KEY WORDS

IS AGILE PROJECT MANAGEMENT APPLICABLE TO

CONSTRUCTION?

Robert Owen1, Lauri Koskela2, Guilherme Henrich3 and Ricardo Codinhoto4ABSTRACT

This paper briefly summarises the evolution of Agile Project Management (APM) anddifferentiates it from lean and agile production and ‘leagile’ construction. The significantbenefits being realized through employment of APM within the information systems industryare stated. The characteristics of APM are explored, including: philosophy, organizationalattitudes and practices, planning, execution and control and learning. Finally, APM issubjectively assessed as to its potential contribution to the pre-design, design and constructionphases.

In conclusion, it is assessed that APM offers considerable potential for application in pre-design and design but that there are significant hurdles to its adoption in the actual constructionphase. Should these be overcome, APM offers benefits well beyond any individual project.KEY WORDS

Agile, Project Management, Construction

Salford Centre for Research and Innovation (SCRI), University of Salford, 4 Floor, Maxwell Building, TheCrescent, Salford, Greater Manchester, M5 4WT UK. Fax +44 161 2954587

1234

th

PhD ‘Candidate, Phone +44 161 295 4143, r.l.owen@pgr.salford.ac.ukProfessor, Phone +44 161 295 7960, l.j.koskela@salford.ac.uk

PhD Candidate, Phone +44 161 295 4143, g.henrich@pgr.salford.ac.ukResearch Assistant, Phone +44 161 295 4284, r.codinhoto@salford.ac.uk

52Robert Owen, Lauri Koskela, Guilherme Henrich and Ricardo Codinhoto

INTRODUCTION

The evolution of Agile project management (APM) has its foundations in the managementscience of Deming and is described more fully elsewhere (Owen and Koskela, 2006b). However,it is important to point out that APM, as it has evolved within information systems development,is synonymous neither with agile manufacturing, despite some common roots and similarcharacteristics, nor with ‘lean’.

Whilst some see agility as a state of mind, others focus on methodologies; those whoimplement ‘agile’ frequently confuse it with ‘lean’. In terms of manufacturing, lean and agileare different, as pointed out below (Sanchez and Nagi, 2001):

‘Lean manufacturing’ developed as ‘a response to competitive pressures with limitedresources. Agile manufacturing, on the other hand, is a response to complexity broughtabout by constant change. Lean is a collection of operational techniques focused onproductive use of resources. Agility is an overall strategy focused on thriving in anunpredictable environment. …… Flexible manufacturing systems (offer) reactiveadaptation, while’ agile manufacturing systems offer ‘proactive adaptation’.

Whilst some see agility as a state of mind, others focus on methodologies; those whoimplement ‘agile’ frequently confuse it with ‘lean’. In terms of manufacturing, lean and agileare different, as pointed out below (Sanchez and Nagi, 2001):

‘Lean manufacturing’ developed as ‘a response to competitive pressures with limitedresources. Agile manufacturing, on the other hand, is a response to complexity broughtabout by constant change. Lean is a collection of operational techniques focused onproductive use of resources. Agility is an overall strategy focused on thriving in anunpredictable environment. …… Flexible manufacturing systems (offer) reactiveadaptation, while’ agile manufacturing systems offer ‘proactive adaptation’.

In other words, to be agile an enterprise or project must be structured appropriately to proactivelyand quickly adapt to change, seizing such opportunities to enhance value outcomes. It shouldbe noted here that ‘lean construction’ contains some aspects of both lean and agile productionand the ‘Last Planner’ method (Ballard, 2000) can even be seen as partially agile. An alternativeand interesting view of merging lean and agile techniques (‘leagile’) has been proffered (Naimand Barlow, 2003); however, this only considers the ‘pull’ nature of agile customer demand onthe lean construction supply chain and does not embrace APM holistically. A summarycomparison of some of the similarities and differences between lean construction and APM isshown in Table 1.

Proceedings IGLC-14, July 2006, Santiago, Chile

Is Agile Project Management Applicable to Construction?53

Table 1: A Comparison between Lean Construction and Agile Project Management

UK Lean ConstructionToyota ProductionMethods/ Egan/Construction LeanImprovementProgrammeWaste Reduction &Bench MarkingSupply Chain

Relationship Change/Just In Time/PerformanceMeasurement/Customer Pull

Repeatability/ reliabilityPush/ Top Down

IGLC LeanConstructionToyota Production

Methods/ Koskela TFVTheory/ Theory of

Constraints/ ComplexityTheory/ SystemsThinking

Waste Reduction, Flow& Value

Agile ProjectManagementReaction to InformationSystems poorperformance/

Complexity TheoryEmergent Value & RapidFeedback

Evolved from

Key Tenets

Signature Methods

Collaborative Working

Embedded Customer/

& Distributed

Empowered, Multi-Management (Last

disciplinary Teams

Planner)/ Customer PullReliability

Partial (Design & LastPlanner) but evolvingPartial (Last Planner)but evolving

ReliabilityYes

Essential

Continuous Type 2Learning

Decisions Delayed UntilLast ResponsibleNoMoment

Yes

Metrics summarised in Boehm and Turner (2004), Shine (2003) and Stapleton (2003) showeddramatic changes in customer and business satisfaction in the information systems industrythrough the use of APM. Specific contributions included significant improvements inproductivity, quality, predictability and both development personnel and managerialorganisational skills. There was also a significant reduction in cost, though that was, in APMterms, a secondary effect.APM CHARACTERISTICS

APM can be expressed in terms of an underpinning philosophy; this philosophy affects howAPM is most effectively applied, whether in the field of information systems or elsewhere.Some outline rationales for adopting APM have already been explored (Owen and Koskela,2006b); there is potentially more to offer in construction than the application of ‘agile’ customerpull on the customisation of products (Naim and Barlow, 2003). The application of APM canbe described in terms of organisational attitudes and practices, its impact on project planningand execution, and its impact on methods of control and organisational learning, as shown inTable 2.

Theory

54Robert Owen, Lauri Koskela, Guilherme Henrich and Ricardo Codinhoto

Table 2: Project Management Method Comparators

Philosophy

Attitude to Chaordic Change

Management StyleOrganisation TypeWork Group StructureApproach to RiskNature of PlanningRequirements CaptureWork Package StructureDevelopment ApproachQuality ApproachCustomer Involvement

Value DeliveryProject MetricsAttitude to Learning

Organisational Attitudes &Practices

Planning

Execution

Control & Learning

PHILOSOPHY

The area of philosophy and metaphysics has been well reviewed in terms of production (Koskelaand Kagioglou, 2005). APM fits within the philosophy of process metaphysics through itsembracing the emergence of value as a process throughout the project life cycle. On the otherhand, traditional manufacture, construction and sequential project management tend towardsan emphasis on defining value up-front as something to strive to deliver.In terms of project management theory, APM is very much aligned to the discovery ofemergent business needs to provide value, primarily using management-as-organising (but ina very light-weight manner, i.e. facilitation). Project flow is typically managed in a regular(‘time boxed’) manner to ensure that value is generated continuously throughout the project.APM is strongly focussed on the scientific experimentation model as its primary controlmechanism. (Koskela and Howell, 2002a). In terms of practice, APM could almost be consideredas the opposite of a systems engineering approach, which relies on detailed knowledge elicitationand modelling up-front (Bonaceto and Burns, 2005).ORGANISATIONAL ATTITUDES AND PRACTICES

APM requires significant changes to traditional attitudes and practices, not least being theattitude to change itself.Attitude to Chaordic Change

It is significant that lean production and APM have different aims: lean seeks essentialrepeatability, whilst agile seeks essential reliability (Highsmith, 2004). An alternative view isthat lean focuses on efficiency and requires stability, whilst the agile process focuses oneffectiveness (Naim and Barlow, 2002). APM recognises change as inevitable and as an

Proceedings IGLC-14, July 2006, Santiago, Chile

Is Agile Project Management Applicable to Construction?55

opportunity to enhance delivered value (Owen and Koskela, 2006a); APM can be seen asharmoniously blending characteristics of both chaos and order, thus being ‘chaordic’ (Hock,2000).

On the other hand, change has traditionally been seen as a threat to conformance to plan.The difficulty with such a view is that we are unable to adequately describe an interactivesystem in the first place (Owen and Koskela, 2006a), so to try to persist in initial exhaustiverequirements definition is to ignore the opportunity to improve understanding throughout theproject.

Management Style

Traditional hierarchical management relies on the largely unidirectional flow of communication,thus denying opportunities for retrospection which are afforded through APM (Boehm andTurner, 2003).

The ‘command and control’ style of management used in many industries and projectsinhibits trust and hence reliability, whereas APM provides, together with facilitating leadership,an effective managerial motive force and fosters creativity.Organisation Type

In view of the stress on small, facilitated and empowered teams, McGregor’s theory Y(McGregor, 1960) practices of consensual management (Massie and Douglas, 1992) areobviously more relevant to agile than theory X, traditional western autocratic organisations.Ouchi’s theory Z (Ouchi, 1981) attempts to merge the best of theory Y into modern westernorganisations, adding a large amount of freedom and trust of workers. However, it also assumesthat workers have strong loyalty and an interest in team-working and in the organisation itself.Therefore, although theory Z ‘pragmatic oriental’ practices of collective decision making,employee-employer relationships and long-term employment organisations would prove anatural management fit with agile techniques such as Scrum (a Japanese-derived managementmethod – see Koskela and Howell, (2002a) and Schwaber (2004)), it fails to cross the culturaldivide inherent in many western enterprises.Work Group Structure

In APM the use of small, empowered, multi-skilled teams is a common trait. These teams relyon tacit knowledge and dense communications to create innovative solutions through swarmintelligence (Bonabeau and Meyer, 2001) and are typically self-managing; the project managerprovides leadership and facilitates progress. It can be argued that spontaneous, self-organisingteams would prove even more productive (Anderson and McMillan, 2003).

APM relies on a flatter, team-based structure rather than traditional close, hierarchicalmanagement. The removal of tiered management effectively removes communications protocoloverheads, as well as reducing unnecessary systems noise and the probability of compoundingerrors.

Approach to Risk

For apparent pragmatic reasons, organisations are often ‘risk averse’. As a consequence,financial, legal and insurance frameworks have evolved to ensure that risk is passed as far

Theory

56Robert Owen, Lauri Koskela, Guilherme Henrich and Ricardo Codinhoto

down the contractual food chain as possible (Pietroforte, 1997), resulting in companies divestingthemselves of organic expertise, as in much of the construction industry. Such risk distributionand management mechanisms are obviously causal in the development of adversarial businessrelationships.

Written evidence supplied to The House of Commons by IBM (Cornielle, 2004) states: ‘Asystem designed to pass off risk to one party, without also encouraging the other party to workto mitigate that risk, is doomed to failure’.

The alternative and agile approach is that risk is passed to whichever actor is most capableof resolving it, irrespective of initial contractual relationships. In other words, risk is apportionedfrom a total value maximisation perspective, rather than a(n apparent) local financial riskmanagement perspective. Such actions typify the network of trust which is necessary for trueteamwork.PLANNING

Most project managers will recognise the need to following a well-prepared plan, and theensuing fight to get back on plan when things go wrong. However, this is not the APM way.Nature of planning

Traditional project management tools expect a sequential plan to be prepared in detail for theentire project, and then to be rigidly followed. Deviation from plan is expected to be resolvedwith an overall aim to get back on plan.

APM also expects that a plan is prepared but at a level that is both realistic for the plannersto act on in the short term in order to deliver early value, and to mitigate risk for the entireproject. Where possible, decisions are delayed until the ‘last responsible moment’.Requirements capture

APM recognises that change is inevitable during a project and therefore embraces it as anopportunity for enhancing customer-perceived value. This is particularly important in the caseof information systems as they are so difficult to visualise, and: ‘We can not completely specifyan interactive system.’ - Wegner’s Lemma (Wegner, 1995)

However, many construction projects could also be categorised as interactive systems andresearch shows that significant uncertainty remains as to what is to be constructed, as late asthe start of construction. (Howell et al., 1993).

Traditional requirements capture can be shown to be less than optimal: ‘The definition anddissemination of initial objectives was not significantly related to the success or failure of aproject.’; and: ‘Successful projects were able, over their lifetime, to resolve the initial uncertaintyassociated with their technical and commercial goals and objectives.’ (Baker et al., 1986)APM recognises that changes throughout the project force scope control to be an ongoingtask: project scope should only be defined as far as we are currently truly able to comprehendand prioritise it, from the perspectives of value realisation and risk mitigation. We can thenuse project team (including the customer) learning for control and rapid feedback.Work package structure

Instead of the traditional task-based WBS being used to realise the project plan, APM utilises

Proceedings IGLC-14, July 2006, Santiago, Chile

Is Agile Project Management Applicable to Construction?57

its dynamic work backlogs. Typically, ‘stories’ are first developed to describe the wider businessvalue priorities, facilitate rough estimation and showing how value will be validated.

A prioritised backlog is then produced of value generation and risk mitigation tasks whichwill change during the project based on organisational learning. Each value generation or riskmitigation task is structured and sized in such a way that its package can be delivered within aregular and sustainable time(boxed) period (30 days for some APM methods) (this can beconsidered as analogous to Takt time in the Toyota Production System (Liker, 2004)).EXECUTION

Agile project management can be seen as ‘management as organising’ (Koskela and Howell,2002b), indeed, an agile project manager is very much seen as a facilitator who enables small,self-organising multi-disciplinary teams to decide for themselves how they satisfy their valuegoals.

Development approach

Traditionally requirements capture methods usually discourage further adaptation once theplan is running. After the requirements are described they are broken down and recompiledinto logical groups, often to create delivery milestones. It is obviously in the developers’ interestto strive to meet these milestones as payment is usually attached to them; change is seen asadding risk.

APM relies on incremental and iterative development with continuous learning beingessential to the evolution of the optimal value (to the customer) within the constraints of timeand cost. Thus, the ‘iron triangle’ of traditional project management is turned on its head, asshown in Figure 1 (after Cockburn, 2003).

Figure 1: Traditional versus APM

Quality approach

Compared with structured sequential (Royce, 1970) or ‘waterfall’ PM, APM was reported asdelivering defect rate improvements of 61% in two case studies (Bowers et al., 2002), whilst83% of 131 respondent companies in an online survey reported better or significantly better

Theory

58Robert Owen, Lauri Koskela, Guilherme Henrich and Ricardo Codinhoto

quality (Shine, 2003). The reasons for improvement in both defect rates and perceived qualityare not fully understood; however, it is probable that defects are caught and corrected muchearlier because of the nature of APM teams, work structures and feedback mechanisms. APMalso concentrates on evolving customer perceptions, rather than conformance to an early plan.Customer involvement

Every project manager must see customer involvement as vital at the requirements collectionand specification stage, and most would carry this through to the design phase. In the author’sown experience, many managers of sequential development projects act as if customerinvolvement at later stages of production is a necessary but irritating obstruction to efficientcompletion of the plan; such involvement must be managed and minimised.

However, APM emphasises that customer involvement right through the project is key tothe organisational learning required to iteratively and incrementally produce the best possiblevalue yielded through Type 2 learning. The result of such change must be reflected in thereported 83% improvement in business satisfaction from using APM (Shine, 2003).Value delivery

A major focus of APM is the early and sustained delivery of value, as seen by the customer orstake holders.

At the end each timebox recognisable (by the customer) value must be delivered; feedbackand learning are core to the dynamic realisation of customer value. At the end of the projectthe customer has received what they by then realise are their dynamically prioritised valuedeliveries, rather than what the supplier and they would have originally identified undertraditional processes. The structure of APM value delivery is shown at Figure 2. This contrastsstarkly with the traditional approach of value residing with the developer/ contractor until thecustomer is prepared to accept the phase/ project as complete.

Figure 2: APM Value Delivery

Proceedings IGLC-14, July 2006, Santiago, Chile

Is Agile Project Management Applicable to Construction?59

CONTROL AND LEARNING

APM offers the possibility of process improvements in the use of metrics for project control.However, the area of organisational learning offers even greater potential, both during theproject and in the medium and long term.Project metrics

Whilst measurement is never waste if it enhances value or learning, excessive measurementcan interfere with work and metrics should both be meaningful and have context. Excessiveemphasis on data collection (Armour, 2006) not only adds waste to the production process butalso can lead to information overload to such an extent that analysis paralysis is reached andderived information therefore becomes stale and obsolete before it is available.

On the other hand, APM relies on minimal, accurate and up-to-date metrics, displayed as‘big visible pictures’, to facilitate project control at a glance. An exemplary example of this isthe Toyota product development environment, Oobeya (big room in Japanese) (Tanaka, 2005)which consolidates meaningful metrics for control and involves every functional area in oneplace.

Attitude to learning

Traditional project management employs Model 1 single loop learning which ‘make(s)organizational assumptions and behavioural routines self-reinforcing – inhibiting “detectionand correction of error” and giving rise to mistrust, defensiveness and self-fulfilling prophecy’(Edmondson and Moingeon, 1999)

On the other hand, Model 2 learning organizations have governing values which include:valid information, free and informed choice and internal commitment. Model 2 strategies are:sharing control, and participation in design and implementation of action. Model 2 learningorganisations are rare, though they foster double loop learning, as shown in Figure 3. (Argyrisand Schön, 1996)

Figure 3: Single & Double Loop Learning

Agile project management has emerged from double loop learning (Model 2), i.e. byquestioning the governing variables such as methodologies. Double loop learning continuesthroughout the agile project through the formal use of iterative development and through theinformal learning inherent in small interactive multi-disciplinary teams. Any Model 2organisation relies on the sharing of control, design and implementation of action, using

Theory

60Robert Owen, Lauri Koskela, Guilherme Henrich and Ricardo Codinhoto

minimally defensive relationships and is in direct contrast to the traditional command andcontrol management system and waterfall development methodologies.

As with the Toyota Production System (Liker, 2004), it is essential for an agile workforce andtheir employers to have long-term bi-lateral commitment and for the workforce to be encouragedand trained in the application of Model 2 learning.

Since the 1970s, particularly in the construction industry, main contractors have employedless labour to give them more flexibility and to offset risk (Pietroforte, 1997). Employeesgenerally no longer benefit from the in-depth training and industrial fidelity inculcated bytraditional apprenticeships. This mobility and de-skilling of the workforce is an obviousimpediment to agile learning.

POTENTIAL AREAS FOR IMPROVED AGILITY WITHIN CONSTRUCTIONWithin the construction industry, it is possible to consider the pre-design, design and actualconstruction phases separately. These have been mapped against the APM analysis, above, inorder to assess the degree to which APM might be useful to the industry.PRE-DESIGN PHASE

In the pre-design phase of a construction project, the three prime issues are: conceptdevelopment; planning covering procurement strategy, time and cost; and the preparation of abrief (Best and de Valence, 1999). The contents, organization and management principlesused in the pre-design phase vary considerably across projects and client organizations, evenif (especially larger) clients have created standard procedures for this phase. There is oftenconsiderable complexity in the pre-design phase (Pennanen and Koskela, 2005). This phasebeing a basis for subsequent phases, the output of pre-design should be comprehensive andintegrated (Morris, 1991), as well as consistent.

The key findings in prior literature on the pre-design phase suggest that, in practice, theapproach in pre-design tends to be either too programmatic or too chaotic (following theadhocracy prescription of Mintzberg (1983). This results in incomplete, inconsistent orotherwise suboptimal guidance for the subsequent phases of the project.

Whether agile principles are applicable in the pre-design phase is discussed based on threecriteria, presented in diminishing order of validity:

•••

Agile principles have successfully been used, implicitly or explicitly.Problems have been identified in prior literature, to which agile principlesarguably provide a solution.

Agile principles can be argued to be applicable, based on general knowledgerelated to the pre-design phase.

In the following, the application of a number of agile principles to the pre-design phase isanalysed, based on the fore-mentioned principles. The starkly contextual nature of any statementand advice should be noted.

Philosophy. In the pre-design phase, a considerable number of issues are in a flux, and thewhole process is emergent. Thus, process metaphysics can advantageously be used as a basisfor conceptualizing this phase.

Proceedings IGLC-14, July 2006, Santiago, Chile

Is Agile Project Management Applicable to Construction?61

Attitude to Chaordic change. New opportunities constantly emerge (Blomberg, 1998) andnew risks are constantly identified; thus the situation is characterized by chaordic change.Management style/work group structure. It is advisable to organize through an empoweredteam any large and complex pre-design effort, with frequent mutual communication.Hierarchical decision making has been found to cause problems, for example in the pre-designphase of primary healthcare facilities (Tzortzopoulos et al., 2006).

Customer involvement. As requirements capture is a central task in the pre-design phase,customer involvement is clearly highly recommended, if not essential.

Nature of planning. Due to the complexities and uncertainties involved in the pre-designphase, anything other than lightweight planning is probably nugatory. Indeed, Blomberg (1998)finds very little formal planning in the early phases of successful projects.

Development approach. Due to the needs for integration and customer involvement, aniterative and incremental development approach can be – and is often recommended for thepre-design phase.

Requirements capture. The distinction between stable requirements (to be captured up-front), volatile requirements (for which options need to be kept open) and evolving requirements(for which learning is to be cultivated) is highly relevant in construction projects. Consequencesof failures to categorize requirements in this way and reliance on immature requirements arereported by Tzortzopoulos, et al. (2006).

Arguably, agile principles and methods promise the potential of an improved approach forthe pre-design phase, being simultaneously appropriately structured but also flexible enoughto allow opportunities to be seized and creative solutions to be devised.DESIGN

Design is the intermediate phase where the concept generated during the pre-design phasewill be developed and transformed into solutions (specifications and prescriptions) to guideconstruction, operation and maintenance of the building (Kagioglou et al., 1998). As such,two main issues emerge: the integration between design and production, and the dynamicprocess of requirements capture.

Despite the possibility of generalizing the main phases of the design process, the contentdeveloped during this phase varies from project to project, and is also varied through iterationinherent to designing. It is to these two key issues that it is believed that APM can contribute value.Philosophy. In the design phase, contemporary methods and approaches, such as ConcurrentEngineering and Last Planner are essentially based in delivering value throughout the process(Kamara et al, 1997); (Codinhoto, 2003). Issues regarding the identification of trade-offs,processes of analysis and synthesis, and also decision-making are in flux. Therefore, processmetaphysics is the appropriate basis for conceptualizing this phase (Koskela; Kagioglou, 2005).Organisational Attitudes and Practices. The construction industry, in general is characterizedby the establishment of a new team of companies for each new project. Therefore, the designteam varies from project to project and the categories Y, X and Z can not easily be applied inconstruction. However, it is possible to say that some Type Z characteristics e.g. collectivedecision making and improved employee-employer relationships can be observed in somelong-term partnerships (Kamara; et al, 1997); also observed in the Heathrow airport Terminal5 project (College, 2005).

Theory

62Robert Owen, Lauri Koskela, Guilherme Henrich and Ricardo Codinhoto

Regarding iterative and incremental development of value, it seems that this is a naturalprocess in the design phase. However, to delay decisions to the ‘last responsible moment’would prove problematic in the construction setting, as currently structured with its discretephases. Such an attitude might also imply difficulties in product development coordination(Clark and Fujimoto, 1991). The design process is highly interactive and before any changethe design team should consider the impact of the change on the product, and also on thedesign process itself (Crawford and Benedetto, 2000).

Planning. Design planning has been considerably investigated in manufacturing and inconstruction. Approaches, methods and tools vary greatly. Therefore, there is a range of solutionsthat fit into one of the two categories “light” and “heavyweight”. On the one hand, DesignStructure Matrix (DSM) and Analytical Design Plan Technique (ADePT) (Austin et al., 2000)constitute examples of heavyweight planning. On the other hand Last Planner (Ballard, 2000)can be considered lightweight.

Regarding the client requirements capture in the design phase, it seems that a considerablenumber of methods, e.g. Quality Function Deployment (QFD) are focused in detailing therequirements exhaustively and at the beginning. Studies as presented (Miron and Formoso,2003) show that there is still a gap regarding the process of requirements capture during thewhole design phase, indeed, research shows that, as late as the start of construction, significantuncertainty remains as to what is to be constructed (Howell et al., 1993).

Finally, the use of work breakdown structures is the current basis of work package structuresin construction. The division of the tasks according to products and sub-products to be deliveredis a common practice in construction. The process protocol developed by Salford University(Kagioglou et al., 1998) is one example.

Execution. Regarding the approach to development, the design phase can be either sequentialor iterative. The adoption of one or another will vary according to the project as discussed byKamara et al (1997). On the one hand, the adoption of iterative approaches will result infrequent value delivery for clients. On the other hand, sequential approaches are characterizedby product delivery at the end; as a consequence errors and corrections are frequent (Prasad,1996). Also quality is delivered considering both the perception of value by the customer andother stake holders, and defect reduction (e.g. Design for Manufacturing and constructabilityanalysis). Client involvement during the design phase is common practice in construction.Control and Learning. The construction design is constantly measured according to differenttypes of metrics, for instance, cost, maintainability and sustainability of solutions. However,the relations between different metrics are not completely understood. Regarding attitude tolearning, the process will vary within each new project. Therefore, it is substantially based onchange management within the temporary organisation, with knowledge retained largely atthe individual, rather than wider organisational level.

In conclusion, the adoption of APM principles in the design phase is very appropriate tothe challenges that face the construction industry, for instance, the development of high qualityand complex products at lowest possible cost. However, its adoption will vary according withthe complexity and uncertainty related to the project; it would be particularly appropriatewhere solutions to requirements evolve or are likely to change through the project. Therefore,the projects that will gain more benefit from APM are projects in which a considerable numberof clients are involved, requirements are conflicting and constantly generate trade-offs, andearly delivery of value is a priority.

Proceedings IGLC-14, July 2006, Santiago, Chile

Is Agile Project Management Applicable to Construction?63

CONSTRUCTION

The construction phase differs from design in some relevant ways; these will be pinpointed toanalyze the applicability of APM in construction. Firstly, we have to be aware that theconstruction phase utilises a wider disparity of employers and employees, this workforce isalso one of the most poorly prepared in terms of professional qualifications and have amongstthe lowest comparative salaries (Koch, 2005). Thus, trying to apply new managementmethodologies in construction, we are confronted by a large culture problem which mustchange in order to enable training and learning to achieve multi-skilled and self-managingteams, as proposed by APM philosophy. Furthermore, construction is usually characterized byhaving a great number of sub-contractor and casual workforces; this is a significant impedimentto inculcating strong loyalty from workers (Howell and Koskela, 2000). Despite all theseconstruction culture problems, the authors believe that there is room for use of APM inconstruction on the site level, at least for planning, when managers can respond quickly to anychange that might occur in the scope of the project.

However, construction shows considerable inertia towards cultural evolution. Therefore,at the lower levels of execution, an APM implementation is likely to be harder because theconsequences of changes can cause big impacts and their cost can be too high for this non-coherent workforce to assume. But, as mentioned previously, APM applicability will alsodepend on the scale of the project and the organisation type. Construction has started to moveforward with a positive change in philosophy by taking into consideration more human aspectswithin some production management methods (e.g. the Last Planner System (Ballard, 2000)).The APM approach relies on up-front value definition, and early and continuous deliveryof value to customers. In design APM fits perfectly, but in the execution phase it is morecomplicated to achieve as there are many more interdependent activities. In brief, the authorsbelieve in the APM concepts and think that it could be a powerful tool for construction managers,mainly for planning in the production phase of construction. However, for managingconstruction execution, a great amount of effort would be needed, beginning with a culturechange within the sector.CONCLUSIONS

APM revolves around the embracing of change as an opportunity for improved, early andsustained value delivery; it requires organisations to be more proactive than lean organisations.To achieve this, stake holders, including the client/customer utilise the power of organisationallearning. However, for the organisation to learn organically it is vital to employ a highlytrained and team-based workforce to use their swarm intelligence; bi-directional loyalty andmutual long-term commitment are necessary for such learning.

There seems to be considerable potential for gains to be made from the adoption of APMin the pre-design and design phases of construction; iterative and incremental developmentcan facilitate creative solutions, particularly to complex and uncertain requirements. However,the fractured and temporary nature of the actual construction organisation is likely to impedethe desirable continuation of these practices through to construction and support.

A construction industry restructured to embrace a more highly trained and motivated workforce should yield an improvement in customer-perceived value delivery, forming the basis oflong-term trust networks. The bid-to-win ratios of construction companies would then be

Theory

64Robert Owen, Lauri Koskela, Guilherme Henrich and Ricardo Codinhoto

improved; reducing the feast and famine nature of each company’s own contracts, and formingthe basis of an iterative and incremental improvement of fortunes for such innovators.

Whether or not it is adopted within construction, APM acknowledges that change isinevitable. Agile Project Management delivers real benefits to those organisations which thriveon change, and which foster a culture where workers can contribute to organisational learning(and hence, profitability).REFERENCES

Anderson, C.; McMillan, E. (2003) “Of Ants and Men: Self-organized Teams in Human andInsect Organizations”, Emergence: Complexity and Organization, 5, (2), pp. 29-41.

Argyris, C.; Schön, D. A. (1996) “Organizational Learning 2 : Theory, Method and Practice”,Addison-Wesley, Reading, Mass ; Wokingham, pp. xxix, 305p.

Armour, P. G. (2006) “Counting Boulders and Measuring Mountains”, Communications ofthe ACM, 49, (1), pp. 17-20.

Austin, S.; Baldwin, A.; Li, B.; Waskett, P. (2000) “Analytical Design Planning Technique(ADePT): A Dependency Structure Matrix Tool to Schedule the Building Process”, ConstructionManagement and Economics, 18, pp. 173-182.

Ballard, G. (2000) “The Last Planner System of Production Control”, Thesis (Doctor ofPhilosophy), The University of Birmingham, Birmingham.

Beck, K.; et al (2001) “Manifesto for Agile Software Development” [online]. Available from:http://www.agilemanifesto.org/ [Accessed 9 Jan 2006].

Best, R.; de Valence, G. (1999) “Getting it Right at the Start”, Building in Value: Pre-designIssues Best, R. and de Valence, G.(Eds.). London, pp. 1-9.

Blomberg, J. (1998) “Myter om Projekt”, Nerenius and Santerus, Stockholm.

Boehm, B.; Turner, R. (2003) “People Factors in Software Management: Lessons FromComparing Agile and Plan-Driven Methods”, Crosstalk - The Journal of Defense SoftwareEngineering, (Dec 2003).

Boehm, B. W.; Turner, R. (2004) “Balancing Agility and Discipline : A Guide For ThePerplexed”, Addison-Wesley, Boston ; London.

Bonabeau, E.; Meyer, C. (2001) “Swarm Intelligence: a Whole New Way to Think AboutBusiness”, Harvard Business Review, (May), pp. 107-114.

Bonaceto, C.; Burns, K. (2005) “Using Cognitive Engineering to Improve SystemsEngineering”, Mitre Technical Paper [online]. Available from: www.mitre.org/work/tech_papers/tech_papers_05/05_1361/05_1361.pdf [Accessed 30 May 2005].

Bowers, J.; May, J.; Melander, E.; Baarman, M.; Ayoob, A. (2002) “Tailoring XP for LargeSystem Mission Critical Software Development”, Lecture Notes in Computer Science, 2418,(2418/2002), pp. 100-111.

Clark, K. B.; Fujimoto, T. (1991) “Product Development Performance: Strategy, Organizationand Management in the World Auto Industry”, Harvard Business School Press, Boston, Mass.Cockburn, A. (2003) “Process: the Fourth Dimension (Tricking the Iron Triangle)”, Humansand Technology [online]. Available from: http://alistair.cockburn.us/crystal/articles/ptfd/processthefourthdimension.htm [Accessed 10 Jan 2006].

Codinhoto, R. (2003) “Guidelines for Integrated Management of the Design and ProductionProcesses in Construction Projects”, Dissertação de Mestrado (Mestrado em Engenharia Civil)– Núcleo Orientado para a Inovação da Edificação, Programa de Pós-Graduação em EngenhariaCivil, Universidade Federal do Rio Grande do Sul, Porto Alegre, pp. 176.

Proceedings IGLC-14, July 2006, Santiago, Chile

Is Agile Project Management Applicable to Construction?65

College, B. (2005) “Relational Contracting - Creating Value Beyond the Project”, LeanConstruction Journal, 2, (1), pp. 30-45.

Cornielle, J. (2004) “Memorandum Submitted by IBM (SC 15)”, Select Committee on Workand Pensions Written Evidence.

Crawford, C. M.; Benedetto, C. A. (2000) “New Products Management: International Edition”,Macgraw Hill, Boston, Mass.

Edmondson, A.; Moingeon, B. (1999) “Learning, Trust and Organisational Change”,IOrganizational Learning and the Learning Organization : Developments in Theory andPractice Easterby-Smith, M.;Burgoyne, J. and Araujo, L.(Eds.). SAGE, London, pp. viii, 247p.Highsmith, J. A. (2004) “Agile Project Management : Creating Innovative Products”, Addison-Wesley, Boston, pp. xxvi, 277 p.

Hock, D. (2000) “The Art of Chaordic Leadership”, Leader to Leader, 15, (Winter 2000), pp.20-26.

Howell, G. A.; Koskela, L. (2000) “Reforming Project Management: The Role of LeanConstruction”, 8th Annual Conference on Lean Construction, Brighton, UK. IGLC.

Howell, G. A.; Laufer, A.; Ballard, G. (1993) “Uncertainty and Project Objectives”, ProjectAppraisal, 8, (1), pp. 37-43.

Kagioglou, M. Cooper, R. Aouad, G. Hinks, J. Sexton, M., et al. (1998) “Final Report: GenericDesign and Construction Process Protocol”, University of Salford, UK.

Kamara, J. M.; et al (1997) “Considerations for the Effective Implementation of ConcurrentEngineering in Construction”, Concurrent Engineering in Construction, London. The Institutionof Structural Engineers, pp. 33-44.

Koch, C. (2005) “Failures in Building Operations - An Occasion for Learning?” InternationalConference on Operations and Global Competitiveness - EurOMA, Budapest, Hungary.Koskela, L.; Howell, G. A. (2002a) “The Theory of Project Management: Explanation to NovelMethods”, 10th International Conference on Lean Construction, Gramado, Brazil. IGLC.Koskela, L.; Howell, G. A. (2002b) “The Underlying Theory of Management is Obsolete”,PMI Research Conference, PMI, pp. 293-302.

Koskela, L. J.; Kagioglou, M. (2005) “On the Metaphysics of Production”. 13th InternationalConference on Lean Construction, Sydney. IGLC.

Liker, J. K. (2004) “The Toyota Way : 14 Management Principles From The World’s GreatestManufacturer”, McGraw-Hill, New York ; London, pp. xxii, 330 p.

MacCormack, A.; Kemerer, C. F.; Cusumano, M.; Crandall, B. (2003) “Exploring TradeoffsBetween Productivity and Quality in the Selection of Software Development Practices”, IEEESoftware, pp. 75-85.

Massie, J. L.; Douglas, J. (1992) “Managing : a Contemporary Introduction”, Prentice Hall,Englewood Cliffs, N.J, pp. 540 p.

McGregor, D. (1960) “The Human Side Of Enterprise”, Penguin, Harmondsworth, pp. xiv,246p.

Mintzberg, H. (1983) “Structure in Fives: Designing Effective Organizations”, Prentice HallInternational, Englewood Cliffs, pp. 312.

Miron, L. I. G.; Formoso, C. T. (2003) “Client Requirement Management in Building Projects”,11th International Conference on Lean Construction, July 2003 Blacksburg, Virginia. IGLC.Morriss, P. (1991) “Project Specification and Initiation”, Guidelines for the Management ofMajor Construction Projects, National Economic Development Office, HMSO, London.

Theory

66Robert Owen, Lauri Koskela, Guilherme Henrich and Ricardo Codinhoto

Naim, M.; Barlow, J. (2003) “An Innovative Supply Chain Strategy for Customized Housing”,Construction Management and Economics, V, pp. 593-602.

Ouchi, W. G. (1981) “Theory Z : How American Business Can Meet the Japanese Challenge”,Addison-Wesley, Reading, Mass.

Owen, R. L.; Koskela, L. (2006a) “Agile Construction Project Management”, 6th InternationalPostgraduate Research Conference in the Built and Human Environment, 6/7 April 2006 Delft,Netherlands. Research Institute for the Built and Human Environment, University of Salford.Owen, R. L.; Koskela, L. (2006b) “An Agile Step Forward in Project Management”, 2ndSpecialty Conference on Leadership and Management in Construction and Engineering, 4 - 6May 2006 Grand Bahama. CIB / ASCE.

Pennanen, A.; Koskela, L. (2005) “Necessary and Unnecessary Complexity in Construction”,1st International Conference on Built Environment Complexity (Becon), 11-14 September2005 University of Liverpool.

Pietroforte, R. (1997) “Communication and Governance in the Building Process”, ConstructionManagement and Economics, 15, (1), pp. 71-82.

Prasad, B. (1996) “Concurrent Engineering Fundamentals: Integrated Product and ProcessOrganisation”, Prentice Hall, New Jersey.

Royce, D. W. W. (1970) “Managing the Development of Large Software Systems”, Proceedings,IEEE WESCON, (August), pp. 1-9.

Sanchez, L. M.; Nagi, R. (2001) “A Review of Agile Manufacturing Systems”, InternationalJournal of Production Research, 39, (16), pp. 3561 - 3600.

Schwaber, K. (2004) “Agile Project Management with Scrum”, Microsoft Press, Redmond,Wash., pp. xix, 163 p.

Shine (2003) “Shine Technologies Agile Methodologies Survey Results” [online]. Availablefrom: http://www.shinetech.com/dosearchsite.action?quickSearch=true&searchQuery.queryString=agile+survey [Accessed 20December 2005].

Stapleton, J.; Consortium, D. (2003) “DSDM : Business Focused Development”, Addison-Wesley, Reading, Mass. ; Harlow, pp. xxvii, 239 p.

Tanaka, T. (2005) “Quickening the Pace of New Product Development” [online]. Availablefrom: http://www.toyota-engineering.co.jp/qvsystem-eng.htm [Accessed 7 December 2005].Tzortzopoulos, P.; Cooper, R.; Chan, P.; Kagioglou, M. (2006) “Client Activities at the DesignFront End”, In Design Studies. In Print.

Wegner, P. (1995) “Interactive Foundations of Object-Based Programming”, IEEE Computer,28, (10), pp. 70-72.

Proceedings IGLC-14, July 2006, Santiago, Chile