
contextare reflected in the reportof a meeting ofthe Major
Projects Association,which in summary notes that:“Manual
uploading of data between software programmes allows
understanding of progress and visibility” (MPA, 2013: p. 3
this report there are concerns that project managers feel
as the project controls themselves produce massive data
solutionsproposed are manualinterventions,to identify key
information or manually transfer data.
Configuration management is discussed little in the pro
management literatures,and where it is mentioned,notably in
Morris’ work, it is as implemented through the digital sys
used on projects. In the move toward the “era of the pap
project”, Morris describes configuration management: “e
ed to include the configuration ofthe totalprojectdocumen-
tation handling process.”(Morris,1997:preface,note4);
noting increasing focus on information rather than docum
The Project Management Body of Knowledge Guide likew
refers to the configuration management system as a com
of the projectmanagementinformation system,containing
versionsand baselinesfor all projectdocuments(PMBOK
guide,2013:p. 28).Standards for asset managementlikewise
indicate this need to manage changes to assets assessin
and consequences of change.While we know that users resist
configuration managementprocesses and these are notfully
implemented (e.g.Ali and Kidd,2014),we know little about
how change is managed on complex projects, as we mov
an era of ‘big data,and assetinformation becomes used as a
project deliverable. We approached our empirical work w
question about how changes in assets and asset informa
managed in this context as summarised in Fig. 2.
3. Methods
Using a casestudymethodology(e.g. Eisenhardtand
Graebner, 2007; Stake, 1995; Yin, 1994), we analysed ch
managementpractices in the three separate organizations a
cases,and then compared and contrasted the findings acr
casesto developfurtherinsight.The organizationswere
selected as leading organizations that deliver complex pr
systems using digital technologies to manage the large v
of associated information.Data was collected from the three
organizations,and their interactionswith each other as
summarized in Table 2. Our initialcontactwith participants
indicated ourinterestin examining thetransition from an
“as-designed” configuration baseline to an “as-built” con
ration baseline;as information was delivered to owners and
operators.The preliminary analysis was based on a desktop
review of leadingconfigurationmanagementactivityin
through-lifeengineering and scoping interviewswith 1–2
personnelfrom the CAD managementand/orconfiguration
management teams from Airbus, CERN (online) and Cros
(in-person) as wellas a visitto Airbus premises.The scoping
interviews used a protocol with starter questions includin
is configuration managementdefined? Whatare the processes
Flexibly linking
to other sources
of data, in which
change may be
uncontrolled as
separately
owned
How is change
managed as asset
information becomes a
deliverable on
complex projects?
Information for decision-making in delivery
Project processes / ‘as designed’
Asset information as a deliverable
Project outcomes / ‘as built’
Project digital systems
Aggregation and links with
increased data volume,
velocity and variety
Operation digital systems
Fig. 2. Contextfor our research question on managing changeas asset
information becomes a deliverable on complex projects.
343J. Whyte et al. / International Journal of Project Management 34 (2016) 339–351

involved in setting up this configuration management? What is
the approach to configuration control? What are the challenges in
terms of integration of data from as-design configuration baseline
to the as-built configuration? At the end of projects did you find
differences between the as-built with the as-designed?
The comparisonacrosscases was facilitatedby an
afternoon workshop hosted in the Crossrail offices in London.
The workshop in Crossrailoffices involved twenty partici-
pants,with atleasttwo participants from the CAD manage-
mentand/orconfiguration managementteamswithin each
organization studied.This day was recorded with video;49
photographs and notes as wellas through the distribution of
presentation from each organization afterwards (totaling 94
slidesfrom the3 companies).It was an opportunity for
presentations about the context for the research; for feedback
of preliminaryfindings; presentationson configuration
management from the three companies and discussion of key
theoretical challenges.
Following the workshopwe analyzedand iterateda
detailedtable that comparedconfigurationmanagement
practices in the three collaborating organizations,in relation
to topics that had arisen in the discussion at the workshop: 1)
background(e.g. overview,infrastructuretype, scopeof
works,budgets);2) lifecycle (e.g.typicallifecycle duration;
developmenttime);3) complexity (e.g.physicalassets;asset
information);4) configuration managementmotivation (cor-
porate motivation, industry guidance, teams); 5) approach and
systems(e.g.lifecycle breakdown,approach,data manage-
ment,information systemsand supporting tool,structure of
configuration items;6) managing change and change control
process(e.g. changeperspective,changecontrolprocess,
conformancesand non-conformances);7) risks;culturaland
social issues (e.g. language, culture). The table we generated was
used to visualize the data fordiscussion in the research team
(Miles and Huberman, 1994),to identify salient similarities and
differences, and to check details with the collaborating firms. We
discussed the findings in this table, which was seven pages long,
to identify why all three firms had a strong interest in configuration
management and to highlight the similarities and differences. To
develop our argument we reorganized our data,bringing it into
dialogue with the existing literatureson the managementof
projects.
4. Findings
The three organizationsstudied have differentlevelsof
experience ofconfiguration management.Airbus has mature
processes and systems, with interest in leading developme
future systems to manage and controlthe growing amountof
data produced in the delivery ofcomplex products.CERN
introduced configuration management in the 1990s (Bachy
Hameri, 1997; Hameri, 1997; Hameri and Nitter, 2002; Ham
and Puittinen,2003) and are reflecting on their approaches to
configuration management,challenges and areas for improve-
ment. As a major project, Crossrail is a temporary organisa
established in 2008.It has a configuration managementteam
and has drawn on industry standards (e.g. ISO, 2003) to ra
embed configuration managementprocesses in the delivery of
assetinformation.Motivationsgiven forusing configuration
managementin managing change in these settings include th
complexity of complex product systems, operational constr
and the need for valid asset information. The regulated nat
each ofthe industries (aerospace,nuclearresearch and civil
engineering), mean these organizations all need to be able
configuration items to be able to revisit designs and compl
future regulation on safety-critical facilities. In Airbus, confi
ration management is a strategic priority. In CERN and Cro
it is an explicit activity that is addressed in organizational s
using the language of managing change.
4.1. Configuration management in Airbus
As an aircraft manufacturer, Airbus employs around 63,0
people in France, Germany, Spain and the United Kingdom
has subsidiaries in the United States,Japan,China and India;
and the final assembly of aircraft is in France, Germany, Sp
and through ajoint venturein China. Customersinclude
commercialairlines.Satisfying theirevolving needs requires
the design and manufacture ofnew additionsto the fleet.
Airbus invests abouttwo billion euros annually in research,
developmentand technology activities.It operatesin the
highly regulated aerospace industry,in which each aircraft
manufactured requires an individual certificate of airworthi
showing itconformsto the approved design,and hasthe
relevantdocumentation,inspectionsand teststo ensureit
Table 2
Sources of data for the individual cases and the comparison across cases.
Organization Interactions and observations Associated documentation
Data on each individual case
Airbus Online call (2 participants and 2 researchers) and day-long
visit to Toulouse (10 participants and 2 researchers); email
clarifications.
Presentation discussed in online call(20 pages)and 3
publically available presentations (79 pages).
CERN Online call and presentation (2 participants and 2 researchers);
email clarifications.
Internal documents (23 pages, QA procedure; status repor
2 conference papers (6 pages); and presentation (29 page
Crossrail In person interview (1 participantand 2 researchers)and
ongoing collaboration.
Access to internal system and documents on configuration
management.
Comparison across cases
Airbus; CERN and Crossrail 3 hourworkshop with 20 participants,recorded in 49
photographs (some video) and notes; email correspondence.
Presentations from the workshop (94 slides).
344 J. Whyte et al. / International Journal of Project Management 34 (2016) 339–351

will be safe in operations.Aircraftcomponentshave strict
definitionsthatincludearchitecturalnatureand materials.
There are rules on how components are aggregated into sub-
assemblies and assemblies to ensure the correct serviceable part
is available for every configuration and operating condition.
Configuration managementis one of the 14 Airbuskey
competences, considered in procurement of the supply-chain and
included as a requirement in contracts.It is well established in
Airbus.There is a Centre ofCompetence where 80 internal
employees and 70 externalconsultants work to editmethods,
process and tools; and a significant broader capability, with more
than 8000 professionals with related tasks, including 800 internal
and 500 external configuration management professionals.
A motivation for configuration managementis the product
complexity.It can take more than a decade to develop a new
aircraft design.Airbus produces several aircraft families,each
with members and versions.No two aircraftare the same as
customerscan selectbetween variantsand theproduction
standardevolvescontinuously.The numberof partsand
combinationsof solutionsgrow with productcomplexity
along with the combination of configurations to be managed.
Product architecture includes component parts, numbering and
links; with 500–1000 conceptdefinitions,with associated
solutions and ways to restore previous solutions.
There is a significant volume of data, where each aircraft has
millions ofparts.The Airbus A380 plane has,for example,
about 4 million parts,with 2.5 million part numbers produced
by 1,500 companies from 30 countries around the world. There
is a growing velocity,with increasing production rates,and
shorterproductdevelopmentlead times.As aircraftdesign
requirescarefulanalysisof performance data,there isalso
significantvariety,with productinformation linked to struc-
tured and unstructured data-setswith testresults,electrical
bonding calculations,requirements from economic analyses,
weight distribution, and weight calculations, etc. Configuration
managementprocesses have substantialdigitalsupport,with
six preferred suppliers for configuration management services.
However,as the numberof data-setshas increased,some
information related to the aircraft is no longer well linked. For
example,the functional baseline with the initialspecifications
is notwell linked with the technicalbaseline,within which
change is managed once requirements are approved.Instead
these are synchronised at a particular point and reset.
In Airbus, a configuration item is seen as “an invariant item
in the productstructure”,where each configuration item is
linked to a design solution. While configuration items stay the
same, the engineering baseline, used to design, gives a different
view of the productarchitecturefrom the manufacturing
baseline, which is used to install things, and both are different
from the data needed in customer services. It can be difficult to
recover a technical functional baseline when a big evolution is
done to a standardaircraft,wherethe interfacebetween
differentdata-sets are notlinked which means the source of
data is noteasily known.A significantmilestone is the move
between conceptand definition phases,where data is moved
from a development environment to a production environment,
which often involves differentdata structures.At this point,
engineersmay be involvedin manuallyupdatingdata,
re-numbering data-sets as well as re-defining links.
Before product evolution can take place in the concept
there isan important“congruency process”,through which
engineering and manufacturing agree on productarchitecture
and terminology.A requestfor a productevolution then
involves four stages: initialization, which is the change re
evaluation,which involves an evaluation study;investigation,
which involves a modification proposaland consideration of
technical, cost, embodiment and production repercussion
implementation and closure,in which there is a fulltechnical
repercussionssheet,technicaldossierand modifications
approval sheet.There are also processes for managing chan
in productidentification,and releasinginformationfrom
engineering and manufacturing information in the definit
stage.All these process involve a finalstep thatverifies the
implications of change to the product structure at higher
within this complex productsystem;and ensures consistency
acrossall domains(including systems,electrical,testand
technical data).
Airbuswantsto improve efficiency ofthe productdata
managementby achieving:scalability,through reuse of data;
agility,through integration of data,and adaptability,throughflexibility in integrating data changes.Configuration manage-
ment challenges include:
• Differentproductarchitectures– Developing agreement
betweenstakeholdersin engineering,managementand
customer services takes a long time. In Airbus’ experie
developing a complete architecture can take up to thre
to agree.After the congruenceprocess,the product
architecture is the same in both engineering and manu
ing and a bill of materials is produced.
• A desire to be agile in a highly controlled environment.
strictand rigid system developed through the congruen
processlacksflexibility.Yet, even afterthe congruency
process is complete there are still changes that need t
place based on technology developments and the chan
competitive environment that may not have been inclu
the congruency process.
Other challenges include the increase of product and p
complexity;variety ofsoftwarevendorsand theirlack of
supportfor the interfacesbetween systems;heavy load of
changes to manage; communication to a growing configu
management community; and shorter product developm
time.
4.2. Configuration management in CERN
To achieve its mission of providing scientists from arou
world with tools to study the building blocks ofmatterand
origins of the universe, CERN builds and operates huge p
accelerators on the border between Switzerland and Fran
to Geneva.It has approximately 2,350 staff,2,000 contractors
and 10,000 visiting scientists,with an annualbudgetof about
800 million EUR.As an integrated owner-operator,it has
345J. Whyte et al. / International Journal of Project Management 34 (2016) 339–351

responsibility forthe full assetlife-cycle.Developmentand
delivery of each particle accelerator is a major project. The Large
Hadron Collider(LHC), for example,involved millionsof
high-tech components installed in a 27 km long circular tunnel,
100 metersbelow ground,with particlesacceleratedto
99.999999% ofthe speed oflight, corresponding to over
11,000 revolutions per second.The design phase took approx-
imately 20 years,and the work was globally distributed with
collaborators in more than 80 countries across five continents.
The material cost was approximately 3.7 billion EUR.
While a lot of the design work and drafting is sub-contracted,
there are about150 fulltime designers using CAD,who are
engaged in designing in 2D and 3D,updating information,and
upgrading designs on the different projects. A digital system for
data management has been used since the late 1990s, with 5,700
active users registered on the system.The volume ofdata is
significant: The whole complex involves 100 million components;
with about1.5 million documents and drawings;1.6 million
individually registered assets;and almost2 million equipment
interventionslogged.The velocityof informationinvolves
about3,000 new pieces of equipmenta month,and 7,000 new
documentsand drawingscreated amonth.In work on an
accelerator in 2009 there were approximately 12,500 equipment
interventions permonth.This data is ofsignificantvariety,it
includes physics parameters, technical specifications, layouts and
equipment codes (in specification); simulations, bill of materials,
documents and drawings,(in design);manufacturing processes
and testprocedures and results (in manufacturing);installation
and safety procedures;and “as-installed”documentation (in
installation); and radiation measurements,material composition,
recycling procedures; and waste management (in dismantling).
There is substantialinterestin configuration management
within CERN with an ongoing initiative to update processes of
changecontrol.The manufacturingteam has a mature
understanding ofconfiguration management,and usesthe
work breakdownstructureto manageassetsthroughthe
supply-chain,and follow up changes.Manufacturing software
is linked to the digitalsystem fordata management,where
manufacturing data on critical equipment can be entered. There
is interestin improving thisfor the design data including
production drawings;and a desire to formalize the feedback
from operation and maintenance teams as input for new designs
and standardizing of parts. Challenges include the:
• Extended life-cycles – Installations have lifecycles of more
than 50 years, so they use historical data and formats, such
as microfilm drawings,and collectmissing information by
scanning and photographing the system.New assetinfor-
mation will be relied on by operators in the mid-to-late 21st
century, so needs to be self-explanatory and complete.
• Large, complex and advanced installation – Many pieces of
equipmentunique,designed specially,and the resultof
many years of research and development. New accelerators
and experiments must be installed into the overall system of
tunnels and facilities, for example a new accelerator reused
the main part of an old tunnel, with some interventions such
as ground construction work to upgrade it.
• Support for a scientific culture – The nature of the scienc
means tolerances may be measured in microns (one mic
is a thousandth of a milimetre) in an installation of sever
kilometers.Yet unlike the militarycontextin which
configuration managementwas developed,there is nota
centralised command and control culture. The ethos is b
around international research collaboration.
• Major operational constraints – There is no access to tun
and equipmentwhen scientific experiments are in progress.
Thereis a long shut-down formaintenance work every
1–2 years,with a shorttechnicalstop every 1–2 months.
Even then access is curtailed by the need to limit installe
maintainersexposure to radiation;and to cool-down and
warm-up equipment after and before experiments.
• Regulationsfor nuclearinstallations– Installers and
maintainers have to waitto enter certain zones because of
the radiation generated.Radiation effects are calculated for
each material. Where equipment or components are inst
or removed, these have to be tracked. CERN is classified
a nuclear installation,so regulations are similar to a power
plant,requiringtraceabledocumentationof equipment,
interventions and procedures.
• Parallel design work – Upgrades are managed by machin
versions(2007, 2009, 2012, 2015 etc.) as different
acceleratorconfigurationsare worked on in parallel.
Changesincludeplannedconfigurationchangesacross
versions;and interventions to fix things.When equipment
is installed,it needs to fitwith whatis there.Where there
have been unexpected changes,installers have had to cut
and weld equipment when installing it,modifying it on the
spot to make it fit.
There is particularfocus on managing such ‘non-confor-
mities’, where changes happen to the design during the us
testing,installing or modifying the equipment.Such changes
may resultin a design notbeing exactly within the specifica-
tion.How the equipmentdoes notconform to specification is
documented. Checks are made to establish whether equipm
can be used, accepted as it is, or corrected, and how it mig
corrected.Some equipmentis bespoke and extremely expen-
sive, so it may be accepted despite not conforming to spec
tion if it can be adapted on site.‘As-built’information is
required formaintenance and disposal,as well as the next
generationdesign,so test dataand new versionsof the
drawings need to be entered into the systems,along with the
location of the installed equipment in the tunnel.
4.3. Configuration management in Crossrail
Crossrail acts as the delivery client for a new railway, wh
is due to begin operation in 2018.This £15.8 billion complex
projectaims to hand-overa physicaland digitalrailway.It
involvesupgradingexistingrail networks;buildingnew
stations;boring 42 km of new tunnels across centralLondon
from Paddington Station to Liverpool Street Station; installi
extending and commissioning a wide array ofunderground
electricaland mechanicalsystems;and deliveringthe
346 J. Whyte et al. / International Journal of Project Management 34 (2016) 339–351

associated asset information to owners and operators. Contracts
with the supply-chain require delivery of assetinformation as
well as assets. There is a program of briefing the supply-chain
on how to deliver this; and the quality of received information
is benchmarked,initially every quarter,and now every
2 weeks.High quality assetinformation isrequired asthe
railway is expected to operate for over 100 years.
Configurationmanagementis importantin ensuringa
consistent,validated setof assetinformation asa project
deliverable.Through delivery there hasbeen a smallteam
focused on configuration management in a broader information
management team. They are involved in developing processes
for establishing and maintaining the integrity of configuration
items in the preparation for hand-over. Digital system is used to
manage data,where the projectmanages a significant volume
of data,expecting to generate 2–3 million recordsin asset
databases, 1 million model and drawing records; and quarter of
a million GIS records.Assetinformation is stored within a
repository,which is managed by data managementsoftware,
with document,modeland geographic information linked to
this,butrequiring differentsoftware to be viewed and edited.
There is hence a significant variety of linked data in the form of
assetinformation (and allthe associated variables,such as
author,approver,datesand versioning);digitaldocuments,
such as operationaland maintenance manuals,plans,require-
ments,2D designs;parametric building information models;
and geographic information, such as asset locations.
Physical assetsinclude rail-tracks,trains, shafts and
buildings. The contractor is responsible for providing labels to
identify assets as configuration items where practical, as well as
equipment/serialnumberlabels on allequipment,unless the
size restrictsapplication.The labellingof assetsenables
tracking of these items.Assets are also related to other assets
to represent vital ‘powered by’ or ‘controlled by’ relationships.
The associated information is controlled by contractors until
they deliver‘as-built’information to Crossrail.From there
onwardsin the lifecycle,Crossraillocksdown information
associated with configuration items,equipment/serialnumber
labels, and controls any further changes internally, to ensure the
integrity ofassetinformation in theirdigitalsystem.The
delivery client can introduce configuration items when design
has matured,from detailed design onwards.Crossrailapplies
configuration controlatthe ‘as-built’ stage so changes to the
configuration before handover are consistently maintained for
the owner.There are currently around 155 thousand client
configuration items, and by March 2016 the number is expected
to rise over 600 thousand. Challenges include:
• Rapid deploymentin a temporary organization – Interna-
tionalstandardsfor configuration and assetmanagement
were important,where obtaining the ‘buy-in’to establish
and maintain integrity is hugely challenging in the middle of
a complex mega project.
• Complexity and culture of delivery – There are conflicts and
interfacesamongthe many processesand procedures
currently used in the industry forchanges in programme
baseline,design management,red-line(marked-up)and
as-builtdrawings.There are challengesin understanding
theseand creatingthe culturethat get managersand
engineersto understand and usetheseexisting change
processes.
• Multiple change processes – the configuration manage
team has concluded that having multiple change proce
as is currently the case within the industry is notideal,
particularly where each change process may differ slig
and in some the same person is assessing the impactof
change and making the decision ofwhetherit should or
shouldn’t be rejected.
• Establishing a requirements-led orientation – There is a
need for a change from typicalmethods in the industry to
mitigatethe risk of scopecreep,by having assetsthat
conform to assetinformation,which in turn conform to
requirements.The red-line procedure,for example,covers
the process ofannotating changes to drawings afterthey
have been released forconstruction.The task,which is
conducted by the site contractor,is intended to highlight
approved,post-design changes from the originaldrawings,
as reflected by an inspector of the built asset. The ann
drawings are then transferred to Crossrail. These recor
used to laterupdate the as-builtdrawingsprior to data
handover to the operator.
• Determining operational requirements – There are ong
discussions with the future owners and operators regar
the type and format of asset information that will be us
operations;and a need to future-proofthis information
because of the long operational life.
An interface to the digital system has been reconfigure
facilitate this delivery ofthe large volume ofasset-specific
information.This specifieshow assetinformation isto be
identified,named,labelled,stored,synthesized and managed.
Additionalrules fornumbering and naming complementthe
use of internationalconfiguration managementstandardsto
provide robustmethods of maintaining assetinformation.By
following a standard structure,and definitions,the aim is to
organize asset information within the digital system and
it overto operatorsin ways thatwill be usefulto future
operation and maintenance. The interface helps users to
assetrequirementsto contractors,as well as capturethe
configuration items that those contractors return.Asset labels,
equipmentlabelsand serialnumbersare used to represent
configuration items,as defined in three differenthierarchies
based on location,function and classification.The system
allows metadata searching; and provides the ability to ex
link groups of assets to form a single system.
4.4. Approaches to managing change
While each organization,Airbus, CERN and Crossrail,
operates in a different industry, similarities in their appro
and experienceof using configuration managementreveal
shared characteristics and challenges ofmanaging change in
complex projects as assetinformation becomes a deliverable.
Table 3 summarises and compares the relationship betwe
347J. Whyte et al. / International Journal of Project Management 34 (2016) 339–351

projectdelivery and operations;the dataaggregation and
connectionsin projectdelivery;and approach to managing
change to provide asset information as a deliverable; and these
topics are discussed in turn below.
In summary,each ofthese organizations is interested in
configuration managementin operations as wellas delivery,
though they vary in the extentto which they have lifecycle
responsibilities. Because Airbus has service contracts it retains
an interestin aspects ofconfiguration management,such as
conformity,throughoutthe lifecycle foraircraftthatit has
designedand manufactured.CERN is an integrated
owner-operator, with responsibility for the whole lifecycle. As
the delivery client,Crossrailhands over assetinformation for
operation and maintenance. For each organization, the complex
products – aircraft, particle accelerators and railways – have a
long operationallife (more than 20,50 and 100 years)so
information on assets, such as material, provenance, and design
rationale,need to be available to enable efficientand safe
operation.
Projectdelivery involves managing change in assetinfor-
mation as data-sets are aggregated and re-used through life. The
volume,velocity and variety of data bring new challenges of
version control,linkages across projectstages and with other
data-sets; and ways of structuring and organizing. While there
is increasing integration between data-sets in project delivery,
digitalsystemsare notseamlessly integrated,but heteroge-
neous,with majortransitions in the use ofdata through the
projectlife. For example,in Airbus, there isa significant
transitionbetweenthe developmentand the production
environment,in which there ismanualwork to restructure
and re-link data (there may be instances where the product does
not change,butconnections between configuration items is
differentin engineering and production).All three organiza-
tions carefully manage the upload of information to their di
system because of the importance of its integrity. Where e
mentproviders have relevantassetinformation,these organi-
zationswill often duplicatethatinformation in theirown
systems rather than link to it, because they need to ensure
still be available in decades to come,when the manufacturer
may have differentequipmentfor sale and may notmaintain
legacy information.
Changeis managedthroughthe digital system,with
configuration managementsoftware providing workflows for
defining baselines for particular assets and assetsystems;and
for assigning roles and responsibilities for approving chang
This use of configuration management starts earlier in Airb
than in Crossrail, where the use of configuration managem
mostpronounced in the controlof as-builtinformation.Each
organization has a substantial and distributed supply-chain
has access to inputrelevantinformation,and to be involved
in the approvalof changewithin projectdigitalsystems.
Design globally distributed in both Airbus, with 4 million pa
supplied from 30 countriesfor A380,and CERN,with 80
countriesacross5 continentsinvolved in the LHC design.
Crossrail has a substantial supply-chain of contractors invo
in construction.Contractorswithin thissupply chain have
permissions to input ‘as-built’ asset information into Crossr
digital systems, with information approved for purpose befo
is made available.
Baselines are sometimes interpreted as an agreed descr
of the complex product system at a point in time. This stud
clarified the types of baselines now used in Airbus, CERN an
Crossrail increasingly focus on assets or groups of assets. T
is mostclearly articulated in Airbus,which uses functional,
Table 3
Comparing how change is managed as asset information becomes a deliverable.
AIRBUS CERN CROSSRAIL
Relationship between project delivery and operations
Organization type Service provider, with multiple customers.Integrated owner-operator Delivery client,handing over infrastructure
provider(s).
Responsibility Design,manufacturing,and servicing in
operation.
Full life-cycle. Design, construction and handover.
Information as a deliverableFor customer services to monitor
operational aircraft.
For maintenance and upgrades to
accelerators.
For station and railway operators and
maintainers.
Responsibility for operation
information
Final configuration of customer’s aircraft;
updating servicing information.
Asset tracking; work management
information; disposal information.
Asset information at hand-over; not
responsible for updating and maintaining.
Data aggregation and connections in project delivery
Interfaces in digital systemsBetween the development environment
and the production environment.
Across machine version; and between
manufactured and installed equipment.
Between the delivery client’s design and
as-built asset information.
Approach to managing change
Configuration managementUsed in the concept, product identification
and definition stages.
Principles mostfamiliarin manufacturing;
initiatives to reduce non-conformities.
Used to manage and control as-built
information; principles used in design.
Configuration items Agreed by manufacturing and engineering
through a ‘congruency process.’
Different configurations of the machine
managed in parallel.
Identified by the delivery client and labelled
by contractors.
Complex productsystem –
hierarchy and baselines
Functional baseline for requirements;
engineering baseline for design;
manufacturing baseline for production and
installation; Customer service baseline for
technical information.
Different generations of machines have
baselines that are managed in parallel,
making it important to get information about
unplanned changes back into the design
information.
Configuration items first identified at
detailed design stage.Contractors manage
the information up to an as-built baseline.
348 J. Whyte et al. / International Journal of Project Management 34 (2016) 339–351

engineering,manufacturing and customer service baselines to
manage related changes in relevantstages ofthe life-cycle.
Thus there isno longera simple baseline,but functional,
productand other baselines managed on differenttimescales.
Although in the 1950s configuration managementenabled a
move backwards to the baseline, in the era of big data this is not
straightforward.Re-baselining may notaccountfor different
evolutionsin data unlessthese are alllinked;and there is
significant work in industry to achieve this.
5. Discussion: renewed importance of configuration
management in an era of ‘big data’
For managers in these organizations thatdelivercomplex
projects,configuration managementhas become more,rather
than less,importantas we enteran era of‘big data’.New
challengesarise asassetinformation hasbecome a project
deliverable;as data increases in volume,velocity and variety;
and as itis aggregated and re-used;with connections(and
potentialconnections)acrossinternally and externally held
data-sets. The organizations perceive a greater need for control
through configuration management.The analyses suggestthe
need for integrity, in assets and in asset information, is a reason
for this renewed emphasis on and interestin the associated
controlprocesses,as complex projects manage a significant
volume and variety of assetinformation and hand this on to
ownersand operators.Ensuring integrity in operationsis
essential in industries that are regulated and safety-critical but
organizational complexity,large distributed supply-chains and
time-pressures increase the challenge of projects delivering the
asset information to support this.
Configuration managementhas its origins in the mid-20th
century. However, what is meant by configuration management
has changed significantly. There has been a shift from relatively
slow paper-based processes to fasterdatabase oriented prac-
tices; and extension of configuration management practices to
cover the life-cycle through articulation of multiple baselines
and aspects ofthe product.It is through such changes that
configuration managementhas increased in importance as an
approach to managing changein the delivery ofcomplex
projects,in an era of big data,rather than left behind with the
paper-based processes of the late 20th century.There is some
evidence that the speed of interaction with this data is growing,
with for example Crossrail increasing the frequency with which
it benchmarksits supply-chain from every three monthsto
every two weeks.This is notjustaboutdoing change control
better,butalso aboutproviding more visibility ofquality of
change controlby differentparts ofthe supply-chain.Such
changes are within the paradigm of configuration management:
they are beginning to be reflected in related standardsand
guides, with configuration management described as a compo-
nent of the project management information system (PMBOK
guide,2013);and,during the timeframe ofour study,the
military developing and releasing anew interim standard
(DOD, 2013),having considered theuse of a 3D model,
definition of as-designed, as-built and as-maintained baselines,
and definition of the product(Windham,2012).Thus change
management is no longer a paper-based process, as imp
Fig. 1, butit is predominantly concerned with digitaldata.
Digital workflows become important to manage the integ
of information; and conformity between requirements, sp
cations and asset information.
As previous research has shown that users often don’t
the prescribed processes involved in configuration manag
(Ali and Kidd,2014;Burgess etal.,2003;Kidd, 2001;Kidd
and Burgess, 2010), there are opportunities to consider w
new digitaltechnologiesmightenable otherapproachesto
managingchange.Airbus, CERN and Crossrailall face
challenges in implementing control processes; and are ac
engaged in developing new strategies. In complex projec
anticipate limits to the extent to which ‘big-data’ will bre
mould of established approaches to enable radically new
and flexible form of organizing envisioned by Levitt(2011).
However, we recognise the possibility of a broader shift a
from baseline planning,as has occurred in software projects
(Levitt,2011);or a complete transition away from principle
such asdecomposition and hierarchy,as advocated by the
military(Albertsand Hayes,2003).Otherapproachesto
managing change mightbe to mine data-sets to identify in-
formation relevant to the operational performance of ass
to seek new scalableapproachesto managing changein
non-criticaldocuments,where Wikipedia,mightsuggestsa
modelin which changes are made,and then corrected;which
contrastswith themorebureaucratic,pre-authorisation ap-
proach of configuration management.
While managing change is important to project manag
configuration management has had limited attention in li
on complex projects. Research on complex projects has i
discussed how systems integration capabilities are mobil
innovation in construction (Gann and Salter,2000);and in
manufacturing settings such as aircraftengine controlsystem
(Brusoni et al., 2001) and flight simulation (Miller et al., 1
This researchon systemsintegration,like the work on
configuration management, traces its history to the USA
programme in the 1950s (Sapolski, 2003). Change mana
through ‘configuration management’ relates to systems i
tion,as itinvolves the decomposition of the complex produ
system to identify assets, and then manage change withi
assets and their associated asset information.
6. Conclusions
While prior researchhas arguedthat digitally-enabled
approachesbreak themould of established approachesto
projectmanagement,enabling rapid,flexible forms of project
organizing; in this study we find Airbus, CERN and Crossr
using relatively hierarchical, asynchronous, sequential pr
es to managechange.We concludethatthe unstructured,
uncontrollednatureof ‘big data’ presentschallengesto
complex projects that deliver assets. Thus this paper con
by uncovering limits to flexibility where integrity is impor
The potentialto use ‘big data’in these contexts presents a
conundrum,similarto thatdiscussed by Lagoze in science.
While there is the potential for analytics to provide comm
349J. Whyte et al. / International Journal of Project Management 34 (2016) 339–351

advantage by revealing smallpatterns,‘big data’ represents a
paradigm shift that disrupts notions of integrity and force new
ways of thinking and doing to re-establish it. It challenges the
existing approachesto ensuring theintegrity ofassetsin
regulated and safety critical environments.
There are practical implications. The first is that as managers
in complex projects begin to deliver asset information, as well
as assets,they should expectchangesin both assetsand
associated asset information,and plan to manage this change.
The second is that managers should be aware of the challenges
thatan ‘era of big data’ presents to this process of managing
change.Configuration management provides a set of tools for
maintaining integrity in this context,and implementation of
configuration managementhas changed and is changing as a
result of digital technologies. As well as using these processes,
there may be contextsin which there are opportunitiesfor
managers to seek new proactive approaches to using data from
projects to understanding future scenarios. Managers seeking to
benefitfrom ‘big data’,mustdo this while maintaining the
validity ofthe information on which the delivery and main-
tenance of complex product systems rely.
There are also implications for research.This study returns
attention to Morris’ interests in the history of projectmanage-
ment; the centrality of change control to good project manage-
ment; and the shift from project execution to broader questions of
managementof projects.There are a number of directions for
further research. First, more needs to be done to understand the
idea of a baseline.Morris critiquesthe ethosof important
guidelines, such as the PMBOK Guide as to: “plan and then put
on cruise control” (Morris, 2013: p. 282), where this misses the
challenges at the front-end of projects; and the constant need for
updating and modifying plans during project delivery. We need
to understand more about the process of agreeing baselines; how
baselines are used to managing changes in the configuration of
assets,and how they are controlled across complex supply-
chains.Second complex productsystems may have different
hierarchicaldescriptions and hence more needs to be done to
understand how configuration items are identified.There are
questions aboutwhen and how the complex productsystem
becomes decomposed into assets thatare then controlled,and
also what information needs to be known about assets. Mapping
different approaches might lead researchers to set out frameworks
for understanding the kinds of change management that are most
effective in differentcircumstances.Third,more needs to be
understood aboutthe process of ensuring the validity of asset
information in digitalsystemsthatare constantly changing,
where responses are required rapidly.Here,researchers might
compare models, in which changes are made, and then corrected;
with the more bureaucratic,pre-authorisationapproachof
configuration management. The broader visibility and intercon-
nections between data-sets provided in an era of big data may
alter the utility of different approaches.
Finally, there are theoretical connections to be made between
the literatures.Our work hasrevealed particulardisconnects
between the literatures on configuration managementand the
strand of work on systems integration within the literature on
complex projects.Further studies might explore their historical
and contemporary interconnections between these,and situate
concepts within broader literatures on modularity and prod
architectures that may be useful in understanding change.
further research will continue to chart interconnections, de
by Morris,between the evolution of projectmanagementand
developmentsin systemsengineering,modern management
theory, and the evolution of the computer.
Conflict of interest
The university and authors have a strong research collab
ration with Crossrail, which is indirectly and directly involve
in other research, has funded consultancy and has connect
through advisory boards. We wish to confirm that there are
known conflicts of interest associated with this publication
there has been no significant financial support for this work
could have influenced its outcome.
Acknowledgements
The authors acknowledge the strong contributions ofthe
configuration managersinvolved in this research,Airbus,
CERN and Crossrail, the centres of configuration managem
competencein thesefirms, and the EPSRC Centre for
Innovative Manufacturing in Through-Life Engineering Ser-
vices (EP/I033246/1)at Cranfield University,which funded
this research as a feasibility study on “Configuration Manag
mentin Through-Life Engineering”.The research wascon-
ducted by the team as part of the Design Innovation Resea
Centre(EP/H02204X/1)and Technologiesfor Sustainable
Built Environments Centre (EP/G037787/1) atthe University
of Reading.
References
Alberts, D.S., Hayes, R.E., 2003. Power to the Edge: Command… Control…
the Information Age. CCRP Publishing.
Ali, U., Kidd, C., 2014.Barriersto effectiveconfiguration management
application in a projectcontext:An empiricalinvestigation.Int. J. Proj.
Manag. 32, 508–518.
Bachy, G., Hameri, A.-P., 1997. What to be implemented at the early stage
large-scale project. Int. J. Proj. Manag. 15, 211–218.
Barton, D., Court, D., 2012. Making Advanced Analytics Work For You. Harv
Bus. Rev. 90, 78–83.
Bersoff,E.H.,1984.Elements of Software Configuration Management.IEEE
Trans. Softw. Eng. 10, 79–87.
Billingham,V., 2008.Configuration management:Controlling yourproject
assets.In: Billingham,V. (Ed.),Projectmanagement:How to plan and
deliver a successful project. Studymates Ltd., Abergele.
BIM 2050 Group,2014.Built Environment2050:A Reporton Our Digital
Future. Construction Industry Council, London.
BIS/Industry Working Group,2011.Building Information Modelling (BIM)
Working Party Strategy Paper.GovernmentConstruction ClientGroup,
London.
Boyd,D., Crawford,K., 2011.Six Provocation forBig Data,A Decade in
InternetTime.Symposium on the Dynamics of the Internetand Society.
Oxford Internet Institute.
Boyd, D., Crawford, K., 2012. Critical questions for big data: Provocations f
cultural,technological,and scholarly phenomenon.Inf Commun Soc 15,
662–679.
350 J. Whyte et al. / International Journal of Project Management 34 (2016) 339–351

Braglia, M., Frosolini, M., 2014. An integrated approach to implement Project
ManagementInformation Systemswithin the Extended Enterprise.Int.
J. Proj. Manag. 32, 18–29.
Brouse,P.S.,2008.Configuration management.In: Sage,A. (Ed.),Systems
engineering and management for sustainable development. EOLSS, Oxford,
pp. 214–242.
Brusoni,S., Prencipe,A., Pavitt, K., 2001. KnowledgeSpecialization,
Organizational Coupling,and the Boundaries of the Firm: Why Do Firms
Know More Than They Make? Adm. Sci. Q. 46, 597–621.
Burgess,T.F., Byrne,K., Kidd, C., 2003.Making projectstatus visible in
complex aerospace projects. Int. J. Proj. Manag. 21, 251–259.
Burgess,T., McKee,D., Kidd, C., 2005.Configuration managementin the
aerospace industry:a review ofindustry practice.Int. J. Oper.Prod.
Manage. 25, 290–301.
Chen,H., Chiang,R.H.L., Storey,V.C., 2012.BusinessIntelligence and
Analytics: From big data to big impact. MIS Q. 36, 1165–1188.
Constantiou, I.D., Kallinikos, J., 2015. New games, new rules: big data and the
changing context of strategy. J. Inf. Technol. 30, 1–14.
Davenport, T.H., Barth, P., Bean, R., 2012. How ‘Big Data’ Is Different. Sloan
Manag. Rev. 54, 22–24.
Davies, A., Hobday, M., 2006. The Business of Projects: Managing Innovation
in Complex Productsand Systems.CambridgeUniversityPress,
Cambridge.
Davies,A., Mackenzie,I., 2014.Projectcomplexity and systems integration:
Constructing the London 2012 Olympicsand ParalympicsGames.Int.
J. Proj. Manag. 32, 773–790.
Davies, A., Gann, D., Douglas, T., 2009. Innovation in Megaprojects: Systems
Integration atLondon Heathrow Terminal5. Calif. Manag.Rev. 51,
101–125.
DOD, 1978.Configuration Control:Engineering Changes,Deviationsand
Waivers, DOD-STD-480A.
DOD, 2013.Interim Standard Practice:Configuration management.Depart-
ment of Defense (MIL-STD-3046(ARMY); AMSC 9275 AREA SESS).
Eisenhardt,K.M., Graebner,M.E., 2007. Theory buildingfrom cases:
opportunities and challenges. Acad. Manag. J. 50, 25–32.
Estublier,J., 2000.Software configuration management:A roadmap Confer-
ence on the future of software engineering. The International Conference of
Software Engineering, Limerick, Ireland, pp. 279–289.
Faucher, J.-B.P.L., Everett, A.M., Lawson, R., 2008. What do we know about
knowledge? In:Koohang,A., Harman,K., Britz,J. (Eds.),Knowledge
Management:TheoreticalFoundations.Informing Science Press,Santa
Rosa, CA, pp. 41–78
Floridi, L., 2012.Big Data and TheirEpistemologicalChallenge.Philos.
Technol. 25, 435–437.
Galbraith,J.R., 1973.Designing Complex Organization.Addison-Wesley,
Reading, MA.
Galbraith, J.R., 1977. Organization Design. Addison-Wesley, Reading, MA.
Galbraith, J.R., 2014. Organization Design Challenges resulting from Big Data.
J. Organ. Des. 3, 2–13.
Gann,D.M., Salter,A.J., 2000.Innovation in project-based,service-enhancedfirms:the construction of complex products and systems.Res.Policy 29,
955–972.
Gonzalez,P., 2002.A Guide to Configuration Managementfor Intelligent
Transportation Systems. Department of Transport, USA.
Hameri, A.-P., 1997. Project management in a long-term and global one-of-a-
kind project. Int. J. Proj. Manag. 15, 151–157.
Hameri, A.-P., Nitter, P., 2002. Engineering data management through different
breakdown structuresin a large-scale project.Int. J. Proj. Manag.20,
375–384.
Hameri, A.-P., Puittinen, R., 2003. WWW-enabled knowledge management for
distributed engineering projects. Comput. Ind. 50, 165–177.
Hobday, M., 1998. Product complexity, innovation and industrial organisation.
Res. Policy 26, 689–710.
Hobday,M., Davies,A., Prencipe,A., 2005.Systemsintegration:a core
capability of the modern corporation. Ind. Corp. Chang. 14, 1109–1143.
ISO, 2003. Quality managementsystems:Guidelinesfor configuration
management, BS ISO 1007:2003. BSI, London.
Kidd, C., 2001.The case for configuration management.IEE (Institution of
Electrical Engineers) (Review September).
Kidd, C., Burgess, R.G., 2010. Managing configurations and data for effe
projectmanagement.In: Morris,P., Pinto,J.K. (Eds.),Wiley Guide to
Projects, TechnologySupply-chainand ProcurementManagement,
pp. 108–123.
Kitchin, R., 2014. The Data Revolution:Big Data, Open Data, Data
Infrastructures and Their Consequences. SAGE Publications Ltd.
Lagoze, C., 2014. Big data, data integrity, and the fracturing of the cont
Big Data Soc. 1, 1–11.
Laney,2001.Data Management:Controlling DataVolume,Velocity and
Variety, Application Delivery Strategies.
LaValle, S., Lesser, E., Shockley, R., Hopkins, M.S., Kruschwitz, N., 2011.
data,analytics and the path from insight to value.Sloan Manag.Rev.52,
21–31.
Levitt,R., 2011.Towards projectmanagement2.0.Eng.Proj. Organ.J. 1,
197–210.
McAfee, A., Brynjolfsson, E., 2012. Big Data: The management of Revolu
Harv. Bus. Rev. 60–68.
Miles,M.B., Huberman,M.A., 1994.An expanded sourcebook:Qualitative
data analysis. Second ed. Sage Publications, London.
Military,1988.Configuration Control:Engineering Changes,Deviations and
Waivers, MIL-STD-480B.
Miller,R., Hobday,M., Leroux-Demers,T., Olleros,X., 1995.Innovation in
complex systems industries: the case of flight simulation. Ind. Corp. C
4, 363–400.
Moreira,M.E., 2010.Adapting configuration managementfor agile teams.
Wiley, West Sussex.
Morris, P., 1997. The Management of Projects. Thomas Telford, Londonfirst published in 1994, preface 1997).
Morris, P., 2013. Reconstructing project management. Wiley, Chichester
MPA, 2013.Are we any good atprojectcontrols -whatare the cross-sector
challenges for the future? Report of seminar 174 (1 Great George Str
Pinto,J., 2013.Lies,damned lies,and projectplans: Recurring human errors
that can ruin the project planning process. Bus. Horiz. 56, 643–653.
PMBOK guide, 2013. A guide to the project management body of knowle
5th edition. Project Management Institute, Newton Square, PA.
Sapolski, H., 2003. Inventing systems integration. In: Prencipe, A., Davie
Hobday, M. (Eds.), The Business of Systems Integration. Oxford Unive
Press, Oxford, pp. 15–34.
Shenhar,A.J., Dvir, D., 2007.Re-inventing projectmanagement.Harvard
Business School, Cambridge, MA.
Stake, R.E., 1995. The Art of Case Study Research. Sage, Thousand Oak
UK Government, 2013. Construction 2025. HM Government, London.
Viitanen,J., Kingston,R., 2013.Smartcities and green growth:Outsourcing
democratic and environmentalresilience to the globaltechnology sector.
Environ. Plann. A 45.
Whyte,J., Levitt,R., 2011.Information managementand the managementof
projects.In: Morris,P.W.G.,Pinto,J.K., Söderlund,J. (Eds.),The Oxford
handbook of project management. Oxford University Press, UK, pp. 3
Williams, T., 2009. Configuration management: Controlling change in co
projects.In: Williams,T. (Ed.), Construction Management:Emerging
trends and technologies.Delmar CengageLearning,New York,
pp. 177–190.
Williams, N., Ferdinand, N.P., Croft, R., 2014. Project management matu
the age of big data. Int. J. Manag. Proj. Bus. 7, 311–317.
Winch,G., 2010.Managing construction projects.2nd edition.Blackwell
Publishing, London.
Windham,J., 2012.DoD Military Standard forConfiguration Management.
ACDM Conference, Destin, FL.
Wozny, R., Black, K., Guess, V., 2014. The CMII/IPE Model and PLM/PDM
Tool: Functionality Needed to SupportImplementation.White Paper,
CMII-875A. CMII Research Institute.
Wu, X., Wu, G.-Q.,2014.Data mining with Big Data.IEEE Trans.Knowl.
Data Eng. 26, 97–107.
Yin, R.K., 1994.Case Study Research: Design and methods.Sage,Thousand
Oaks, CA.
351J. Whyte et al. / International Journal of Project Management 34 (2016) 339–351