Evaluating the effect of construction process characteristics to the applicability of lean principles
The Authors
Abdulsalam A. Al-Sudairi, King Faisal University, Dammam, Saudi Arabia
Acknowledgements
The author would like to thank King Faisal University for the research grant that supported the development of this research study.
Abstract
Purpose – This paper seeks to test the applicability of lean principles to simple construction processes using discrete-event simulation.
Design/methodology/approach – Quantitative construction data and process mapping of plastering and block-laying processes were first gathered and established from construction project through field observation and interviews with those involved in the selected projects. Then a simulation model was built to mimic the aforementioned processes to study the impact of certain lean principles. The simulation models became like an experimentation tool where lean principles (e.g. focus on actual objects and map the value stream) were introduced to evaluate their impact on such processes.
Findings – Lean principles are effective not only in complicated processes, as proved in previous studies, but also in simple processes. Enhancing the flow of construction materials means the less time they will spend in the value stream and as a result the leaner a process will be. In fact, simple processes are good candidate for lean improvements.
Research limitations/implications – Simulating lean principles did not bring different construction processes to the leanest level of performance. There are other factors that govern each process. Rework, uncertainty, labor skills, site conditions and location are some examples of such factors that need further analyses for leaner construction processes.
Originality/value – Many studies focused on complicated processes to investigate the applicability of lean principles to construction. Results of these studies affirmed the great potentiality of such principles in improving construction processes. This study readdressed the issue of lean applicability to construction by focusing on simple processes, which are block-laying and plastering.
Article Type:
Research paper
Keyword(s):
Lean production; Process planning; Simulation.
Journal:
Construction Innovation
Volume:
7
Number:
1
Year:
2007
pp:
99-121
Copyright ©
Emerald Group Publishing Limited
ISSN:
1471-4175
Introduction
Evaluating the applicability of lean principles to construction is not new. Many studies (Tommelein, 1997; Tommelein, 1998; Crowley, 1998; Al-Sudairi, 2004; Salem and Zimmer, 2005) proved that these principles made construction processes more efficient where in certain examples project duration was reduced by 20-55 percent and productivity increased 10-37 percent (Kartam et al., 1995). Generally, these studies utilized simulation as a tool to evaluate the impact of implementing certain principles to relatively complex processes (e.g. pipe-spool installation, steel erection and so forth).
This study extended the issue of lean applicability to construction by focusing on two different processes (block-laying and plastering). Construction encompasses different processes with different requirements, structures, and logics. Accordingly, construction processes could range from simple to complex that may influence the potential improvement of lean principles. For example, laying concrete blocks is not like erecting steel members. The flow of concrete blocks is not as critical as the flow of steel members. Almost each steel member has few designated locations in a building structure as opposed to concrete blocks that could go any where in a masonry wall unit. One may argue that the application of lean principles to simple processes is not as effective as in complex ones. Having said that the primary objective of this study is to seek an answer for the following question: do process characteristics influence lean application to construction?
Therefore, two simple processes were selected and modeled using discrete-event simulation. The selected processes included plastering and block-laying; they are considered simple because of two main reasons. First, they require simple equipment. That is, both processes mainly depend on the craftsmanship of workers. Second, the materials used in the two processes are totally interchangeable. The degree of material interchangeability is one characteristic that may pose certain requirements to construction processes in which this study adopted to characterize processes.
In fact, Tommelein (1998) emphasized that flow data must be more or less detailed depending on whether the material of concern will be available in large quantities, interchangeable units (e.g. concrete blocks; electrical conduit), in modest quantities, possibly with some degree of interchangeability (e.g. windows, structural steel), or in small quantities of units with unique properties (e.g. engineered materials such as pipe spools). Thus, the construction crews may encounter difficulties if the material of concern is less interchangeable. This is because such crews must find resources that match among those available to them and they must ensure that the right material gets put in the right place (Tommelein, 1998). These difficulties do not exist in block-laying or plastering processes because the construction crew neither needs to keep track of concrete blocks/mortar-mix on site nor to wary where to lay/plaster them in the masonry wall unit.
In order to reveal an understanding of lean applicability to construction, simulation models were created for the aforementioned processes. The simulation models became like an experimentation tool where lean principles (e.g. focus on actual objects and map the value stream) were introduced to evaluate their impact on such simple processes.
To construct simulation models, data were collected from different construction sites located in Dammam metropolitan area, Eastern Saudi Arabia. The data were collected from 13 case studies. Appendices 1 and 2 summarizes the data gathered from such projects.
Previous work
Focusing on information and materials flows is one of the main principles of lean production theory (Crowley, 1998). In fact, Farrar et al. (2004) considered lean production as a flow-based management. The amount of information associated with construction materials vary considerably. For instance, keeping track of concrete blocks or tiles is not as important as tracking pre-cast units or steel members. Specifications of the latter materials are vital to construction crews off- and on-site. Recognition of this fact encouraged many construction researchers to evaluate the application of lean principles to different construction processes with different types of materials.
Iris Tommelein (1997, 1998) examined pull-driven techniques to pipe-spool installation process using simulation. Custom built spools were first engineered off-site and then delivered to site to be installed in a specific place of the project, which indicates the level of complexity for this process. The lean-production “pull” technique has shown to improve performance of construction process by reducing buffer size and project duration when compared to traditional process. As a result, the original project duration of 397 days was reduced to 304 days when pull techniques was implemented. This shows that lean principles are indeed applicable to complex processes.
Similarly, Alves and Tommelein (2004) focused on another complicated process, namely detailing-fabrication-installation of HVAC ductwork. According to Alves and Tommelein this process is complicated because ducts are usually large, so they are difficult to store, and parts are tested separately but the entire work must be put together only at its final location. Again simulating this process revealed the potentiality of lean principles to improvement.
Al-Sudairi et al. (1999) and Tommelein and Weissenberger (1999) examined steel erection processes in relation to buffer size and where to locate buffers in order to reduce non-value adding activities leading to leaner processes. Another similar study by Al-sudairi et al. (2000) investigate the effect of lean production principles on construction projects that manifest different degrees of project complexity. Three structural steel projects were evaluated using discrete-event simulation. The projects' configurations ranged from simple to complex. Application of lean practices made the case study processes more efficient in proportion to the project complexity.
From the previous summary of the literature, one may realize that all of the aforementioned studies focused on complex processes. One common aspect of pipe-spools, HVAC ducts and steel members is that they are less interchangeable. This study readdresses the issue of lean applicability to construction with respect to different construction processes (block-laying and plastering) in order to develop a broader understanding of lean principles to construction. The materials of concern in these two processes are concrete blocks and mortar mix that are totally interchangeable.
Methodology
Figure 1 shows the research methodology adopted to construct valid simulation models for analyzing lean applications. On-site, off-site interviews and field observations of 13 case studies provided an empirical basis for such models. These case studies are projects under construction located in Dammam metropolitan area, Eastern Saudi Arabia. Included projects are houses and apartment buildings constructed in lump sum contracts. These projects are low rise buildings with floor area ranging from 200 to 800 m2 that together represent the diversity of residential buildings in Saudi Arabia.
A total of 60 in-depth interviews were carried out with experts and practitioners who were involved in the studied projects or experienced similar ones. According to Miller (1996) the number of interviews carried out in this study is sufficient. The interviewees were randomly selected based on one criterion that they should have at least five years of experience. The questions aimed at understanding the logic and the structure of block-laying and plastering processes.
Field observations included personal observations and video recording. Personal observations were generally useful in understanding the studied processes as a whole, whereas video recording were useful in timing certain activities especially in the block laying process.
Finding an appropriate distribution to duration of activities was done through two methods:
- fitted distribution; and
- estimated parameters of a given distribution.
There are many computer packages used for fitting a statistical distribution to a sample data. The availability of such packages makes the process of fitting distributions to a sample of observed data quick, easy and accurate (AbouRizk and Halpin, 1990). Stat:Fit is one commercial package that fits a wide variety of distributions to sample observations, which was used in this study.
Estimating the minimum, maximum and most likely parameters of a β distribution is the second method used to represent randomness in activities duration. The reason behind using β distribution is because of its adequacy and flexibility for most construction activities (AbouRizk and Halpin, 1992; AbouRizk et al., 1994; Alkoc and Erbatur, 1997).
Interviews and field observations are interrelated techniques. For instance, it is not enough to establish process maps using only one technique. A preliminary process map was first established through thorough observations that was then refined and validated through discussions with selected interviewees; i.e. project superintendents, engineers and foremen because they are aware of the whole studied processes.
Constructing the initial as-is simulation model required both duration data and process map (Figure 1). It is considered initial because it is not yet verified and validated. Model verification means that the model does not contain logical errors and it operates as expected. On the other hand, model validation means that the model reflects the actual system. Reaching a verified and valid model was not straightforward. The initial as-is model went through several iterations.
Introducing lean principles to the valid as-is model led to a new model, which is referred to as the lean model. Results of both as-is and lean models were compared with each other to evaluate the potentiality of lean principles with respect to block-laying and plastering processes.
Process mapping and data collection
The purposes of process mapping are to:
- identify activities, decisions, queues and resources required;
- clarify process sequence and logic; and
- seek opportunities for improvements (Damelio, 1996; Hines and Rich, 1997).
Based on field observations and discussions with practitioners, process maps for block-laying and plastering were established. Such maps played a role in visualizing the work process flow and made discussion with interviewees very easy and fruitful. It was possible to illustrate the logic and the interrelationships of activities in the two studied processes. The author then facilitated further discussion by asking the following questions, which are mostly developed from Back and Bell (1994):
- What must be finished before this activity can begin?
- Can this activity occur concurrently with any other?
- What resources are required to perform this activity?
- What are the deliverables of these processes?
- How are the deliverables transmitted internally and externally?
- How often does certain activity have to be redone?
- How long does it take to finish an activity?
Process mapping went through two main levels. Firstly, a macro-process map that aimed at major sub-processes as shown in Figures 2 and 3. The block-laying process consists of five major sub-processes:
- order material;
- transport material;
- unload/store material;
- prepare working area and resources; and
- lay blocks.
On the other hand, plastering process consists of six major sub-processes where the first four are similar to those in block-laying; the last two sub-processes are roughening and plastering.
Secondly, a micro-process map that aimed at detailing sub-processes was established (Appendix 1). There are 24 actions (i.e. decisions, activities and queues) in the block-laying process and 35 actions in the plastering process. In both levels of process mapping, several interviews were conducted with those involved in performing the job in order to make sure that maps do mimic reality.
After having the logic of process flow diagram completed, it is time to determine quantitative data related to each activity, and decision. To do so, a detailed comprehensive field survey was conducted. As mentioned in the methodology section there were several techniques used to obtain the necessary information. One of the techniques is time-lapse motion picture. This technique proved to be advantageous in studying activities associated with short cycle lengths (AbouRizk and Halpin, 1992), which is the case in block-laying and plastering activities. The technique is based on filming such activities and then measuring cycle times for each activity. This is done by reviewing the tape and defining and quantifying the different cycle times to identify activities' durations.
It is apparent from the detailed process maps in Appendix 1 that both processes contain many activities. For the sake of research only major repetitive activities were measured using time sheets that records cycle time for different cycles of such activities; data of such activities are presented in a form of histograms as shown in part (a) of Figures 4-7. These duration histograms are then statistically analyzed to find the best fitted distributions for these major activities (part (b) of Figures 4-7).
The best fitted distributions for applying mortar to one block, to lay one block, roughen one square meter, and to plaster one meter square are Log-Logistic, Pearson V, γ, and Pearson VI, respectively. The parameters of these distributions are clearly presented in Tables AI and AII of Appendix 2.
According to Law and Kelton (2000) finding the best fitted distribution is one of the most efficient ways to a credible simulation model rather than relying on empirical distributions. This is because empirical distributions have drawbacks:
- they cannot generate values outside the range of the observed data; and
- empirical distribution may have irregularities.
The remaining activities' durations were estimated by experts where they asked to give three times (most likely, maximum, and minimum) for each activity as shown in Table I and II of Appendix 2. These activities usually happen once or twice in the whole day (e.g. setting up scaffold), so that is why the author relied on estimate rather than measurement.
Simulation model
To model the block-laying and plastering processes data acquired from site surveys and interviews into simulation requires transferring the information into modeling notation. Each simulation package has its own form of activity notation or language which describes the precedence logic of the process network (Back and Bell, 1994). For this study, Extend + BPR® was selected as the simulation modeling package because of its simplicity of use and its adaptability in modeling lengthy complex processes (Abdulhadi, 1997). Hansen (1997) specified other features of Extend + BPR which are:
- utilizes iconic (graphical) blocks that are connected to form a model;
- utilizes an underlying IF … THEN … ELSE language to determine the behavior of the model;
- uses dialog windows to specify behavior and enter process variables; and
- is object oriented and allows attributes to be assigned to items.
Attributes are characteristics of items, such as color, height, quality and so on.
The most important parts of any Extend + BPR model are the blocks, the libraries where block are stored, the dialogs associated with each block, the connectors on each block, and the connections between blocks. A block specifies an action or process; it is used to represent a portion of a model. Some blocks may simply represent sources of information. Others may modify information as it passes through them. In other words, a block is a high-level modeling element for building simulations without programming (Extend-user's Manual, 1997). Information comes into the block and is processed by the program that is embodied in the block. The block then transmits information to the next block in the simulation.
Moreover, almost all blocks in Extend + BPR have input and output connectors, the small squares attached to the sides of a block (Figures 8 and 9). Input and output connectors are usually pre-defined; their function is known in advance. Connection lines are used to hook blocks together; they show the flow of information from block to block through the model (Extend-user's Manual, 1997).
To simulate construction processes, a modeler needs to identify flow unit. A flow unit could be a cubic meter of concrete or a steel member (Halpin and Riggs, 1992). In the block-laying model the flow unit is a concrete block. However, the flow unit in the plastering model was not straightforward; for simulation purposes, it was assumed to be one square meter, which almost represents a catchments area of a plasterer. Thus, the simulation models created for this study are designed to examine the flow of concrete blocks and square meters of surfaces for block-laying and plastering processes, respectively. By analyzing the simulation outputs from each of the process models, it is possible to evaluate the effect of implementing lean principles to this specific type of materials.Successful modeling is wholly dependent on the development of a base-line model that accurately depicts the present work flow process and the interrelationships among various tasks (Back and Bell, 1994). In this study, the base-line model is referred to “as-is” model. The as-is model can be computer simulated and experimentally modified (Houshyar and Nuila, 1994; Ardhaldjian and Fahner, 1994). Before experimenting with simulation to evaluate the effect of lean principles, it is necessary to verify and validate these models. Verification basically implies examining two questions (Back and Bell, 1994), which are:
- Does every transaction go where it is suppose to go under every condition? For example, in block-laying model a mason cannot start laying blocks unless there are enough mortar and concrete blocks. If this is not the case, it means that there is an error in the model. Fortunately, there are features offered in Extend + BPR, like animation that helped a lot in viewing the correct sequence of activities and the logic of resources usage.
- Does every transaction do what it is suppose to do under every condition? This is crucial in almost all simulation models because changes to the as-is model are expected. Implementing lean principles to either block-laying or plastering processes implied changes to the as-is model, which will be discussed in depth in the forthcoming sections. In fact, Chisman (1992) stated that the true benefit of simulation modeling is the ability to explore what-if analysis with respect to a defined process.
However, verification does not guarantee that the model is valid (Houshyar and Nuila, 1994). Validation means that the model is almost behaving like the actual system. A comparison of the model outcomes and the data gathered from both processes on site was made as shown in Figures 10 and 11. Figures 10 and 11 show two sets of data: actual and experimental of total cycle time to lay one block and to plaster one meter square, respectively; the actual data were gathered from studied projects for both processes whereas experimental data were gathered from simulation models. Notice the close correspondence between the behavior of the actual processes and the simulated processes. Results in Figures 10 and 11 prove that simulation models are valid and ready for evaluation.
In both models, block-laying and plastering, the order material and the unloading sub-processes were not included in simulation. Also, both models were constructed so that changes imposed by implementing lean principles can be easily manipulated and hence measure their impact on the selected processes as it makes it possible to answer the question of this study.
Lean principles and simulation
This study simulated few lean principles as listed in Table I and how they have been managed in the simulation models. The principle “focus on actual objects” correspond to the concept of “flow unit” that has been mentioned in the previous section. The block-laying and plastering simulation models were constructed to focus on concrete blocks and m2 of plastered surfaces, respectively. That is, both models simulate the movement (i.e. the processing time and waiting time) of such actual objects from the start as raw materials until the end as being part of the wall unit. The time an actual object spends in the process was one of the essential measurements used for evaluation.
One of the goals of process mapping was to identify value adding activities from those activities that do not. Value adding activities are those that increase the economic value of a process and valued by the customer (Anupindi et al., 1999). Also, value adding activities transform inputs to outputs and are necessary to meet customer's requirements (Tenner and DeToro, 1997). For instance, laying block or roughening surfaces for plastering are examples of value adding activities.
On the other hand, non-value activities are those that do not directly increase the economic value of a process and they are not valued by the customer (Anupindi et al., 1999). This is because such activities increase cycle time and add cost rather than value (Tenner and DeToro, 1997). Laguna and Marklund (2005) classified non-value activities into two types. Type-1 of non-value adding activities, sometimes referred to as control activities, are necessary to the logic and the structure of a process. Preparing work area and transporting and storing materials and equipment are common example of type-1 of non-value adding activities. This type of activities if not eliminated should be kept to the minimum (Womack and Jones, 1996).
In the simulation models, the magnitude of such activities was reduced by reconsidering the sequence of major activities so that there is an overlap instead of having them running serially. In other words, activities are performed in parallel. In practice plastering starts after roughening all surfaces. In many cases workers cannot immediately start plastering, because they have to wait for roughened surface to get harden. However, by dividing the project into smaller segments, workers in one segment can start roughening while plastering is taking place in another segment at the same time. In doing so, waiting time between roughening and plastering is noticeably reduced leading to leaner process.
In addition, block-laying and plastering processes considerably contained type-2 of non-value adding activities. One common example for both processes is that materials go through many storages which increase double handling and distance to working area (Appendix 1). Another example is that workers sometimes cannot start the work because of the lack of certain materials especially in the plastering process. In fact, Ballard (1997) and Alarcon (1997) concluded that one the major causes of lower productivity is lack of materials.
Gregory Hansen (1997) emphasized that no process is straightforward where there are usually conditions that require deviations from normal processing. It is possible that tasks begin based on conditions and terminate based on conditions. Material availability was one condition that influenced the progress in block-laying and plastering processes. Considering such condition in simulation required further investigation.
In both simulation models availability of materials were randomly modeled. To randomly measure material availability, different production cycles (which was laying one block and plastering one square meter) were observed to quantify the impact of materials availability. There were 87 and 79 production cycles for block-laying and plastering, respectively. Out of the 87 production cycles in the block-laying process 12 production cycles, which means 14 percent, were starving for material. Likewise, 10 out of 79 production cycles, which means 13 percent, were starving for materials in the plastering process. Lack of material is one source of waste (i.e. type 2 of non-value adding activities) that was modeled by measuring its probability. In other words, there was a conditional stoppage for resources waiting for the proper materials (Figure 12). With good coordination between suppliers and work at the construction site, delays caused by material could be reduced significantly.
Analysis of results and discussion
Both simulation models allow for the evaluation of lean principles and measure their impact on the studied processes. This is accomplished by comparing cycle time and efficiency of both the as-is models and the lean models. Process efficiency is measured by comparing time consumed by value adding activities to the total cycle time (Tenner and DeToro, 1997), as shown in equation (1). Equation 1 Table II summarizes results of the block-laying model. Applying lean principles to this simple process enhanced its performance. In practice, it took 58 seconds on average to lay one block and 48 seconds in the lean model. This means that there is 17 percent reduction in cycle time that led to better utilization of resources and increased process efficiency.
The block-laying process did not reach 100 percent efficiency because of many non-value adding activities (type 1), such as moving block to working area, which also existed in the lean process. This tells that there is an opportunity for further improvements. That is, lean principles are very effective to construction regardless of different process characteristics.
Likewise, Table III summarizes results of the plastering process, which coincide with results of the block-laying process. There is a noticeable reduction in the total cycle time where it traditionally took 178 hours on average to plaster one square meter as opposed to 136 hours in the lean model. Reduction of cycle time in the plastering process reached 23.6 percent indeed prove that lean principles are effective in simple processes.
In terms of process efficiency in the plastering process, one may notice that there is great improvement that reached 50 percent. This is because the waiting time between roughening and plastering was originally large where in some cases is larger than the time of roughening. The low process efficiency in the plastering process even in the lean model, which is 60 percent, is due to the nature of this process. The time of plastered surfaces to get harden is considered in this study a non-value adding activity. This is because there are other techniques, which are beyond the scope of this study, where cement-mortar can get strength faster leading to less waiting time.
Integrating lean principles holds a significant potentiality in improving construction processes. Eliminating non-value adding activities and not protecting production from uncertainty does not lead to a lean process. Material availability was one source of uncertainty that led to many stoppages and hence increased cycle time.
Understanding a process and its requirements and methods is very important in moving towards leaner processes. Applying lean principles did not lead to a waste-free process in either block-laying or plastering. These processes depend on cement-mortar especially the latter one where waiting time of plastered surfaces to get harden is inevitable. Such waiting time was a major barrier to a continuous flow. Expediting the setting time of plastering and roughening activities requires not only changes in the process but also in the construction methods/techniques that would definitely lead to a leaner process.
In summary, lean principles are not only effective in complicated processes, as proved in previous studies, but also in simple processes. Enhancing the flow of construction materials regardless of their interchangeability means the less time they will spend in the value stream and as a result the leaner a process will be. In fact, simple processes are good candidate for lean improvements.
Conclusion
Many studies focused on complicated processes to investigate the applicability of lean principles to construction. Results of these studies affirmed the great potentiality of such principles in improving construction processes. This study readdressed the issue of lean applicability to construction by focusing on simple processes, which are block-laying and plastering.
Data, including process maps, of block-laying and plastering was collected through interviews with practitioners and field surveys (i.e. personal observations and video recording) on selected projects located in Dammam metropolitan area, Eastern Saudi Arabia. Utilizing video recording is indeed an advantageous technique to time construction activities. However, avoiding worker's bias is essential to data credibility. This data modeled in Extend + BPR, which is a simulation package, to form the experimental tool to further evaluate the applicability of lean principles to construction.
Results of the simulation models showed that lean principles enhanced the performance of the selected processes. For instance, process efficiency increased 21 percent and 50 percent for the block-laying and plastering processes, respectively. Application of lean principles enhanced the flow of work in both processes by reducing non-value adding activities leading to shorter cycle time when compared to the traditional process. This means lean principles are very effective to construction regardless of different process characteristics.
Focusing on actual objects helped in detecting and eliminating other non-value adding activities (e.g. waiting time of resources) that were difficult to find in a process map. For successful focus on objects, this study utilized process mapping and simulation because both tools complement each other.
Process mapping is one useful tool to study and understand flow of construction materials. Using this tool, it is possible to indicate queues, decisions and check points that are usually one source of non-value adding activities. Process mapping is a good diagnostic tool but it fails to provide sophisticated analysis and evaluation. For the purposes of analysis and evaluation, this study utilized simulation to investigate the benefits of lean application to construction. For instance, segmenting work of major activities, which is very difficult to do in other tools, provided a flow with less waiting time when compared to the traditional process.
Simulating lean principles did not bring different construction processes to the leanest level of performance. Each process has its own logic and structure that may require changes in order to improve it. In addition, there are factors that govern each process. Rework, uncertainty, labor skills, site conditions and location are some examples of such factors that need further analyses for leaner construction processes.
Equation 1
Figure 1Research methodology
Figure 2A macro block-laying process map
Figure 3A macro plastering process map
Figure 4Data and fitted distribution for applying mortar to one block
Figure 5Data and fitted distribution for laying one block
Figure 6Data and fitted distribution for roughening one m2
Figure 7Data and fitted distribution for plastering one m2
Figure 8The first part of the block-laying model
Figure 9Arranging resources (plasterer and helper) and entering stochastic times to activities in the plastering model
Figure 10Comparing actual total cycle time of laying one block to simulation result
Figure 11Comparing actual total cycle time of plastering one m2 to simulation result
Figure 12Modeling material availability and mixing mortar activity in the block-laying model
Figure A1Basic shapes used to map construction processes
Figure A3Plastering process map
Figure A2Block-laying process map
Table ISimulated principles and their consequence to block-laying and plastering processes
Table IISimulation results of the block-laying model
Table IIISimulation results of the plastering model
Table AISimulation input to the block-laying model
Table AIISimulation input to the plastering model
References
Abdulhadi, N.H. (1997), "Simulation as a tool for re-engineering construction processes", Department of Civil, Environmental and Architectural Engineering, University of Colorado-Boulder, Boulder, CO, .
AbouRizk, S., Halpin, D. (1990), "Probabilistic simulation studies for repetitive construction processes", Journal of Construction Engineering and Management, Vol. 116 No.4, pp.575-94.
AbouRizk, S., Halpin, D. (1992), "Statistical properties of construction duration data", Journal of Construction Engineering and Management, Vol. 118 No.3, pp.525-44.
AbouRizk, S., Halpin, D., Wilson, J. (1994), "Fitting beta distribution based on sample data", Journal of Construction Engineering and Management, Vol. 120 No.2, pp.288-305.
Alarcon, L.F. (1997), "Modeling waste and performance in construction", in Alarcon, L. (Eds),Lean Construction, Balkema, Rotterdam, .
Alkoc, E., Erbatur, F. (1997), "Productivity improvement in concreting operations through simulation models", Building Research and Information, Vol. 25 No.2, pp.82-91.
Al-Sudairi, A. (2004), "Simulation as an aid tool to the best utilization of lean principles", in Bertelsen, S., Formoso, C. (Eds),Proceedings 12th Ann. Conf. Intl. Group for Lean Construction IGLC-12, Helsingor, Denmark, .
Al-sudairi, A., Diekmann, J., Songer, A. (2000), "Interplay of project complexity and lean production methods", Proceedings 8th Ann. Conf. Intl. Group for Lean Construction IGLC-8, Brighton, .
Al-Sudairi, A., Diekmann, J., Songer, A., Hyman, B. (1999), "Simulation of construction processes: traditional practices versus lean principles", in Tommelein, I. (Ed.), Proceedings 7th Ann. Conf. Intl. Group for Lean Construction IGLC-7, Berkeley, CA, .
Alves, T., Tommelein, I. (2004), "Simulation of buffering and batching practices in the interface detailing-fabrication-installation of HVAC ductwork", in Bertlsen, S. and Formoso, C. (Eds), 12th Ann. Conf. Intl. Group for Lean Construction IGLC-12, Helsingor, .
Anupindi, R., Chopra, S., Deshmukh, S., Mieghem, J., Zemel, E. (1999), Managing Business Process Flow, Prentice-Hall, Upper Saddle River, NJ, .
Ardhaldjian, R., Fahner, M. (1994), "Using simulation in the business process reengineering effort", Industrial Engineering, July, pp.60-1.
Back, W., Bell, L. (1994), Quantifying Benefits of Electronic Technology Applied to Bulk Materials Management, Construction Industry Action Group, Doha, .
Ballard, G. (1997), "Lean construction and EPC performance improvement", in Alarcon, L. (Eds),Lean Construction, Balkema, Rotterdam, .
Chisman, J. (1992), Introduction to Simulation Modeling Using GPSS/PC, Prentice-Hall, Englewood Cliffs, NJ, .
Crowley, A. (1998), "Construction as a manufacturing process: lessons from automotive industry", Computers & Structures, Vol. 67 pp.389-400.
Damelio, R. (1996), The Basic of Process Mapping, Productivity Inc., Portland, OR, .
Extend (1997), Simulation Software for the Next Millennium, Imagine That Inc., Washington, DC, User's Manual for Extend, .
Farrar, J., AbouRizk, S., Mao, X. (2004), "Generic implementation of lean concepts in simulation models", Lean Construction Journal, Vol. 1 No.1, pp.1-23.
Halpin, D., Riggs, L. (1992), Planning and Analysis of Construction Operations, Wiley, New York, NY, .
Hansen, G. (1997), Automating Business Process Reengineering: Using the Power of Visual Simulation Strategies to Improve Performance and Profit, Prentice-Hall PTR, Upper Saddle River, NJ, .
Hines, P., Rich, N. (1997), "The seven value stream mapping tools", International Journal of Operations & Production Management, Vol. 17 No.1, pp.46-64.
Houshyar, A., Nuila, V. (1994), "Well-designed simulation: a tool for superior management, part 1", International Journal of Modeling and Simulation, Vol. 14 No.3, pp.129-33.
Kartam, S., Ibbs, W., Ballard, G. (1995), "Reengineering construction planning", Project Management Journal, June, pp.27-37.
Laguna, M., Marklund, J. (2005), Business Process Modeling, Simulation, and Design, Pearson Prentice-Hall, Upper Saddle River, NJ, .
Law, A., Kelton, D. (2000), Simulation Modeling and Analysis, 3rd ed., McGraw-Hill, Boston, MA, .
Miller, E. (1996), "What is the standard of care?", Journal of Management in Engineering, Vol. 12 No.6, pp.40-6.
Salem, O., Zimmer, E. (2005), "Application of lean manufacturing principles to construction", Lean Construction Journal, Vol. 2 No.2, pp.51-4.
Tenner, A., DeToro, I. (1997), Process Redesign: The Implementation Guide for Managers, Addison-Wesley Longman, Inc., Glen View, IL, .
Tommelein, I. (1997), "Models of lean construction processes: example of pipe-spool materials management", Proceedings of Construction Congress V, October 5-7 in Minneapolis, MN, ASCE, VA, USA, pp.405-13.
Tommelein, I. (1998), "Pull-driven scheduling for pipe-spool installation: simulation of a lean construction technique", Journal of Construction Engineering and Management, Vol. 124 No.4, pp.279-88.
Tommelein, I., Weissenberger, M. (1999), "More just-in-time: location of buffers in structural steel supply and construction processes", in Tommelein, I. (Ed.), Proceedings 7th Ann. Conf. Intl. Group for Lean Construction IGLC-7, Berke, CA, .
Womack, J., Jones, D. (1996), Lean Thinking: Banish Waste and Create Wealth in Your Corporation, Simon & Schuster, New York, NY, .
Appendix 1. Process maps
Appendix 2. Simulation inputs
Corresponding author
Abdulsalam A. Al-Sudairi can be contacted at: aalsudairi@kfu.edu.sa