Presented in the Thirteenth International Working Seminar on Production Economics, Igls/Innsbruck, February 16-20, 2004

A Solution for the Material Delivery Problems in Construction Projects

Timo Ala -Risku*, timo.ala -risku@hut.fi Mikko Krkkinen*, mikko.karkkainen@hut.fi

*Helsinki University of Technology TAI Research Centre P.O. Box 5500 FIN-02015 TKK FINLAND

Abstract: The new project management methods emerging for construction projects generate new kinds of challenges for the delivery process of the project materials. The basic philosophy of the emerging methods is to create short-term schedules for project tasks based on a constraint analysis of project resources. This approach places two requirements for the material deliveries: Transparency to material availability, and short response times in the supply chain. In this paper, we propose a potential solution for managing the material logistics of construction projects. The solution consists of a shipment tracking-based approach tested in pilot implementations to provide inventory transparency, and a pro-active delivery approach for efficient material deliveries validated in expert group discussions. Keywords : supply chain management, logistics, inventory transparency, construction

Introduction
New project management methods addressing the shortcomings of traditional methods by adding flexibility to the execution of construction projects create new challenges to material delivery processes (e.g. Ballard, 2000; Choo et al., 1999; Chua et al., 2001; Koskela and Howell, 2002). The new methods acknowledge the impossibility of creating an exact schedule beforehand for a large, complex project. Instead, the methods use continuous planning on single construction task level. The basic philosophy of the methods is to create short-term schedules for project tasks based on a constraint analysis of project resources. The approach places two requirements for the material deliveries: The analysis of material constraints requires transparency to material availability in site inventories and other parts of the supply chain, and the short time-span of planning demands for short response times in the supply chain. The aim of this paper is to present a potential solution for managing the material logistics of construction projects. The solution consists of a shipment tracking-based approach to provide inventory transparency, and a pro-active material delivery approach for timely material availability. In the first section of the article we will review literature related to the new project management methods and its implications for material replenishments of construction projects. In the second section, we present our research problem and discuss our research design. The proposed solution for material replenishments of construction projects is presented in the third section. And, in the final sections we make concluding remarks on the proposals and outline directions for further research.

Literature review
In this section we review literature dealing with shortcomings of traditional project management practices, proposed developments, and their implications for logistics. First, we make a short overview on the criticism calling for developments in the traditional project management approach, and present a framework developed by Ballard (2000) the Last Planner System that introduces a novel approach in line with the suggested

development directions. Then, we discuss the material delivery related capabilities that are needed for the Last Planner approach to be effectively utilised.

Challenges in prevalent project management practices

The prevalent project management methods have recently been stated as inadequate for controlling the progress of construction projects (e.g. Ballard, 2000; Choo et al., 1999; Chua et al., 2001; Koskela and Howell, 2002). Some of the strongest opinions have been presented by Koskela and Howell in their statements that the underlying theory of project management is obsolete and there in fact exists no explicitly stated theory on project management (Koskela and Howell, 2001; Koskela and Howell, 2002). Their main objection is that planning, execution, and control are not utilised in practice as suggested by PMBOK Guide of Project Management Institute. In similar vein, Johnston and Brennan (1996) have argued that the general interpretation of project management is mana gement-as-planning based on a strong causal connection between generating a plan and the resulting operational activity. They conclude that such a straightforward coupling is not feasible, since no plan can ever be detailed enough to enable mere execution with no feedback from the environment. Koskela and Howell (2002) note that reliance on a general plan leads to poor short-term planning. These observations are confirmed, for example, in a study of six construction projects in England and Brazil (Santos et al., 2002). Koskela and Howell (2002) recognise that maintaining a comprehensive up-to-date plan is difficult. Therefore, being based on an out-of-date plan, the tasks pushed to execution often cannot be performed, as they lack either predecessor tasks or other inputs. As Johnston and Brennan (1996) state, That this approach works at all is largely attributable to tacit knowledge and improvisation at the operational level. From the supply chain point of view however, this management approach including last-minute improvisation leads to inefficient practices to guard against material shortages. Materials are often ordered either very late leading to buffering at supplier to guarantee service level, or too early leading to buffering at the site (Vrijhoef and Koskela, 2000).

Therefore, there has been a clear need for a more interactive management method, where the next steps to take in a construction project are determined from the current status of the project, not from predefined and outdated schedules obsolete for controlling practical actions. More flexible project management practices have been developed to overcome the challenges of traditional project management approaches (e.g. Ballard, 2000; Chua et al; 2001). In the following, we give the Last Planner System of Ballard (2000) as an example of such approaches. Last Planner - a novel method for managing projects The Last Planner System has been developed by Ballard (2000) since 1992. It is a project execution system that uses the overall project plan as the general frame, but suggests that the day-to-day activities of the actual production should be managed by a more flexible approach accounting for the actual progress of the project. The main philosophy is to ensure that all the prerequisites needed for performing a distinct task are in place, before it is assigned to a work group. This is the task of the last planner in the chain of plans of increasing detail in a project. Therefore, Last Planner has been used to mean both the system and the person responsible for the final arrangement of tasks.
Project Objective

Information

PLANNING THE WORK

SHOULD

CAN

LAST PLANNER PLANNING PROCESS

WILL

Resources

Production

DID

Figure 1. The Last Planner System (Ballard, 2000)

The four main categories for any executable assignment are SHOULD, CAN, WILL, and DID (Figure 1). 4

SHOULD: assignments that need to be done in the near future according to the overall project plan

CAN: assignments that have all their prerequisites ready: e.g. previous project steps are completed, necessary materials are at hand, and work force is available

WILL: the tasks that are commenced before the next planning round DID: the assignments that are completed

The Last Planner System includes rules and procedures, and a set of tools that facilitate the implementation of those procedures. The two main procedures are Production Unit Control and Work Flow Control. Production Unit Control coordinates the execution of work within production units (e.g. construction crews) and includes processes for directing workers through continuous learning and corrective actions. Work Flow Control coordinates the flow of design, supply, and installation to proactively arrange work to flow in the best achievable sequence and rate. From the supply chain perspective, the most important feature of the Last Planner System is the way the work flow control utilises a lookahead process. The traditional use for a lookahead process is only to highlight what SHOULD be done in the near term. However, within the Last Planner System, the process serves several functions, e.g. work flow sequence shaping, decomposing master schedule activities into assignments and revising higher level schedules as needed, and maintaining a backlog of work ready to be performed. These are accomplished through specific processes, such as matching load and capacity and constraints analysis. The vehicle for the entire lookahead process is a schedule of potential assignments for the next 3 to 12 weeks (Figure 2), the extent depending on project characteristics, such as lead times for acquiring materials or labour force.

Master schedule activities entering 6th week 1 Assignments Screen assignments & make ready each week enough work to maintain 2 week workable backlog 2 3 4 5 6 Explode scheduled activities into work packages on entry to the lookahead window

Workable Backlog

Reasons why planned work not completed

Notify coordinator of constraints status

Figure 2. A schematic illustration of the lookahead process, showing work flowing through time from right to left (Ballard, 2000).

Constraints analysis is used on the potential assignments to determine what must be done to make it ready for execution. If the planner thinks that constraints of a specific assignment (e.g. materials, prerequisite work, or equipment) cannot be removed in time, it is shifted to a later date. The assignments enter the lookahead window at the far end, and move forward as time progresses. As all of their constraints are removed, they enter the workable backlog. The objective is to constantly have a backlog of work that CAN be performed, and based on the backlog the actual workplan (WILL) for the next period is created. The aim of the constraints analysis of Last Planner is to ensure that all resources are available for a given project task at the time of the execution. Therefore, from a material replenishments perspective information about the availability of the relevant material is crucial for the constraints analysis to function. Material flow management related challenges of the last planner -method Two distinct material flow management related challenges can be associated with using the last planner methodology: The Last Planner needs to have access to as comprehensive 6

as possible materials availability information for individual project tasks, and the material delivery lead-times need to be as short as possible to enable timely response to the shortterm planning of the approach. The two challenges give pressures for increasing the site inventory, as the materials on site are readily available for possible changes in production plan. However, excessive site inventories create problems for materials handling at site (Alves and Tommelein, 2003) and the materials have a big risk of getting lost or stolen (Krkkinen et al., 2003). And most importantly, the inventories at sites are often not registered in information systems, but need to be visually controlled to make sure they are available (Halmepuro and Nysten, 2003; ISI Industry Software, 2003). This is a serious problem for the constraint analysis of the Last Planner approach. To avoid increasing site inventory levels, the closely situated suppliers can requested finished goods inventory for the project. A more scalable solution is to establish temporary intermediate storages near the actual site to have materials available at short notice. This approach is actually directly analogous to the supplier hubs already common in the field of discrete manufacturing. Obviously, transparency to inventories in these locations needs to provided for the Last Planner. Creating supply chain transparency for project deliveries is however challenging: the networks supplying construction projects are short -term, often set up for single projects only, and the project delivery network often involves a large number of rather small companies. The most significant challenge of generating transparency for project supply network is that the networks are mostly created for one project only, and disbanded after the project is completed (Dainty et al., 2001). This has resulted in low long-term commitment, which is needed with most information technology development projects (Voordijk, 1999). Thus, solutions seeking to increase inventory transparency in project supply chains should be quick and easy to take into use and not demand significant investment due to the short time span of the usability of the solution (Cheng et al., 2001). Supply networks of construction projects are also usually characterised by including several small or medium sized enterprises (SME). For example, almost 98% of UK construction companies are SMEs, employing 24 or fewer workers (Dainty et al. 2001). 7

The ability to make investments in IT infrastructure is rather limited in the whole for these small companies, and thus can rarely engage in heavy integration efforts (Anumba and Ruikar, 2002; Elliman and Orange, 2000).

Research problem and design

Based on the gap in the current body of literature we formulated the following research problem: How to construct an effective material delivery model for construction projects with short-term task level scheduling? Due to the specific requirements for the solution pointed out by the current literature we developed two more detailed research questions: RQ1) How to gather and convey material availability information to project task scheduling? RQ2) How to organise efficient material deliveries for project materials when operating with a flexible project schedule? The research reported in this paper is based on the Innovation Action Research paradigm, first presented by Kaplan (1998). In innovation action research the aim is to initially document major limitations in contemporary practice, identify a new concept to overcome the limitation, and to continually apply and improve the concept through publication, teaching and active intervention in companies. The general flow of innovation action research is illustrated in Figure 3.

Feedback and learning

2. Teach and speak about the innovation

3. Write articles and books Invitations

Cases 1. Observe and document innovative practice

Management and organizational phenomena Advanced implementation Intermediate Initial implementation Base case

4. Implement concept in new organizations

New practices

Create changes in practice

Figure 3. Innovation Action Research process (Kaplan, 1998)

This research is still on the first stages of the development cycle. The roots of the research are founded in summer 2001, when we worked with a group of supplier companies to construction projects. The companies had severe difficulties in building transparency to their supply chains, and were thus not able to inform their customers on the availability of the goods they were supplying. One of the major difficulties was to manage the information of what goods had already arrived at the project site, as well as locating goods at site inventory. During the project, a simple tool for tracking the goods was developed. Thereafter we have been continuously working with the problems associated with building transparency to short-term supply networks. Later in another case study, a supplier company had problems with the delivery accuracy of project materials. This caused significant difficulties in the management of the construction activities. It was even claimed that material shortages occurred much too often for successful operations to be possible. To uncover a solution, we performed an initial literature survey, focusing on logistics literature. In the survey, we did not find any solutions directly answering the material delivery challenges of the case company. Our research team thus developed a potential solution model to these challenges. The solution model was basically a project finite scheduling approach that took into account 9

the constraints proposed by material delivery problems. In a more extensive literature survey, now focusing also on project management literature, to link our proposed solution to the current body of knowledge, we identified the Last Planner project management approach. The Last Planner proved to be a more comprehensive approach than our scheduling model, as it incorporates also other scheduling constraints than material availability (e.g. prerequisite tasks, personnel availability). Also, there has been software development based on the Last Planner approach making it easier to take the system into use (e.g. Choo et al. 1999). Nevertheless, we recognised that our replenishment model initially developed to work with our own finite scheduling approach is also a complementing element for the Last Planner System, and addresses the transparency and lead-time requirements it poses to the material delivery process. In this paper, we present the delivery model and its linkages to the Last Planner System. The transparency solution has been tested and developed in two installations, but we do not yet have any experiences of the finite scheduling model in a real life setting. The initial scheduling model has been validated in expert group discussions in one supplier and one installation company operating with construction projects. Therefore, our proposed solution can be considered to be in the early stages of its innovation action research development cycle.

The proposed material delivery solution

We will present our proposal in two parts. First, we address the transparency needed for determining material constraints. And second, we demonstrate how the finite scheduling of project tasks can be used to help the supplier to provide timely deliveries. Visibility to material constraints The first challenge of the material delivery process in relation to Last Planner type project management approaches is to provide the planner reliable information on material availability for project tasks. This means that the material needs of each task have to be known. The materials needed for a task can be thought of as a bill of materials (BOM) for a task. An example BOM for a project task 29 is illustrated in Figure 4. 10

Project Task_29

2 x ITEM A ITEM XYZ 5kg ITEM C

Figure 4. An example bill of materials for project task 29.

Transparency to materials in the project supply chain is needed to enable checking whether the identified task materials are available or not. If the Last Planner can be provided with availability information, the material constraints for the project task can be determined. The situation when project tasks and materials are linked and transparency to materials is achieved is illustrated in Figure 5. When equipped with the availability information, those project tasks in which material delays are detected, can be transferred to later dates where the delays are satisfied.
All materials available at site Detected delay of some material(s) Materials in interim storage Material status not defined

Task_01 Task_13 Task_54 Task_02 Task_08 Task_28 Task_56 Task_06 Task_14 Task_16 Task_24 Task_66
Present day Day 2 Day 3 Day 4 Day 5 Day 6

Task _78 Task _29 Task _17 Task _65 Task_27 Task _69 Task _21 Task _07 Task _25
Day 7 Day 8 Day 9 Day 10 Day 11

Figure 5. Illustration of the results of materials constraint analysis.

However, as discussed earlier, providing transparency to the materials is not very straightforward in construction projects, whether the inventory is located at the site or elsewhere in the supply chain. To address this challenge, we have developed a tool for creating inventory transparency based on shipment tracking. The tool has been designed for site inventories and short-term storages, where traditional inventory transparency solutions are extremely difficult to apply, and which are the most critical inventories for the finite scheduling of the project (Krkkinen and Ala-Risku, 2003). The tool is also suitable for building transparency to inventories located in any other part of the supply 11

network (e.g. at the suppliers or sub-suppliers). The material inventory information with the tool is established by tracking the incoming shipments (materials received at the site / warehouse) and outgoing materials (materials installed / sent from the warehouse) (Krkkinen and Ala-Risku, 2003).

Shipment tracking system

Intermediate Storage 1

Tracking database Code Location ID_2364 Intermediate Storage 1 ID_6421 Intermediate Storage 2 ID_1536 Intermediate Storage 1 ID_5423 Site Inventory ID_1416 Site Inventory

Timestamp 10.11.2003 9.11.2003 9.11.2003 3.11.2003 10.11.2003

Intermediate Storage 2

Figure 6. Inventory transparency provided with a tracking tool.

When building tracking based inventory transparency, deliveries arriving at the project site or an intermediate storage have to be equipped with an identifying code (e.g. order number or task number), which can be used to link the shipment that arrived to the materials it contains. These codes are registered with a tracking system, when packages arrive or are taken out of a given storage location. The tracking software then conveys the tracking code, the location of the inventory and the time of the read to a tracking database, illustrated in Figure 6. The codes read to the tracking database need to be linked to the materials in the shipment. Using the link between materials and tasks (cf. Figure 4) enables the use of the tracking information in the constraint analysis for the finite schedule. If these functionalities are built, the following information can be interrogated from the databases: What is the location of the different goods needed for a certain project task What materials are at a given location (e.g. site inventory or intermediate storage inventory) 12

What is the location of a certain shipment

This information is sufficient for checking the availability of material for the project tasks, and it can also be used to optimise inventory allocation in the supply chain. If a further database table containing the delivery dates promised by the suppliers, the ordering date, and shipping date is also connected to the tracking database, also the following pieces of information can be generated: The on-time delivery rate of a supplier The lead-time of orders (from ordering to receiving of the goods) The lead-time of deliveries (from the dispatch to the receiving of the delivery) And, if the packages are traced in several locations the de-constructed lead-times can be used for performance analysis of the supply chain (Jahnukainen et al., 1995). Material delivery triggering The Last Planner allocates tasks on the finite schedule and removes the material constraints of each task by making sure the necessary materials are ordered before the tasks are allowed to move to the workable backlog. Since the tasks are positioned on the finite schedule, they can also be used to produce forecasts of upcoming delivery needs for the suppliers. This is the best forecast available, as the Last Planner continuously updates the most likely schedule for the project tasks in the near future based on the progression of the project. When having visibility to upcoming material needs, the suppliers are able to inform the Last Planner of potential delivery disruptions. The Last Planner can then adjust the task schedule taking the emerged material constraint in to account. Also, if any changes in the task sequence are made, the supplier gets informed of delayed (or advanced) material needs. Thus, the finite project schedules can be used for enabling the suppliers to better cope with the changes in the required delivery dates of the project material. The efficiency of operations for the Last Planner can be increased by removing the need for frequent material orders. If the upcoming material requirements are conveyed directly 13

to the suppliers (e.g. via www), the suppliers can move to a pro-active delivery of the goods needed for the project tasks (a kind of VMI for projects). The suppliers can be given the responsibility to deliver the goods needed for a certain task to the project site in time. For the pro-active delivery, we have added an extra parameter to the task schedule the project buffer time. The project buffer time is used to ensure that suppliers make the materials available in time for the tasks moving to the workable backlog. This means that suppliers are given a material needs view, where the needs are advanced the amount of buffer days (illustrated in Figure 7).

To be dispatched on the present day 2 x ITEM_A 56kg ITEM_C 1 x ITEM_XYZ 3 x ITEM_B

Present day Day 2

Expected call-offs 1 x ITEM_FZC 5 x ITEM_D 12 x ITEM_L 5 x ITEM_D 12 x ITEM_L 1 x ITEM_FZC 1 x ITEM_FZC 1 x ITEM_FZC 1 x ITEM_FZC 1 x ITEM_XYZ 1 x ITEM_KYL 1 x ITEM_FZC 1 x ITEM_KYL 5 x ITEM_D 5 x ITEM_D 9 x ITEM_M 3 x ITEM_K 9 x ITEM_M
Day 3 Day 4 Day 5 Day 6 Day 7 Day 8

Task_01 Task_13 Task_54 Task_02 Task_08 Task_28 Task_56 Task_06 Task_14 Task_16 Task_24 Task_66
Present day Day 2 Day 3 Day 4 Day 5 Day 6

Task _78 Task _29 Task _17 Task _65 Task_27 Task _69 Task _21 Task _07 Task _25
Day 7 Day 8 Day 9 Day 10 Day 11

Project determined buffer time (3 days)

Figure 7. Pro-active deliveries for project materials.

For example, in Figure 7, the project has determined that it should have the materials available for the next three days. A supplier with a delivery time of two days would be requested to deliver the materials for tasks on Day 5. The materials for upcoming tasks would be given as a forecast to ensure that the supplier is prepared to deliver those according to the schedule. To summarise, this material control system allows material call-offs to be done automatically, according to the project progress and based on the workflow logic. The entire operating model is illustrated in Figure 8.

14

SUPPLIER A Material flow tracking on project task level Rolling delivery plan SUPPLIER B
Interim Interim Interim storages Storage* Storage*

PROJECT SITE

Rolling delivery plan

Project schedule updated based on material availability and other constraints

Project construction schedule Project schedule used to trigger material deliveries

Figure 8. An illustration of the proposed material delivery system.

Conclusions
A solution consisting of two parts was proposed for answering the research problem: How to construct an effective material delivery model for construction projects with short-term task level scheduling? The solution consisted of a tracking based approach for building inventory transparency for short -term supply chains, and a pro-active material delivery model for the materials for specific project tasks. Both of these elements are needed when using the Last Planner System. The presented tracking based method for solving the difficulties of site and intermediate inventory management and transparency can be considered a clear addition to the current body of literature. Its practical relevance is emphasised in project-oriented industries, as it is challenging and often infeasible to establish traditional inventory management systems to temporary storage locations used during projects. Also, we consider the pro-active delivery model for construction task materials a novel idea. However, as we are not too familiar with project management literature and 15

terminology, we acknowledge that similar approaches may already be presented with descriptions that we have not recognised. During our research we identified the change resistance of the project management to be the most difficult issue for taking the finite scheduling model into use. The project management people were unwilling to formally state that disturbances in material deliveries are a legitimate reason for re-scheduling project tasks: It was considered next to stating that it is acceptable to miss promised delivery dates. The potential for heavy claims for the time that installation workers have been present at the construction site unnecessarily due to lack of materials, has served as a motivator for on-time deliveries (Halmepuro and Nysten, 2003). If the project is re-scheduled when material shortages occur, different methods for ensuring availability need to be developed.

Further research
We have identified two relevant areas in the need of further research efforts. A case study with a company using the finite scheduling approach is needed in order to determine the most critical points of the supply chain where transparency is required (our experiences thus far have all pressed the importance of site inventories). Also the effects of the finite scheduling model and the established demand visibility on the material delivery process need to be studied more comprehensively. If a company that has already adopted the approach cannot be accessed, we will pursue piloting the approach for example with manual plan updating methods. The already tested method for building inventory transparency can easily be used for determining the material constraints for the upcoming tasks. Another interesting area for future research is to identify the similarities and differences of material flow challenges in construction projects of various industries. For example, does the execution of previous tasks have less importance when establishing communication or power networks (e.g. when building the upcoming digital television networ k or an electric network) than building construction?

16

References
Alves, T.C.L. and Tommelein, I.D., (2003). "Buffering and Batching Practices in the HVAC Industry." Proceedings of the 11th Conference of the International Group for Lean Construction (IGLC11), 22-24 July 2003, Virginia Tech, Blacksburg, Virginia. Anumba, C.J., Ruikar, K., (2001), "Electronic commerce in construction - trends and prospects", Automation in Construction, Vol. 11, pp. 265-275 Ballard, G., (2000) The Last Planner System of Production Control. A thesis submitted to the Faculty of Engineering of The University of Birmingham for the degree of Doctor of Philosophy. School of Civil Engineering, Faculty of Engineering, The University of Birmingham. Cheng, E.W.L., Li, H., Love, P.E.D., Irani Z., (2001), "An e-business model to support supply chain activities in construction", Logistics Information Management, Vol, 14, No. 1/2, pp. 68-77 Choo, H.J., Tommelein, I.D., Ballard, G., and Zabelle, T.R., (1999). "WorkPlan: Constraint-based Database for Work Package Scheduling." ASCE, Journal of Construction Engineering and Management, Vol. 125, No. 3, pp. 151-160. Chua, D.K.H., and Shen, L.J., (2001), Constraint modelling and buffer management with integrated production scheduler, Proceedings of the 9th Annual Conference of the International Group for Lean Construction, 6-8 August, 2001, Singapore. Dainty, A.R.J., Briscoe, G.H., and Millett, S.J., (2001), New perspectives on construction supply chain integration, Supply Chain Management: An International Journal, Vol. 6, No. 4, pp. 163-173. Elliman, T., Orange, G., (2000), "Electronic commerce to support construction design and supply-chain management: a research note", International Journal of Physical Distribution & Logistics Management, Vol. 30, No. 3/4, pp. 345-360. Halmepuro, J. and Nystn, C., (2003), an interview with Department Manager Jorma Halmepuro and Expediter Christer Nystn from Jaakko Pyry Oy on 23rd January 2003

17

ISI Industry Software, (2003), Consignment Tracking for Heavy Industry, News Release of Oy ISI Industry Software Ab, 10. February 2003, available at: http://www.isi.fi/news/kvaerner.html Jahnukainen, J., Lahti, M., and Huhtala, M., (1995), LOGIPRO : towards world class make-to-order supply chains , Research report, no 162, Helsinki University of Technology, Industrial Economics and Industrial Psychology, Finland. Johnston, R.B., and Brennan, M., (1996), Planning or Organising: the Implications of Theories of Activity for Management of operations, Omega, International Journal of Management Science, Vol. 24, No. 4, pp. 367-384. Kaplan, R.S., (1998), "Innovation Action Research: Creating New Management Theory and Practice", Journal of Management Accounting Research, Vol. 10, pp. 89-118. Koskela, L., and Howell, G., (2001), Reforming project management: the role of planning, execution and controlling, Proceedings of the 9th International Group for Lean Construction conference, Kent Ridge Crescent, Singapore, 6-8 August, 2001. Koskela, L., and Howell, G., (2002), The underlying theory of project management is obsolete, Proceedings of the PMI Research Conference, 2002, pp. 293-302. Krkkinen, M., and Ala-Risku, T., (2003): "Facilitating the integration of SMEs to supply networks with lean IT solutions", Proceedings of eChallenges 2003, October 2224, 2003, Bologna, Italy, pp. 1341-1348. Krkkinen, M., Holmstrm, J., Frmling, K., and Artto, K., (2003), "Intelligent products a step towards a more effective project delivery chain", Computers in Industry , Vol. 50, No. 2, pp. 141-151. Santos, A., Powell, J.A., Sarshar, M., (2002), Evolution of management theory: the case of production management in construction, Management Decision, Vol. 40, No. 8, pp. 788-796. Voordijk, H., (1999), "Preconditions and dynamics of logistics networks in the Dutch building industry", Supply Chain Management, Vol. 4, No. 3, pp. 145-154.

18

Vrijhoef, R., Koskela, L., (2000), The four roles of supply chain management in construction, Journal of Purchasing & Supply Management, Vol. 6, pp. 169-178.

19