|International Conference on Construction Process Re-engineering,
12 - 13 July 1999, Sydney, New South Wales, Australia - The University
of New South Wales
REDUCING CYCLE TIME:
Aguinaldo dos Santos
Carlos Torres Formoso
Miguel Torres Curado
This paper discusses the application of the production management principle of reducing process cycle time in construction sites. In its initial part, the paper presents a theoretical framework in which high abstract principles and low abstract tools can be integrated. The assumption behind this theoretical framework is that the current production management philosophies and methodologies (e.g.: JIT, TQM, Re-engineering) have a common core.
Subsequently the paper discusses the implementation approaches related
to the reduction of cycle time based on a number of case studies carried
out in Brazil and in England. Its main findings are concerned with the
degree of application of this principle in the industry and how it interacts
with other production management principles from the emergent production
Keywords: Cycle Time Reduction; Lean Construction; Production
1. SEARCHING FOR THE THEORY-IN-ACTION
A careful reading of the literature reveals that many of the core concepts and principles behind world-class practices (theory-in-action) can be traced to early theories developed during the Industrial Revolution, whilst others actually contain new knowledge. Particularly in the last thirty years, this theory-in-action has been synthesised in a long array of new ‘methodologies’ such as JIT, TQM, Value Based Management, Reengineering, Lean Production, World Class Manufacturing, Concurrent Engineering and so on.
It is possible to identify a common core among these methodologies that
may vary in arrangement and structure depending on the objectives of production
managers (Womack, Jones & Roos, 1990; Koskela, 1992; Sohal, 1996).
In this context, an attempt to establish a logical structure to this common
core is illustrated in the Figure 1, following a hierarchical abstract
distribution (Koskela, 1992; Lilrank, 1995).
‘Concepts’ are located at the highest level of abstraction in the theoretical framework. They could be defined as the mental image of anything formed by generalisation from particulars (Weihrisch & Koontz, 1993). They set the fundamental structure of our thinking and, in that way, they are the filters that we use to understand and analyse reality. Thus, the description of a concept has to be precise enough, in order to avoid ambiguity and, paradoxically, it also has to be flexible enough to allow adaptation to different purposes.
Based on each concept it is posssible to generate a number of ‘principles’ that are, also, highly abstract in their definition. In general, a principle is declared when the generalisations or hypotheses have been tested or observed in practice and appear to be true. Although not always unquestionably or invariably true, principles have to reflect reality and being considered valid enough to be used to predict events (Weihrisch & Koontz, 1993). In short, a principle describes the pathway to transform reality towards the vision set by a concept.
An 'approach for implementation' is the direct and pragmatic answer to the question: "How to implement this principle?". Thus, in the proposed theoretical framework an approach has a much lower level of abstraction in comparison to concepts and principles. It has a straightforward application and, in many instances, the expected result of the action is already in the definition of the approach itself.
In practice, concepts, principles and approaches for implementation
are mixed in different ways, making part of what is called in this paper
‘methodologies’. Methodologies may be found in practice given the
structure and content for a particular practice or as the knowledge embedded
in a complete production system, as it is the case of the Toyota Production
System. JIT, Re-engineering and TQM are examples of methodologies conforming
to this definition. The development and application of each 'approach for
implementation' is supported by a number of tools and tecniques
that are designed to help the determination of specific answers to specific
2. CORE CONCEPTS
The core notion behind the modern production management theories is the concept of value. Nowadays there is an increasing need to expand the customer's perception of a product's worth vis-à-vis its price. In this sense, within this model the increase of value is obtained through systematic analysis of customer wishes and subsequent transformation of these wishes in product and service specifications (Koskela, 1992). Only the customer can define value and it is only meaningful when expressed in terms of a specific product, at a specific price and specific delivery time (Womack & Jones, 1996).
In contrast, the traditional production management philosophy understands production as a set of conversions from inputs to outputs. The conversion model accepts that production can be divided into sub-processes, which are also conversion activities (analytical reductionism). Additionally, this model admits that cost can be minimised by minimising the cost of each sub-process; and that the output value of a process is associated with the costs (or value) of its input (Koskela, 1992).
The conversion model has very deep roots in the Western thinking. Descartes second rule already proposed the division of a problem into as many parts as possible and necessary in order to devise a solution for it. One of the most important practical implications of this model lies on its excessive emphasis upon processing activities and lack of attention on the non-value adding activities that, usually, take significative part of the production resources. Therefore, the direct consequence of relying solely on the conversion model for managing a production system is the generation of waste.
Finally, among the main core ideas of modern production management is the concept of flow where production is understood as constituted of processing, waiting, inspecting and transporting activities. According to this model, processing activities are the only ones that actually add value to the customer and, consequently, waiting, inspecting and transporting activities should be eliminated from the production system (Koskela, 1992; Womack & Jones, 1996). It is one of the principles derived from this core concept that is explored in the next section.
THE PRINCIPLE OF CYCLE TIME REDUCTION
The cycle time of a process is the period required for a particular piece of material or information to traverse a flow and it is composed of processing, inspecting, waiting and moving activities (Koskela, 1992). Thus, the principle of cycle time reduction consists on the elimination of non-processing activities (inspection, waiting, moving) and the increase of efficiency in processing activities from a given process cycle. When looking at all business processes, the process cycle can be measured by the time between an order is placed and the time the customer receives the product or service.
The reduction of cycle time is one of the basic improvement rationales behind the new production management philosophies. One of the most traditional approaches for achieving it, is acting on the processing activities themselves, by changing the technology (conversion model). However, non-processing activities (inspecting, waiting and moving) are usually the most time-consuming ones in production systems and, above all, they do not add any value to the end customer. Thus, the elimination or, at least, the minimisation of non-processing activities is, in general, the most significative step to reduce cycle time (Koskela, 1992).
3.2 Impact on Production Systems
Parkinson’s law dictates that work tends to expand to fill all the time available for it (Weihrisch & Koontz, 1993). In general, this expansion creates waste in the form of movements, waiting and rework. Because of that, compressing cycle time can drives the reduction of waste in production systems (Koskela, 1992). Additionally, Koskela (1992) recognises that the benefits of reducing cycle time are: fast delivery to the customer, reduced need to forecast future demand, and decreased disruption of the production process due to changed orders. Shorter cycle times offer an easier management because there are fewer customer orders to keep track of.
From the perspective of continuous improvement and learning, the time compression has a very important benefit: the cycle deviation-detection-correction becomes shorter. People perceive the results of their actions sooner and, consequently they can act sooner if any correction is necessary. Therefore, shortening the cycle time has a clear connection with lower process variability (Koskela, 1992).
In traditional companies, with rigid departments and numerous organisational layers, the cycle between deviation, detection and correction sometimes may never happen. This is mainly due to the lack of communication or to the existence of long channels of communication which produce distortion in the content of the messages (Koskela, 1992).
"Implementation approaches" are, essentially, ways of doing things or mechanisms for turning abstract principles and concepts into reality. The literature shows various strategies for reducing cycle time, but very often they are not strictly focused on cycle time and for that reason, there is a risk of them not being effectiven. The list of approaches proposed by Koskela (1992) was adopted in this research project because it presents a more strict relationship with the principle. A summarised definition for each of these approaches is presented below:
Table 1 - Approaches for Process Cycle
One important pioneer in the application of these principles was Frank Bunker Gilbreth. He was also interested in the search for the best way of doing a given task. As a building contractor, he became interested in the study of needless, ill-directed and ineffective motions in construction processes. In his most famous study, he analysed the bricklayer’s motions, reducing them from 18 to 5. With these improvements, he doubled the productivity of bricklayers without the need for increasing their efforts (Gilbreth, 1911). The motion studies of Gilbreth allied with the time studies of Frederick Taylor gave birth to the Scientific Management School.
Later, in the Japan post-war, these principles were pushed towards new
limits due to the shortage of resources and the need for increase quality
and productivity. In the same period new approaches were developed such
as the reduction of batch size, the reduction of work-in-progress and the
isolation of value adding activities from supporting activities. Two of
the most proeminent figures of this period are Shigeo Shingo and Taichi
Ohno, developers of the Toyota Production System (see Table 1).
3.4 Theoretical Interactions among the Implementation
It is reasonable to expect that principles derived directly from the same concept should be complementary to each other. Indeed, in theory each principle and the correspondent 'implementation approaches' should reinforce in different levels of intensity all the principles derived from the same concept. Figure 2 indicates the degree of interactions between cycle time reduction and other three principles, based on a literature review carried out in this research project
Figure 2 - Theoretical Interactions among the Four Core Principles
Following this logic, a hypothesis can be made that the maximum effectiveness
of the production management theory in practice would be achieved when
all principles were applied simultaneously in a holistic and coherent fashion.
4. Application of the Principle in Construction
4.1 Research Method
This research can be classified as a ‘testing-out research’. In this type of research, the aim is trying to find the limits of previously proposed generalisations and, subsequently, to specify, modify or clarify their content (Phillips & Pugh, 1994:49/50). As a 'testing-out' research, it has to be carried out in ‘real world’ conditions where the kind of control that exists in a laboratory is not feasible and even not ethically justifiable. Thus, the decision was made to adopt a case study research strategy. The need for a large volume of evidence in order to achieve satisfactory cover up of of all implementation approaches led to the choice of multiple case study design.
Six case studies were carried out, three in Brazil and three in England, based on the bricklaying process. Complementarily, examples of bad and good practice from other studies out at the Federal University of Rio Grande do Sul, Brazil, were used as part of a meta case. A protocol was devised in order to guide the observation sistematically. It includes a number of tools, such as video recording, photography, work sampling, performance measurements, flow chart and open-ended interviews. This protocol also specified the sequence in which they are applied, and guidelines for analysing data and for presenting it. Statistical significance was not considered in the data collection. This protocol will be presented and discussed in a forthcoming paper.
The analysis of the multiple case studies was carried out comparing the practice with the production management theory in study (and vice-versa). In order to do that, the process that Yin (1994) calls replication logic was employed, similar to that used in multiple experiments.
4.2 The Applicability of the Implementation Approaches in Construction
Time (particularly the avoidance of delays) was among the main concerns
of production managers in all case studies. Indeed, the literature also
recognises that delays are often one of the most common and costly problems
in construction projects (Alkass; Mazerolle & Harris, 1995). However,
in the case studies managers gave little attention to actions aiming at
reducing cycle time. The few actions observed in this regard had been produced
previously to construction, during the design and planning stage. Case
Study 02 presented the largest number of patterns matching with the theoretical
propositions, nevertheless it was still lacking integration with other
core implementation approaches, as illustrated in the Figure 3.
Figure 3 - Approaches and Interactions Identified in the Case 02
The principal aim of construction managers was to achieve the right design, under the right budget and schedule specifications rather than to challenge it. Since waste and delays always seem to be included in estimates and schedules, the result was the perpetuation of poor historical levels of performance. Consequently, the benefit of developing more effective solutions continued to be shadowed by inefficiencies in the construction stage.
In theory, minimising distances should be the most straightforward of the approaches for reducing cycle time. However, the fact that construction uses mostly mobile workstations demands a more sophisticated interpretation of this approach in order apply it in an effective way. The minimisation of distances in construction workstations are more likely to be successful if they involve both passive (e.g. training, planning, visual indications, etc) and active means (e.g. physical barriers, limit switches, etc). The passive means are necessary because the maintenance of short distances throughout the process relies heavily on the commitment and attention of the workforce. Ideally, the workplace and workstations optimal arrangement should become entrenched on the workforce practice and knowledge, without the need for managerial control. The active means are those that should be built into the workstation and be easily transported with the workstation to the next workplace. They act independently of the workers own will, but should be designed in a way that make workers benefit from a more effective work.
In most of the case studies investigated little emphasis was given to the reduction batch size. In fact, the idea of reducing the batch size was not even considered as an alternative to reducing cycle time by most of the practitioners. One of the obstacles for implementing this approach was the fact that the short term base relationships between subcontractors and contractors were incompatible with the long term learning process demanded from the reduction of batch size. The mass production paradigm was another barrier for implementing small batches. Construction managers did not understand that the use of resources in their full capacity, producing more than the necessary for the next process, is a waste. In fact, for some managers, waste due to overproduction does not exist in a construction project since the sub-products have to be processed anyway. The benefit of immediate identification and correction of errors was not even mentioned with respect to batch size reduction. This situation indicates that considerable efforts still have to be made in the area of education and training in order to include this practice in the knowledge archetype of managers and workers in the construction sector.
In construction the workstations are normally mobile and the product has different shapes. Therefore the definition of what is the batch unit in the sector is a very complex issue. The size of the batch unit is dynamic, since design and technological innovations can push the already fragmented process even further and consequently, affect the characteristics of transfer batches. Rigid process dependencies, for instance, are often reduced or eliminated with innovations, opening the opportunity for reducing the size of batches. In this respect, it was found that the use of physical or virtual delimitation of batches is an extremely helpful instrument to make the batch unit visible and workable in the construction site.
It was also found that the reduction of work-in-progress seems to be easily misunderstood, especially with respect to the 'consolidation of steps in the process' achieved by technological innovations. Truly, the more segmented and interconnected a process, the more waiting time there will be due to the queue effect. In this sense, technological innovations have surely an important role in reducing the number of visits of each gang to a workplace. However, the principle of reduction of work-in-progress, as defined in this paper, has an emphasis on managerial aspects of production and could be applied even if the technology remains unchanged. The actions are focused on the production flow, aiming at the reduction of process steps waiting to be finished in each single batch unit.
The levels of work-in-progress found in the case studies were strongly linked with the levels of process variability. Indeed, the construction sites that presented the highest predictability of starting and finishing dates and effective communication between all parties were the ones that also presented the minimal waiting time between stages of the process. In theory this could mean a more intense use of standardisation, systematic elimination of root cause of variations, control of plans closer to process owners and poka-yoke devices. However, in the case studies there was little use of these approaches. The best case studies achieved lower variability due to an experienced workforce and to the high quality of materials. Yet, the segmented division of responsibilities and the individualistic bonus system in all case studies resulted in the cumulative re-pass of high process variability from upstream operations to downstream operations.
The lack of continuity (smooth flow) of work across different projects was another important cause of variability in all the case studies and this consequently led to longer cycle time. It was rare to find examples of managers trying to keep a continuous and stable process workload across different projects. It was also found that the problem of continuity was also affected by the lack of flexibility of the workstations, including the workers themselves. Flexibility could offer a viable alternative to keep the workers in touch within the same company (ies) and reduce the effect of changing working environment and practices.
Increased speed of processes by solving control problems was rarely identified in the case studies. At the managerial level, controls were restricted to schedule and inventory control. Such managerial controls did not affect the main adding-value flow of everyday activities. From an operative perspective, activities such as 'measurement' and 'adjusting position' could be considered as controls. Thus, there is substantial room for improvement in the site practice in this respect, taking into account that they do not add any value. In this sense, the process controls should be built into the process, taking no time from the flow of materials. The principle of transparency has an important role in this respect.
Mapping the process was identified as an important instrument to identify the main flow of value adding activities. It becomes much easier to identify the exact location of supporting activities with such information at hand. Mapping also helped to compare the process design with other alternative designs and to identify the opportunities to change the process order. The process activities were sequential in most of the case studies and, according to the observations, the main reason was the high process variability and the push mode of production. The variability of upstream operations were generally absorbed or corrected by downstream operations. However, the change from a sequential order to a parallel one certainly demands better design, low process variability, higher level of skills from the work force and more accurate suppliers than the usual practice. Changing of process order without these basic conditions can bring even more disruption to the production system as it was witnessed in Case Study 06.
4.3 The Theoretical Interactions among Implementation
The analysis of practices in the six case studies has revealed a common problem: the lack of integration of the practices regards to the theoretical 'implementation approaches'. Indeed, the practices that were affecting more severely the performance of the bricklaying process have confirmed this finding. As Figure 4 shows, these practices were lacking the support from other essential practices in order to offer the best possible results. Consequently, the main outcome of having a partial application of the theory was the sub-optimal performance of the production system.
It is understood that the main reasons for this lack of systemic integration are twofold. First, there may be an actual gap in the knowledge of practitioners in all sites analysed. In this respect, there are considerable learning opportunities for the introduction of new ideas in the constructin practice. The second possibility is that the workers and managers knew about the principles in study but did not have organisational structure and support to apply them thoroughly.
In reality, the empirical observations have shown that the construction
sites analysed had a mix of these two possibilities. The excessive emphasis
on the control of process inputs/outputs had deprived managers from the
experience of improving flows. Therefore, important knowledge has not been
properly developed and consolidated in their mental models. In addition,
in certain points it was evident that the managers and workers understood
and agreed with the propositions. In those occasions, their experience
in previous building sites was often used as example to confirm the validity
of the theoretical propositions.
Figure - Approaches and Interactions Identified in the Case 01
The lack of complementary approaches in the case studies was directing
the production systems towards sub-optimal conditions. Neither contractor
was benefiting from these conditions nor the subcontractor or the client.
In this context, the observations in the construction site led to the conclusion
that the short term attitudes, lack of commitment and win-lose relationships
were the main reasons for the absence of systemic application of all inter-connected
5. CONCLUDING REMARKS
The production management theory could be considered as a condensed diary of experiences, observations and thinking on better ways to manage production systems. The adequate understanding of the evolution and state of this condensed learning is an important factor to devising innovative and effective solutions for production problems.
In spite of this, production managers often tend to ignore the historical evolution of management theories, since they are more oriented towards action in the present. Nevertheless, disregarding the evolution of events and ideas of management throughout history is to risk repeating the same mistakes again or wasting useful ideas. Indeed, careful reading shows that all approaches for reducing process cycle time presented in the current literature can be traced in the past theories.
Individually all approaches for reducing cycle time presented in the
operations management field were found applicable in the construction environment.
Yet, the construction sites revealed very few examples of reduction of
batch size and reduction of work-in-progress. Furthemore, it was observed
a generalised lack of systemic integration among all implementation approaches.
The main outcome of this situation was production systems working under