Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy Available online at www.sciencedirect.com International Journal of Project Management 28 (2010) 756 – 765 www.elsevier.com/locate/ijproman Knowledge integration at the edge of technology: On teamwork and complexity in new turbine development Cecilia Enberg, Lars Lindkvist ⁎, Fredrik Tell KITE, Department of Management and Engineering, Linköping University, Sweden Received 2 February 2010; accepted 25 May 2010 Abstract This paper takes an empirical point of departure in the development of a new steam turbine. Project work here relied on a process of iteration between a small core group of team members with extensive experience and team members with less of that currency. In this project, the core group had a major integrative role, whereas other team members were mainly responsible for the specific tasks assigned to them. Quite a few of the latter category felt uneasy about their role and felt ‘decoupled’ from the project. In our analysis we use the Teamwork Quality (TWQ) construct proposed by Hoegl and Gemuenden (2001). In conclusion, our findings suggest that in highly complex projects of this type, team-based knowledge integration need not presuppose equality of participation and we introduce the notion of a Segregated Team to account for these findings. © 2010 Elsevier Ltd. and IPMA. All rights reserved. Keywords: Knowledge integration; Product development; Projects; Teamwork 1. Introduction Product development projects typically call for the integration of a wide range of specialised knowledge bases. Moreover, it has been suggested that the degree of integration of such dispersed knowledge can explain differences in product development performance and that the firm's knowledge integration capabilities can be vital to attaining competitiveness (Carlile and Rebentisch, 2003; Hoopes, 2001). However, the question of how knowledge integration is enabled in different types of project settings has not received much attention. Instead, there has been a tendency to treat all projects as similar and equal (Shenhar, 2001; Sauser et al., 2009). Carlile and Rebentisch (2003, p. 1182) suggest that “current frameworks of knowledge transfer and integration do not apply with equal explanatory power to both simple and complex knowledge integration tasks” and that those frameworks which exist seem ⁎ Corresponding author. E-mail addresses: [email protected] (C. Enberg), [email protected] (L. Lindkvist), [email protected] (F. Tell). mainly to mirror situations which would be characterised as being rather simple and comprising limited uncertainties. In this vein, the specific purpose of this paper is to explore the issue of team-based knowledge integration in a product development project — the Turbine project — characterised by high levels of uncertainty and complexity. The project managers and project members repeatedly stressed that the task assigned to them involved “working at the edge of technology” or “at the edge of what is possible”. [In this project] you are really at the leading edge of technology and knowledge. […] The nature of work is such that you always try to push the frontiers of what is possible. (Experienced project member, aero) Although, their subjective opinions cannot be taken as a fact, our case study observations largely substantiate them. Among the risks associated with the project were limited knowledge about issues that concerned extensive vibration excitation at high volume flows and only partly validated tools for aerodynamic calculations. In order to develop the new steam turbine, many different domains of technical expertise were needed, ranging 0263-7863/$ - see front matter © 2010 Elsevier Ltd. and IPMA. All rights reserved. doi:10.1016/j.ijproman.2010.05.003 Author's personal copy C. Enberg et al. / International Journal of Project Management 28 (2010) 756–765 from knowledge about mechanical integrity and aerodynamics to blade design, design layout and testing. Project members were strongly specialised and all of them had deep knowledge of their domain of expertise. One basic challenge facing the project was how to handle the contradictory demands of aerodynamic efficiency and mechanical integrity (MI). Thus, apart from working at the edge of their competences, project members from ‘aero’ and ‘MI’ had to manage a difficult trade-off, which could not be resolved in a classical optimization manner. 2. Team-based knowledge integration Turning first to the issue of integration, we should first acknowledge its very long history in the literature. Lawrence and Lorsch (1967), Thompson (1967) and Perrow (1970) were all concerned with the issue of how to achieve activity integration. With regard to what would affect the mode of integration that should be adopted, using slightly dissimilar terminologies, they all recognized contingencies such as complexity, uncertainty, and interdependencies. Building on these classical sources, Grant (1996) took integration-thinking a step forward by adding the ‘knowledge’ issue to the previous focus on activity integration. In his view, due to the difficulties and high cost of transferring and sharing tacit knowledge, such transfer should be avoided where possible. Hence, knowledge integration should be concerned with the use of appropriate mechanisms for managing knowledge complementarities to be chosen in a cost-economizing manner (Grant, 1996; Postrel, 2002; Enberg, 2007; Schmickl and Kieser, 2008). For example, while contingencies such as minimal interdependencies and limited complexity would allow for the use of cheap mechanisms (such as rules, roles and routines), severe “team interdependencies” would call for expensive and communication-intensive mechanisms, such as “group problem solving and decision making” (Grant, 1996, p. 114). As our interest lies in high complexity contexts, his advice would thus be that we have to rely on the communication-intensive and expensive team option. A similar reliance on the team recipe is suggested by Grandori (2001). Like Grant she acknowledges the tacit-explicit knowledge distinction and looks upon the choice of (knowledge) integration mechanisms as an economizing endeavour. However, unlike Grant, who takes type of interdependences as a main contingency, she focuses on knowledge-related contingencies, such as the degree of knowledge differentiation and knowledge complexity. The first dimension, differentiation, refers to differences in technical specialities and cognitive orientations among individuals and includes differences in languages, the perception of relevant information, the theories and practices used and the types of results pursued. As differentiation increases, “the possibility of mutual understanding, of succeeding in decoding the messages, of utilizing the knowledge of others, decreases” (Grandori, 2001, p. 390). The second dimension comprises two components; namely computational complexity and epistemic complexity. Computational complexity refers to the number of elements and symbols involved and the possible connection between them. The second component, epistemic complexity, adds the aspect of 757 knowledge tacitness. Thus, when knowledge differentiation is high and is combined with high computational/epistemic complexity — teams are appropriate, says Grandori (2001, p. 394). However, neither Grant nor Grandori provide much detail as to the nature or characteristics of teamwork that may actually bring about a functioning process of knowledge integration in such kinds of difficult development contexts. Typically, in the literature on teams and project teams, the need for close interaction between all members is emphasized. Katzenbach and Smith (1993, p. 9), for example, state that “real teams are deeply committed to their purpose, goals, and approach. High-performance team members are also very committed to one another. Both understand that the wisdom of teams comes with a focus on collective work-products, personal growth…”. Moreover, Nonaka and Takeuchi (1995, p. 242) argue that “the product development project emerges from the constant interaction of a multidisciplinary team whose members work together from start to finish”. These accounts of teamwork promote the image of an interaction and communication context that underlines its compositional homogeneity and the equality of its members with regard to their authority and ability to contribute. Yet, while such features may be both common and benefit practice, they are expressed quite generally, making it difficult to use to them in order to discriminate between teamwork practices which mirror those characteristics more or less closely. Furthermore, as shown by the extensive overview provided Hoegl and Gemuenden (2001), while the general importance of teams to the success of innovative processes is well documented in the literature, prevailing studies have a tendency to demonstrate the link between the mere existence of a teambased organization and innovative performance, and pay little attention to specifying in more detail what facets of teamwork may have contributed. As a remedy, they propose a Teamwork Quality (TWQ) construct comprising 6 facets: communication, coordination, balance of member contributions, mutual support, effort, and cohesion. Based on their survey study, they also report that TWQ was positively related to project success (including team performance and team members' personal success). In our view, this is a very helpful construct, which is useful in characterising team operations in the Turbine case. In the following analysis, we will use this construct to indicate to what extent a team should be considered ‘integrated’ or ‘segregated’. As defined in most dictionaries, segregation refers to the, more or less voluntary, physical separation of groups and individuals (NE, 2010). This concept is often used in analyses of segregated housing areas, racial segregation, etc. In such contexts, it often carries an additional connotation of lack of interaction and non-equal opportunities among various groups. In the Turbine project, all the team members were colocated, and our choice to use the TWQ construct to account for segregation facets in the Turbine project is thus a means to recognize these additional interaction/equality aspects. Focusing on its quality of interaction and equality features, we below present the six facets that make up the TWQ construct in a shortened version (Table 1). As will be detailed in the case section below, the Turbine project team relied very much on a process of iteration between Author's personal copy 758 C. Enberg et al. / International Journal of Project Management 28 (2010) 756–765 Table 1 Facets of Teamwork Quality (Hoegl and Gemuenden 2001, pp. 437–438). Communication The possibility for all members to communicate formally and informally with all other members Coordination The synchronization of individual contributions and the need to agree on work break-down structures, schedules, budgets, and deliverables Balance of member The possibility of all members to bring in their views and contributions ideas, unrestrained by hierarchy and dominating others Mutual support The existence of cooperative frames of mind and mutual respect Effort Everyone knowing and accepting the work norms concerning sufficient effort Cohesion Team members' sense of togetherness, belonging, and desire to remain on the team a small core group of experienced team members and other team members with less experience but with a deep knowledge of their specialist areas. Here, the core group had a major integrative role, whereas other members were mainly responsible for the specific tasks assigned to them. Quite a few of the latter felt somewhat uneasy about their role and expressed that they felt ‘decoupled’ from the project. It seems that this division of labour between core and non-core members was an important means of managing a continuous process of trial-and-error, comprising both an intentional and unintentional generation of deviations, which nurtured the process of sense-making predominantly among those more experienced. Yet, although the team appeared to display many signs of segregation, it was a team where all the members were needed and important, and despite the complexities and tensions that accompanied the project, it was brought to a successful conclusion and those involved opined their satisfaction, in particular with the technical solutions finally elaborated. 3. Data collection The study of the Turbine project was conducted at PowerCo,1 a global developer and manufacturer of power generation equipment. The study was inspired by ethnography (Alvesson and Deetz, 2000) and one of the authors was co-located with the project team during the period from July 2002 to July 2003 — i.e. during the concept and detailed development phases of project work. During this period, the project members were intensively engaged in trying to solve the technical problems and set the design layout of the new steam turbine. During the first months of fieldwork, our focus was on the efforts at building an understanding of the context of project work and how product development projects were conducted at PowerCo. The intranet and shared databases were browsed for policies and manuals related to product development and project work. Informal conversations with the two project managers (one on them whose role was more that of a general manager and the other one who had the responsibility for the flow path design) took place on three occasions. Furthermore, secondary material, like presentations, 1 PowerCo is a fictive name. results and notes related to the Turbine project, were studied to gain an understanding of its particulars. In October 2002, we started to attend the weekly project meetings (called aero/MI meetings2). After having been present as an observer at six project meetings, one general progress meeting and one LP-review meeting,3 a first round of interviews was starting by the end of November. Nine semi-structured interviews were conducted; seven interviews with the project members who were most intensively involved in the project at the time (i.e. those who were dealing with the flow path design from an MI or aero perspective) and one with each of the two project managers. Our questions centred on issues of coordination and cooperation and each interview lasted for approximately 2 h. These interviews largely formed the basis of our understanding about the roles and tasks of different project members and the prevailing pattern of interaction between experienced and less experienced project members. After (and partly in parallel with) this first round of interviews, followed a lengthy, intense period of work, extending throughout Spring, when often, several meetings were held each week. Meetings had to be attended somewhat selectively in this period as beginning in February, a second round of interviews, comprising six additional project members were also conducted. Both the project meetings and the last round of interviews helped to further clarify the pattern of interaction between the different groups of project members that had previously been discerned. Project work progressed at a reasonable pace and the last project meeting we attended took place June 24, 2003, as the Turbine project had then passed the concept phase and was halfway through the detailed development phase. By that time a total of 32 meetings had been attended; 27 weekly aero/MI meetings, two LP-review meetings, one concept review meeting, one general progress meeting and one STECO-meeting.4 In May– June 2003 we wrote a lengthy report about the Turbine project, which was distributed to the project managers and the project members. This report then constituted the basis for a workshop at PowerCo in August 2003. Putting all of this together, we believe that we had a good opportunity to gain the kind of in-depth understanding that is a qualitative mark of any single case study (Dyer and Wilkins, 1991) (Table 2). 4. The Turbine project PowerCo develops and manufactures power generation equipment. At the time of the study, its sales approximated €21 billion and the company had about 100,000 employees in more than 70 countries. In about year 2000, there was a significant 2 Aero/MI meetings involved project members representing aerodynamics and mechanical integrity, the project managers and sometimes project members representing other disciplines. Such meetings were held every week and focused on project progress and the agenda for the next week. 3 A general progress meeting involved members and managers of various projects from different locations and several projects, including the Turbine project, were discussed. LP-review meetings (LP refers generally to ‘low pressure’ turbines) served to discuss technical issues related to the Turbine project with experts from different parts of the organization. 4 STECO was the steering committee of the Turbine project. Author's personal copy C. Enberg et al. / International Journal of Project Management 28 (2010) 756–765 759 Table 2 Overview of data collection. Period of time Number and type of meetings July 2002 August 2002 September 2002 October 2002 November 2002 December 2002 January 2003 3 informal meetings with project managers 1 general progress meeting February 2003 March 2003 April 2003 May 2003 June 2003 August 2003 4 aero/MI meetings, 1 LP-review meeting 4 aero/MI meetings 1 aero/MI meeting 2 aero/MI meetings, 1 concept review meeting, 1 LP-review meeting 3 aero/MI meetings, 1 STECO-meeting 1 aero/MI meeting 5 aero/MI meetings 3 aero/MI meetings 4 aero/MI meetings downturn in the market for power generation equipment which resulted in decreased demand for steam turbines. The North American market, which a few years earlier had constituted the most important one in terms of sales, was shrinking drastically. Instead, Europe and Asia were increasingly seen as promising future markets. However, while the emphasis among North American and European customers was on high efficiency and serviceability, Asian customers stressed low cost. This meant that it became unclear what features of steam turbines PowerCO should focus on. This was a difficult situation, as the company was losing market shares on a shrinking market and as they were also lagging technologically behind their competitors. Generally, our steam turbines have no obvious technological advantages and when it comes to efficiency our steam turbines are behind those of our competitors. (Managing Director, R&D department). To deal with this situation PowerCo thus realized they had to develop a more competitive steam turbine. Yet, as suggested above, it was unclear whether customer demand implied that they should focus on offering a turbine at a low initial cost or a technologically advanced turbine with substantially improved efficiency. In the light of this background, PowerCo decided to initiate the Turbine project. 4.1. Project goals and stages The project's target was stated in terms of developing a 50/ 60 Hz Low Pressure (LP) steam turbine with increased efficiency, an increased exhaust area and a blading cost increase of less than 15%. In a sense, this goal mirrored the market uncertainty regarding the trade-off between low cost and efficiency. However, while it expressed the need to increase efficiency, it did not present any indication of how much it should be improved. Yet, it was recognized from the outset that the project had to come up with a significantly ‘better’ turbine, a task that seemed quite risky in particular considering PowerCo's limited knowledge of issues that concerned extensive vibration excitation at high volume flows and the development of reliable tools for aerodynamic calculations. In addition, there Number of interviews Other activities Collecting secondary material. Attending information sessions. Observation of every-day work. Informal conversations. 7 interviews 1 interview 1 interview 6 interviews Writing of company report Workshop were quite a few challenges of a more general character, such as the shortage of personnel, the tight deadline and the tight schedule allowed for the ex works. 4.2. A multi-disciplinary challenge The Turbine project involved highly specialised engineers from different knowledge areas, most of whom belonged to the basic disciplines of mechanical integrity (MI) and aerodynamics (aero) or to related sub-disciplines such as blade design, design layout and testing. Most members of the Turbine project had a similar educational background, as they had a university degree in mechanical engineering or turbo-machinery. However, most of them were strongly specialised within their basic discipline. Moreover, even within each discipline, project members were specialised — in the sense that there were differences in experience and competence. The development of a functioning, reliable and efficient steam turbine involves elements of aerodynamics, design and mechanical integrity. Although at the general level, it is fairly easy to understand the workings of a steam turbine, it involves numerous technical interdependencies that impact on each other in complex ways and often, project members have to handle contradictory demands. For instance, while rotor blades at the inlet area are always equipped with shrouds to reduce losses, rotor blades at the exhaust area are several times longer, less compact and object to higher frequencies. Thus, while putting a shroud at the end of such a blade will most certainly contribute to better efficiency it is less suitable from a mechanical integrity (MI) perspective as it results in higher material stress, which impacts on the durability of the turbine. As suggested by this brief example, there is an inherent trade-off between efficiency (a main concern of the aero-group) and mechanical integrity (a main concern of MI-group). It was even suggested that aero and MI “live in different worlds”. It's difficult […] because efficiency is mainly done by the aero group and we MI guys are just a bit of disturbance for them, so most things that we propose are, how can you say, counterproductive […] because MI in the end gives you some limitations to the aerodynamic freedom and that is Author's personal copy 760 C. Enberg et al. / International Journal of Project Management 28 (2010) 756–765 always an argument between aero and MI […] We [MI] have to go to the limit. (Experienced project member, MI) Furthermore, the changes in the market had an impact on how the trade-off between efficiency and costs should be considered and handled. Market changes in the direction of higher demand for efficiency implied that MI people were put under pressure to “swallow some frogs”, i.e. to go a bit beyond what they consider advisable from the perspective of material integrity and reliability. 4.3. Work at the edge of technology The Turbine project thus involved a couple of trade-offs. First, a market-induced one which was related to efficiency and low cost, and second, a technological one regarding aspects of aerodynamics and MI. In addition, the project required development work which called for ventures into unchartered territory. For example, the project members had to develop the longest rotor blades ever applied in the history of the company. This was a task with extraordinary requirements particularly with respect to issues of mechanical integrity and it led them to investigate the possible use of titanium instead of steel, a material of which they had no former experience, in the rotor blades. Since the project members had to generate new knowledge as part of the endeavour, there were seldom any norms or standards against which they were able to compare and evaluate the results of the different design options. Furthermore, there were no possibilities to actually test the different design options in real life as it would have been too costly and time-consuming to build a prototype for test running. Hence, project members had to rely to a great extent on results from computer modelling and calculations; and they also had to live with uncertainty. The Turbine project was described by the project members, not only as “extremely complex” but they also frequently stated that the task assigned to them involved “working at the edge of technology” or “at the edge of what is possible”. [In this project] you are really at the leading edge of technology and knowledge […] The nature of work is such that you try to push the frontiers of what is possible. (Experienced project member, aero) The Turbine project also involved the development of much of the software that was needed for modelling and calculating the different design options. It was often noticed that this software had its own inaccuracies, and occasionally contributed to increased, rather than decreased, uncertainty. [The software] has its own shortcomings and tolerances… and inaccuracies… so that in the end you will get results that you should always be a bit sceptical to […] we are also working on alternative tracks in order to make sure that if we find serious shortcomings and risks, that we have alternative tracks. […] it's very complex to model all these things and then the programs have their own inaccuracies and so on. (Experienced project member, aero) Actually, it seems that the project members more often than not encountered difficulties when trying to make sense of computational results. On some occasions, the difficulties were related to problems of understanding whether the results obtained were desirable or not, e.g. this was often the case when analysing models and calculations that concerned erosion. Only a few years previous to the study, erosion could not be modelled and had therefore not been an issue when developing a new steam turbine. Project members were not sure whether the problem was due to the software tools used for modelling and calculation, or whether it was actually due to the technology. However, even when the software used was considered reliable, and people thought they “knew the theory behind”, model-based experimentation often produced unexpected results. Such experiments thus provided more information, but increased rather than decreased uncertainty. However, such results were often welcomed as they supplied input for thought and evidenced the need for more in-depth inquiry. Model-based experimentation involving many variables and intricate interdependencies may thus bring about ‘interesting’ deviations, due to the fact that people, erroneously, believe that they know the underlying theory. Experimentation, however, was also used in situations where it was not possible to estimate any clear expectations from theory. Such experiments could sometimes generate outcomes which were far beyond the imagination of even the most knowledgeable project members. Even if people were well aware of the separate impact of each variable on other variables, the system-wide implications of their compound interactions soon became incomprehensible as the number of variables increased. Hence, in modelling various features of the trade-off between issues concerning aero and MI, the variable set-up could result in a multiplying interactivity. 4.4. A team of non-equals In the Turbine project, members with different expertise had different roles and responsibilities. When observing the way in which experienced and less experienced project members respectively participated in project meetings, we noted that not only their knowledge domains mattered, but also their experience. Experienced and less experienced project members were responsible for widely different types of tasks. Less experienced project members were responsible for the execution of well-defined work packages, the results of which were handed over to the experienced project members for analysis. Although less experienced project members were invited to present their work, they were seldom involved in discussing and analysing the results of this work. …that is clear that they [the less experienced project members] also want to do some work on the whole without being controlled, or told what they should do. That means that we have to look for work packages that they can do on their own, like calculations or studies, smaller studies and investigations […] where it is clear, not too many changes in the design, three, four areas and calculate this. (Experienced project member, aero) A few knowledge sharing efforts took place between experienced and less experienced project members but the knowledge shared was limited and unequally distributed. Author's personal copy C. Enberg et al. / International Journal of Project Management 28 (2010) 756–765 Because I think that they exchange a lot of useful information … when they are talking about numbers but I don't understand what it is about, the pressure clearance or whatever. (Less experienced project member, MI) Similarly, when asking the project members how they made sense of computational results, those with much experience suggested that they relied on ‘a gut feeling’, whereas less experienced members replied that they lacked any kind of guidance for evaluating the results of their computations. Unsurprisingly, the less experienced project members considered themselves to be ‘decoupled’ from the project although they were formally full members. The image of ‘decoupledness’ provided by the less experienced project members was also acknowledged by the experienced. 4.5. Experienced members' interaction While the work of less experienced project members was predominantly of an individual character, the work of the experienced members was more interactive. Lukas, one of the more experienced project members, stressed the importance of the fact that he and his fellow team members were co-located and could “discuss together […] until we come up with something and come down to something, which means an alternative, what we can do”. The core team, he explained, should remain a multi-disciplinary one, a team of people with their own deep specialist knowledge, but at the same time in much of their interaction they were dependent on an underlying tacit understanding of each others' working practices. If I make a calculation, if the mean is 300, that means a lot of things to me and I know that if it's Valentin who has worked with it, then I know what he did and the process is clear. If I told him just to make [a certain kind of calculation, explained only briefly] then I know that he knows that I mean a stress calculation 2D and frequency calculation 2D for the system, and that means a process. So there are a lot of things which are unsaid. (Experienced project member, MI) At project meetings, discussions were focused on identifying and making sense of problems from the basis of computational results which were represented in for example graphs and diagrams. Drawings and sketches were also frequently used. Those project members who gave a report of their work frequently put slides and printouts on the table or took a piece of paper and a pencil to illustrate what they meant. The other project members gathered around the table and leant forward to have a closer look. The discussion about what design options were technically feasible or not centred about such illustrations and the computational results which they somehow mirrored. Based on these illustrations, which served as a kind of boundary objects, experienced project members could then decide which values needed to be changed before a design option could constitute “a promising alternative”. Yet, that did not mean that they knew how to change the design in order to reach the desired, or at least an acceptable, solution. 761 The use of illustrations was thus a way of handling the complexity of the issues and the many technical interdependencies that existed. The project members argued that the illustrations helped them make sense of what was said and done in a way which would not have been possible if they were only given verbal information. If you look at the actual things we are talking about, they may be quite complex so they need some kind of simplifications or at least representations in order to make them understandable. You can't say stiff or less stiff or…in a way that helps somebody else…so you have to show them this curve, look at this and this… (Experienced project member, MI) Basically, they somehow had to learn from computational results and step-by-step exclude design options which were not suitable. The development of the new turbine thus mirrored a kind of trial-and-error process and — an iterative process of approaching a good enough design alternative, rather than finding the best solution by means of an optimization endeavour. 4.6. At the end Eventually, the Turbine project resulted in a very successful design concept. The final design solution involved a straight root, a double coupling concept (i.e. both snubber and shroud) of the last stage blade and used a steel material instead of titanium. Proudly, one of the project managers stated that “it is the largest blade you can make with this material”. He went on to say that the discussions about cost versus efficiency and risks had been never-ending, but that the final solution arrived at was better than expected with respect to both efficiency and costs. PowerCo, he continued “can now increase the operating range of its turbines due to better mechanical integrity”. Thus, both aero and MI seemed satisfied with the final solution and even if the time and cost targets had been exceeded, the project managers were happy with the outcome of the project. When the Turbine project formally ended in December 2006, two turbines had already been sold and work on the first commission was due to start by the end of 2007. Luckily, the market situation had then also changed for the better. At the moment, we are struggling to keep our commitments. We are too successful. We have too much to do. We have a very good order intake and I'm afraid that we will end up with penalties as it will be difficult to keep our promises due to the order intake. (Project manager 1) 5. Concluding discussion The Turbine project proved to be a challenging undertaking. New turbines are not introduced on a frequent (or regular) basis; e.g. on the roadmap for the coming five years, the Turbine project, with an estimated duration of three years, was the only development project planned. As a result, the experiences and lessons learned from previous development projects were both Author's personal copy 762 C. Enberg et al. / International Journal of Project Management 28 (2010) 756–765 unequally distributed among project members and likely to become a bit obsolete, if not simply forgotten. Moreover, the project involved working at the edge of technology, which implied creative work to come up with an innovative technology which, to use the terminology of Rosenbloom and Christensen (1994), was not only new to the firm but also to some extent ‘new to the world’. This implied working with materials, which Power Co as well as their competitors had limited experience of, and relying on newly developed methods and tools which were un-sufficiently validated. In the case story, we provide a significant amount of detail about the preconditions that prevailed as well as the way in which knowledge integration was enabled. Below, we turn to the opposite approach and try to account for our observations using just a few conceptions. First, we try to substantiate our claim that this project was associated with profound uncertainty/complexity. Second, we focus on the issue of identifying the prevailing pattern of team-based knowledge integration. Finally, we propose the notion of a Segregated Team and discuss its applicability in contexts such as the one represented by the Turbine project. 5.1. A highly complex project We use the notion of complexity in an over-arching sense as comprising the “overall difficultness and messiness of the overall project” (Williams, 1999, p. 271). In our interpretation then, the Turbine project involves both ‘structural’ and ‘dynamic’ complexities. In the case of the former, the traditional view is to recognize the number of elements and the degree of interdependencies between elements (Simon, 1962; Baccarini, 1996). But, following Grandori (2001) we would also include how varied or ‘differentiated’ these elements are. In our view, the Turbine project scores high on structural complexity. Considering the large number of interdependent subtasks and the high degree of simultaneity among them, the multitude of variables involved in modelling the aero and MI considerations, the presence of complex interaction effects, etc., clearly indicate that the Turbine project scores high on the firstmentioned structural aspect. With regard to ‘differentiation’, we notice that MI and aero members represent quite different fields of expertise with a different focus; to maximise reliability and efficiency respectively. Furthermore, a similar observation can be made with respect to differences between the experienced and less experienced project members. The experienced project members had a gut feeling for what information was relevant and succeeded in making sense of it while the less experienced members had difficulties sorting out the relevant cues from the information provided and understanding its implications. We therefore suggest that knowledge differentiation should be considered to be high. Yet, although the above analysis makes sense, it does not capture the full flavour of the preconditions that prevailed in this case. It was often the case that project members had knowledge about specific cause–effect relationships, but experienced a lack of understanding of the compound cause–effect relations. For example, the degree to which a small change to the shape of the blade influenced frequency families, erosion, mechanical integrity and ultimately, efficiency was largely incomprehensible. As a result, sometimes very surprising but ‘interesting’ outcomes from experiment and tests were obtained due to the impossibility of assessing the multiplying effects of small design changes in one part of the flow path channel on other parts of the system. We suggest that ‘dynamic complexity’ would be a fitting term for the presence of such unforeseeable and unimaginable multiplying effects of small changes. This notion of complexity would be well in line with the discussion in Tsoukas and Hatch (2005), who point out that complex systems tend to be non-linear and fractal, exhibit recursive symmetries, be sensitive to initial conditions and be replete with feed-back loops. In such contexts, there will be “no proportionality between causes and effects. Small causes may give rise to large effects.” (op cit, p. 238). As Weick (2001, p. 155) discusses, making sense of computations and results is difficult in such contexts, since “so little is visible and so much is transient”. In the Turbine project, the turbine existed only in the form of CAD-models during the development process and the only way to test different variants and the assumptions from which they were modelled, was through simulations. This in turn was a difficult undertaking, not least because the tools themselves constituted new technology and were subject to their own shortcomings, tolerances and inaccuracies. As mentioned in the case section, much of the experienced project members' knowledge was tacit, making them rely much on their ‘gut feelings’ and their sense of what other experienced members knew as they interacted. When ambiguity derives from unexpected and confusing information rather than from a lack of information, sense-making is needed; “to remove ignorance, more information is required. To remove confusion, a different kind of information is needed, namely, the information that is constructed in face-to-face interaction that provides multiple cues” (Weick, 1995, p. 99). 5.2. A segregated team Following Grant (1996) and Grandori (2001), group problem solving and teams would constitute appropriate mechanisms for knowledge integration in such highly complex contexts. Both writers think of this mechanism as a last resort expensive choice. Considering “the high costs of consensus decision making given the difficulties of communicating tacit knowledge” (Grant, 1996, p. 115), this kind of mechanism should be used only where the use of cheap options such as rules, routines, etc., do not suffice. Similarly, Grandori (2001) considers the choice of the expensive team option as a necessity in such difficult contexts. It seems that the preconditions that prevailed in the Turbine case, would make it sensible to rely on the team as a basic knowledge integrating mechanism. Grant and Grandori characterise this mechanism as ‘communication intensive’, they do not however penetrate it much further. Hence, in the analysis of the Turbine case we below turn to the TWQ construct by Hoegl and Gemuenden (2001), which represents one way in which the Grant/Grandori framework may be developed. Author's personal copy C. Enberg et al. / International Journal of Project Management 28 (2010) 756–765 With regard to the Turbine project as a whole, our analysis of the six TWQ facets, suggests that this team was very much a ‘segregated’ one. As evidenced in the case description, the experienced and the less experienced project members assumed responsibility for widely different types of tasks. While less experienced project members worked in comparative isolation and were granted little responsibility for activities aimed at knowledge integration, the more experienced project members worked intensively together and had key roles in bringing about knowledge integration. Less experienced members had limited opportunity to voice their ideas and views in discussion with those who were more experienced. This was because at times they did not quite understand the subject being discussed and other times as they were simply not invited. As a result, they were hardly seen or treated as ‘full’ members of the team with equal opportunities to engage in ongoing communication and interaction. Within the core group of experienced members, however, the opposite situation prevailed. Here, people were involved in frequent interaction, discussing their ideas in a respectful manner, etc. As a result, while the Turbine team as a whole appeared as a quite segregated one, in which many of the members was denied having a say in many vital issues and leaving them with a feeling of being decoupled and not full members of the project, the opposite situation prevailed among the members of the core team. The less experienced members did not really constitute a specific sub-group, and while there was certainly some interaction among them, this was rather triggered by efforts to manage their individual assignments. Despite this, all members of the Turbine project were necessary in a complementary fashion and the combined effort was achieved in an iterative manner, in which the less experienced project members were given assignments, formulated by the core team members, who were the main integrators. We would like to refer to the type of team that resulted from this division of labour and interaction, and which constituted the mechanism by which knowledge integration was enabled, as a ‘segregated team’. The six-dimensional TWQ construct by Hoegl and Gemuenden (2001) is used below to illuminate some basic features of the pattern of segregation of the Turbine team (Table 3). Thus, our findings thus suggest that the core group was not at all segregated, but instead should be seen as a fully integrated 763 group. It is here we find the experienced project members with deep individual and idiosyncratic knowledge bases which they were able to apply to make sense of computational results, to define work packages and thereby enable the integration of the specialist knowledge of their less experienced colleagues. The people in this group could communicate freely, have the possibility to influence the organization of work, introduce their ideas, be respected and feel like full and equally valuable members of the Turbine team. Their interaction took place at ad hoc or regular project meetings during which people from aero and MI discussed problematic issues from their different perspectives, using previous experience and exemplars to try out ideas or to find out “what direction to take”. Often, these efforts had little to do with finding one best way. As one project leader said, it was more about finding “a demarcated field on which to play”. Although the experienced project members worked together in an interactive and responsive manner, it was still important that each of them retained their distinctive competences; achieving a high degree of knowledge base similarity was not something they pursued beyond the minimum degree needed to enable their cross-disciplinary communication (Lindkvist, 2005). Hence, their interaction was honouring the virtues of both responsiveness and distinctiveness (Orton and Weick, 1990) in a semi-coupling manner (Lindkvist et al., 1998). 5.3. Why a segregated team? A first indication of efficiency would be that the project was completed within a reasonable time and it was generally regarded successful from both a technical and from a sales point of view. While this by no means proves that the project process could not have been even more efficient, a successful outcome makes it meaningful to try to account positively for the degree of success actually achieved. This would suggest that ‘the team’ and more precisely a segregated team would be an appropriate knowledge integration mechanism in the case of the Turbine project. Hence, while we generally agree with both Grant and Grandori on the appropriateness of the team option in a context characterised by high complexity, we also add to their ideas by pointing out that in such cases the teams may be quite segregated. Furthermore, the division of labour guided by this segregation principle, is not quite the same as the often referred Table 3 TWQ facets regarding less experienced and experienced team members. TWQ facet Less experienced (LE) members Experienced (E) members Communication Members had limited possibilities to communicate with other members of the project team Members had little influence on work organization Members could easily communicate with all project team members Members had extensive influence on work organization Members could freely express their views and ideas Members enjoyed respect for their knowledge and their contributions to problem solving Members were motivated and put in great effort Members felt integrated in the ‘core’ of the team Coordination Balance of member contributions Mutual support Effort Cohesion Members were severely restrained regarding their opportunities of bringing in views and ideas Members enjoyed respect primarily for carrying out assigned tasks Members became somewhat demotivated Members felt ‘decoupled’, did not have a sense of being ‘full’ team members Author's personal copy 764 C. Enberg et al. / International Journal of Project Management 28 (2010) 756–765 to idea of hierarchical decomposition. As stated by Kratzer et al. (2008, p. 540), “hierarchical decomposition serves the managerial purpose of having largely autonomous acting and highly specialized teams whose work proceeds parallel”. Such a view indicates a major initial effort by project leaders, followed by periods of increasingly stronger significance and influential power of the specialised sub-teams. As illustrated by the Turbine case, however, there was no decomposition of the entire project task into subtasks assigned to parallel teams, nor was it the case that the project leaders, or the core team, over time lessened their influence. On the contrary, as the project proceeded, the knowledge integration processes that emerged followed in a more and more subtle and complex trajectory, leaving those less experienced farther and farther away from those more experienced. Instead of converging in understanding and participation, the experienced and less experienced project members drifted even further apart. It was not until the final phase of detail specification, when all basic problems had been solved, that the core team relaxed their impact and control. Generally, there appears to be some rationality behind using a segregated team instead of a more integrated and homogenous one in situations characterised by a high degree of complexity. Apparently, most engineering firms would rely on some kind of a competence based hierarchy among their employees, and as often claimed, sometimes the fate of development projects depends on a very few highly experienced and skilled individuals, who are crucial. As a result, competence will not be uniformly distributed and communicating across different knowledge bases and competence levels will tend to be a costly affair. Excluding those who are inexperienced from discussions should thus save time and money. Expensive communicative effort will then only be granted for the interaction between experienced members, whereas those who are not so experienced may work alone with more focus on their assigned tasks (Enberg et al., 2006). Yet, the roles of the core integrators and the technical specialists were clearly complementary and vital to the operation of the project team. Interestingly, although the Turbine project as a whole scored extremely low on the TWQ indicators (Hoegl and Gemuenden, 2001) it was successfully concluded. Such a result gains support in a multiple case study of construction projects (Baiden et al., 2006, p. 22), showing “that fully integrated teams are not necessary for effective team operation within the industry”. Furthermore, a questionnaire study of software development teams, using the TWQ construct, reports that teamwork quality actually has a negative moderating effect on the relation between creative-thinking skills and team efficiency (Hoegl and Parboteeah, 2007). Hence, using a highly ‘integrated’ instead of a ‘segregated’ team may not always be beneficial. In particular, our own study of the Turbine project here suggests that a segregated team might be an appropriate design solution in a highly complex knowledge integration context. Finally, it should be pointed out that using a strongly segregated team may be a viable solution in the short run, and projects of course by their very nature tend to be just that, i.e. subject to tight limits of both time and budget. However, segregation may come at a price. As described above those who are not really let in and involved will tend to become somewhat demotivated and less committed to achieving the global project goal. Apparently, this may mean that novel and potentially valuable ideas from these ‘outsiders’, are being ignored, and that it becomes less easy to counteract the development of too strong ‘group think’ (Janis and Mann, 1977) tendencies of the core group. In our case little was done to counteract such consequences. Another thing that tends to be neglected in such a segregated team context, is the long-term development of personnel. Here, much more of master–apprentice-like learning could have been carried out as the Turbine project went on. However, although the firm did recognize such a need, it was seen as something that should not be dealt with within the context of the ongoing project. References Alvesson, M., Deetz, S., 2000. Doing Critical Management Research. SAGE Publications, London. Baccarini, D., 1996. The concept of project complexity — a review. International Journal of Project Management 14 (4), 201–204. Baiden, B.K., Price, A.D.F., Dainty, A.R.J., 2006. The extent of team integration within construction projects. International Journal of Project Management 24, 13–23. Carlile, P.R., Rebentisch, E.S., 2003. Into the black box: the knowledge transformation cycle. Management Science 49 (9), 1180–1195. Dyer, W.G., Wilkins, A.L., 1991. Better stories, not better constructs, to generate better theory: a rejoinder to Eisenhardt. Academy of Management Review 16 (3), 613–619. Enberg, C., 2007. Knowledge Integration in product development projects. Ph.D. dissertation, Department of Management and Engineering, Linköping University, Sweden. Enberg, C., Lindkvist, L., Tell, F., 2006. Exploring the dynamics of knowledge integration: acting and interacting in project teams. Management Learning 37 (2), 143–165. Grandori, A., 2001. Neither hierarchy nor identity: knowledge-governance mechanisms and the theory of the firm. Journal of Management and Governance 5 (3–4), 381–399. Grant, R.M., 1996. Toward a knowledge-based theory of the firm. Strategic Management Journal 17, 109–122 (Winter Special Issue). Hoegl, M., Gemuenden, H.G., 2001. Teamwork quality and the success of innovative projects: a theoretical concept and empirical evidence. Organization Science 12 (4), 435–449. Hoegl, M., Parboteeah, K.P., 2007. Creativity in innovative projects: how teamwork matters. Journal of Engineering and Technology Management 24, 143–166. Hoopes, D.G., 2001. Why are there glitches in product development? R&D Management 31 (4), 381–389. Janis, I.L., Mann, L., 1977. Decision Making: a Psychological Analysis of Conflict, Choice and Commitment. Free Press, New York. Katzenbach, J.R., Smith, D.K., 1993. The Wisdom of Teams — Creating the High-performance Organization. McGraw-Hill, London. Kratzer, J., Gemuenden, H.G., Lettl, C., 2008. Balancing creativity and time efficiency in multi-team R&D projects: the alignment of formal and informal networks. R&D Management 38 (5), 538–549. Lawrence, P.R., Lorsch, J.W., 1967. Organization and Environment: Managing Differentiation and Integration. Division of Research Graduate School of Business Administration, Harvard University, Boston. Lindkvist, L., 2005. Knowledge communities and knowledge collectivities: a typology of knowledge work in groups. Journal of Management Studies 42 (6), 1189–1210. Lindkvist, L., Söderlund, J., Tell, F., 1998. Managing product development projects: on the significance of fountains and deadlines. Organization Studies 19 (6), 931–951. Author's personal copy C. Enberg et al. / International Journal of Project Management 28 (2010) 756–765 Nationalencyplopedin (NE), 2010. http://www.ne.se/lang/segregration. Nonaka, I., Takeuchi, H., 1995. The Knowledge-creating Company — How Japanese Companies Create the Dynamics of Innovation. Oxford University Press, New York. Orton, J.D., Weick, K.E., 1990. Loosely coupled systems: a re-conceptualization. Academy of Management Review 15, 203–223. Perrow, C., 1970. Organizational Analysis: a Sociological Review. Tavistock Publications, London. Postrel, S., 2002. Islands of shared knowledge: specialization and mutual understanding in problem-solving teams. Organization Science 13 (3), 303–320. Rosenbloom, R.S., Christensen, C.M., 1994. Technological discontinuities, organizational capabilities, and strategic commitments. Industrial and Corporate Change 3 (3), 655–685. Sauser, B.J., Reilly, R.R., Shenhar, A.J., 2009. Why projects fail? How contingency theory can provide new insights — a comparative analysis of NASA's Mars Climate Orbiter loss. International Journal of Project Management 27 (7), 665–679. Schmickl, C., Kieser, A., 2008. How much do specialists have to learn from each other when they jointly develop radical product innovations? Research Policy 37 (3), 473–491. 765 Shenhar, A.J., 2001. One size does not fit all projects: Exploring classical contingency domains. Management Science 47 (3), 394–414. Simon, H., 1962. The architecture of complexity. Proceedings of the American Philosophical Society 106 (6), 467–482. Thompson, J.D., 1967. Organizations in Action. McGraw-Hill Book Company, New York. Tsoukas, H., Hatch, M.J., 2005. Complex thinking, complex practice. The case for a narrative approach to organizational complexity. In: Tsoukas, H. (Ed.), Complex Knowledge — Studies in Organizational Epistemology. Oxford University Press, Oxford. Weick, K.E., 1995. Sensemaking in Organizations. SAGE Publications, London. Weick, K.E., 2001. Technology as Equivoque: sensemaking in new technologies. In: Weick, K.E. (Ed.), Making Sense of the Organization. Blackwell Publishers, Oxford. Williams, T.M., 1999. The need for new paradigms for complex projects. International Journal of Project Management 17 (5), 269–273.