Content uploaded by Douglas D. Gransberg
Author content
All content in this area was uploaded by Douglas D. Gransberg on Feb 02, 2016
Content may be subject to copyright.
Final Accepted Draft – Published as:
Conte, A.S.I., and Gransberg, D.D., “Lean Construction: From Theory to Practice,” 2001 Transactions,
AACE, Int’l, Pittsburgh, Pennsylvania, June 2001, pp. CSC10.01-CSC10.05.
Lean Construction: From Theory to
Practice
Antonio Sergio Itri
Conte
Lean
Construction
Institute of Brazil
326 Vila
Mariana
Sao Paulo 04012-060
Brazil
phone:
+55.11.5736937
e-mail: asiconte@usp.br
Douglas D. Gransberg, PE CCE
Associate
Professor of
Construction Science
University of
Oklahoma
Construction
Science Division
830 Van Vleet Oval, Room 162
Norman, OK 73019-6141 phone:
405-325-6092
e-mail: dgransberg@ou.edu
Introduction
After World War II, Japan adopted the Toyota Production System (TPS). The system was
based on a scenario of fluctuating demand that required swift assembly line alterations due to
the large number of different parts to be produced. The model developed by Ohno [4] was
infinitely superior to Ford’s mass production system because it provided for the batch
production of small quantities of parts. It required smaller physical areas, fewer resources,
and a smaller inventory of raw material and work in process.
Ohno’s original ideas were based on the adoption of production strategies identified
according to the demand of the down- stream production chain, part of a production plan
that ensured the planned pace was maintained throughout the production process. In other
words, the idea was to achieve a continuous production flow by adopting monitoring
measures for each process phase, aiming to reduce inventories. If we add this rationale to an
attempt to reduce the waste associated to activities that add no value to the final product,
we will obtain a systematic and consistent model—the model that gave Toyota high
performance levels.
In construction, the application of the lean production model stems from a discussion of
Koskela’s work [3], which emphasized the importance of the production process flow, as well
as aspects related to converting inputs into finished products as an important element to reduce
wasted value in jobsites. Production should be seen as a flow that generates value through
conversion processes, characterized by cost, time frame, and the degree of added value. In this
context, considering the high uncertainty typical of the construction sector, it was essential to
adopt management attitudes that are able to make the operating environment stable, reducing
production process variability and significantly increasing the reliability of the production
planning phases, including the jobsite’s internal logistics.
Howell and Ballard’s studies [1] about the “last planner” technique showed that the use
of formal and flexible production planning procedures is the first step to make the production
environment stable, emphasizing, in this case, the use of daily production plans, constraint
analyses-look-ahead, and the percentage of planned and concluded items-PPC as tools for
immediate implementation on any jobsite. A well-executed project mirrors its production
planning precisely. Good production planning comprises planning elements that can
effectively be executed. We would like to add a third aspect: being flexible and able to
readapt your production plan taking into account eventual deviations observed during the
daily sequence of jobsite operations is just as important as planning a project.
The practical results of this technique, expressed by the rise of PPC levels to close to 100
percent, translate into a heightened confidence in short-term planning processes. They
usually can be felt after 2 or 3 months application, and encourage the commitment of the
production team (engineers, foremen, and subcontracted parties), who obtain simple and
efficient tools with which to perform their respective activities.
However, a PPC close to 100 percent does not guarantee that the project is being executed
in compliance with the foreseen conclusion date, cost, and quality. It is fairly common to see
projects with a high PPC that end up moving away from their fore- most targets because
there is no information able to monitor precisely how far the project is from its initial
objectives during every control period. Thus the production teams are unable to change and
adapt their activity plans and recover their grip on the situation originally planned. It also
contributes to make management focus on maintaining the project’s conclusion date, often at
the expense of cost and end result.
LEAN
CONSTRUCTION
IN PRACTICE
In the last 6 years we applied the model proposed by lean construction to more than 20
Brazilian construction companies, always bearing in mind Womack’s basic lean thinking
principles [5]. We sought to define our activities according to the following principles.
Production Planning and
Control
All production planning must be based on maintaining the pace of the work instead of
seeking productivity peaks that improve the performance of a given activity but do not
always ensure the best combination for the project as a whole.
The line of balance (LOB) technique should be used to optimize the study of the pace of
the services to be executed. This technique provides for the immediate identification of
production bottlenecks and eventual buffer insertion points. The aim is to off- set the
differences in pace between the work packages identified for the project. The ideal situation is
when all work packages have the same pace, eliminating inventory that does not really
add value to the end product.
A process design should be devised for every work package planned in the LOB. This
process design will determine the scope of the work to be executed, the daily sequence of
partial targets, the size of the production team, the material, equipment, and tools required,
and the moment when they should be available at the service fronts, the expected quality and
operating performance standards and, last, the attention that must be given to job safety for the
production teams defined. Thus, the solutions adopted can be studied and discussed from
the standpoint of the executive process in a single document, as well as the respective
logistics related to the raw material, labor, equipment, and tools required for the execution of
the services within the specified time frame. The preparation of this document should involve
the engineer in charge, the foreman and all subcontracted parties involved in the scope of
each work package.
Based on the LOB and the process design developed, it is possible to use a new
performance indicator for the project, represented by the projection of a conclusion date
calculated every week. A project with a high PCC whose conclusion dates are under control
is probably in compliance with the expected result for the final execution date, and there isn’t
much uncertainty in the project’s daily operations. This approach gives the production teams
more time to adapt to the next services to be executed and permits the adoption of more
advanced strategies for future negotiations with material and service suppliers, as well as the
study of new technical solutions. It is the key to applying pulled production to construction
and enhances cost and quality performance.
Weekly control of the work should follow the last planner technique. The constraint
analysis involved in the mid-term plans must be very detailed, and will allow the early
anticipation of eventual barriers to the project’s natural pace. There are usually two different
types of barriers, depending on the time when the work package is analyzed.
Type 1: Purchase and Contracting Constraints
Marked by aspects linked to product and/or service, design, technical specs, the purchase
and/or contracting of raw material, labor, equipment, tools, and service specs. These issues
are generally analyzed before the beginning of the work packages in point.
Type 2: Allocation and
Availability
Constraints
Marked by no longer depending on purchasing or contracting but on optimizing internal
jobsite logistics with the aim of ensuring that each planned cycle can be executed
effectively. They should occur after the start of the respective services, and the constraints
linked to material, equipment, tools and especially labor —an important source of
redundancy and waste on Brazilian jobsites —have been analyzed.
A weekly analysis of the PPC leads results in identifying the reasons for the disruption of the
pace observed in the work, and consequently, contributes to systematic learning on the jobsite,
generating a mindset effectively geared to improving competitive - ness.
The project must reassess its strategies every week, after analysis of each subcontracted
party. If necessary, the production logic of the downstream services should be altered. This
can occur in three different ways.
• By inserting or removing resources from the work packages.
• By modifying the relations of precedence among the services to make the
superimposition of activities feasible and reduce the global execution time.
• By reorganizing the different work package activities with the aim of making fewer
simultaneous production cycles feasible.
This significantly simplifies internal jobsite logistics, reducing the size of the work teams
and the possibility of constraints linked to raw material, labor, equipment and tool allocation
and availability.
We can state, based on our experience on Brazilian jobsites, that the third form of
interference in the pace of the work is the least obvious one. It is also the one that tends to
present optimized results for the project due, above all, to the following facts. It provides real
opportunities to reduce the global execution time, because the unification of a work package
fosters the reduction of the equivalent to the sum of the time frames and their respective cycles.
It reduces the opportunity for any interruption of the normal work pace due to internal
logistic problems, and provides real opportunities to reduce production costs. This is because
fewer simultaneous work packages make it easier to estimate the size of the production
support teams, now allocated directly to the foreman instead of only working with specific
tasks. This approach drastically reduces the number of nonspecialized workers at the jobsite,
which in turn, as well as reducing the direct cost of operations, also reduces any waste linked
to non-added value, typical of redundancy when this type of labor is used.
Fewer simultaneous work packages and smaller production teams allow the execution of
more simultaneous services within a single cycle, especially those that do not add value to
the end product but that cannot be eliminated from the original work packages. Fewer
simultaneous work packages drastically reduces the cost of the internal logistics dedicated to
production support; fewer simultaneous work packages reduces the cost of supervision and
quality control by engineers and foremen because fewer service fronts must be contracted at
the same time. The cost related to security requirements in the working environment also
drops, because few service fronts are active simultaneously, and the active ones have
smaller work teams. The cost associated to services, demand for material, and labor drop,
especially because the jobsite’s data-gathering tools become simpler and less vulnerable to
mistakes, as we will see later.
Control Of
Production
Cost And
Raw
Material, Labor,
Equipment,
And
Tools
Consumption
The control of production costs and the material and labor consumption for each task
executed will be based on the demand defined in the process design generated during the
phases of production planning has process designs that effectively mirror field operations.
Thus when a given cycle begins, the management team is able to perceive the daily
requirement of raw material, labor, equipment, and tools for each work front early, generating
a optimizing plan to transport these items and make them available. During this phase
we identify supply bottlenecks and planning mistakes, which should be used as learning
elements for the next occurrence of these work packages. Should there be any planning
mistakes, the process design is adjusted for more or for less, as appropriate. A closed cycle
implies a certain amount of raw material, labor, equipment and tools consumed to execute
a task volume identified in the process design. This explains the ease of gathering data in the
field and, consequently, controlling the production cost as a whole.
The project budget should be organized to provide for the effective comparison of the
real cost based on controlling the execution of the work packages and their cycles. These
reorganized cost budgets are known as production-driven budgets. This approach implies a
complete revision of the standards of each task executed because the productivity of a given
task loses its importance, being replaced by the production team’s capacity to execute the set
of activities defined in the work packages according to the standards determined by the
production plans. The importance of this model lies in the swiftness with which information
is processed, optimizing the project’s decision-making process at all hierarchical levels. The
jobsite will control vital information, speeding up the management pace and reducing the
losses caused by inertia or delays in identifying and resolving problems in the production
process. The transparency generated fosters team commitment and contributes to improve
team members’ quality of life.
Construtora
Hernandez
And The
Gerona
Building—A Practical Case
Of
Lean
Construction
Construtora Hernandez is a 37-year-old family-owned company with main office in
Tatuapé, in the city of São Paulo, Brazil; the firm is dedicated mainly to the construction of
residential buildings. Until 2000, Construtora Hernandez had a built-up area of approximately
210000 m2, comprising approximately 1350 middle and high-class apartments.
The company’s market activity is defined by property acquisition, development, and the
construction of high-rise buildings. It has its own funding system, which makes its operations
feasible. Construtora Hernandez uses its own labor or subcontracts from parties who have
worked with the company for some time. At the start of 1998, Construtora Hernandez began
the construction of the Gerona Building, a high-rise, high-class, 18-story residential building
with four flats per floor and a total built-up area of 14230 m2. When the lean construction
production management model was deployed, approximately 40 percent of the concrete
structure had been executed and bricklaying was about to begin.
Phase 1: Stabilization of the Operating Environment
Initially, our major concern was to stabilize the production environment, defining
production cycles for structural activities and brickwork compatible with the project’s
strategic targets. The reduction of the variability of these tasks was achieved using the last
planner technique and the unconditional support of the company’s directors and production
teams, including the foreman and the main subcontracted parties.
Although during the first weeks there was a high PPC (it rose from 50 percent to
approximately 82 percent in 6 weeks), it was necessary to develop tools to support strategies
aimed at estimating the conclusion date of the project and other extremely important dates,
like the start of the work on the facade of the building and delivering the physical facilities for
the lifts. The idea was to control the progress of the work, ensuring that each control period (a
week, in this particular case) didn’t lose sight of its original target. Optimizing the size and
allocation of the production teams had the aim of reducing the number of workers/hours
required to execute the services.
For this end we devised an Excel spreadsheet using LOB concepts that made it very easy
to run simulations of the plans for each work package, allowing the pace of activities to be
adapted downstream of the production process. The spreadsheet also was meant to ensure
the allocation of work teams for every production cycle and identify idleness or
redundancy in the workers/hours allocated. This led to two management positions:
• the use of surplus workers/hours for unscheduled activities, increasing the speed of
construction; and
• the systematic reduction of the size of the team.
The direct consequence of the use of this tool by the engineer in charge and the foreman
was to reduce the estimated time until conclusion of the project by 2 months (the initial
forecast was 24 months) and the size of the team by 30 percent. In Brazil this signifies an 11
percent reduction of the total project cost —due to the execution technology used in the
country the labor force employed in a vertical building historically represents
approximately 40 percent of the total cost.
From that point on there began weekly control of the conclusion date. We also measured
the interval between the conclusion date foreseen by engineering every week and the
conclusion date the customer expected. Keeping the values of this interval—known as
“lung”—to values close to those defined by Construtora Hernandez at the beginning of
construction guaranteed a small likelihood of delay at the end of the construction, reducing its
global cost, since it would eliminate any activities conducted under pressure or with a
production pace different from the pace planned for each instance [2].
This result was attained 4 months after the start of construction and gave the program great
credibility. A new phase in the development of production management support tools for the
project began, optimizing the time to prepare and discuss daily production plans, especially
the look-ahead and constraint analysis.
There was a weekly 2-hour meeting with the engineer in responsible for supporting
the company’s lean construction pro- gram and divulging the knowledge acquired to other
ventures in progress.
The results were highly positive and the production teams adopted the model proposed. It
made short-term strategies flexible and precise without losing sight of the conclusion of each
phase of the work.
Phase 2: Development of Production Planning Tools and Techniques— Assembly and
Superimposition of
Production
Cycles
The company observed, however, that the natural pace of each of the work packages
planned was uneven. This led to idle - ness and redundancy in the production processes. The
company decided to adopt a new production planning technique consisting basically of
identifying and assembling all the sets of work pack- ages with cycles smaller than the natural
project cycle that could be executed simultaneously by the same work team. The
approach was expected to reduce the construction period and direct production costs.
Below is a practical example that perfectly illustrates this strategy. The idea is to plan the
construction of a vertical building that at present has reinforced concrete structure and
masonry brick- work tasks in progress. We know that during the production phase the
activities in progress and their respective production teams are as follows.
A. Reinforced concrete structure: one floor every 7 days/12 carpenters.
B. Floor finish: one floor every 3 days/2 masons/2 assistants.
C. Preparation for masonry: one floor every 4 days/2 masons/1 assistant.
D. Masonry: one floor every 7 days/4 masons/4 assistants.
It is very clear that the cycle of work packages B and C was systematically interrupted
because the prior services only delivered a new floor every 7 days. The strategy used here
was to assemble work packages B and C with the following results.
• All work packages being executed within 7 days in a perfect pace for all production
teams.
• Reduction of the conclusion date of the work by 4 days.
• Reduction of the production teams, as follows: initial situation, 4 masons plus 3
assistants/closing situation, 2 masons plus 2 assistants.
The gain obtained using this approach in all project production processes had other
benefits, such as the systematic reduction of costs linked to the support teams for production
and internal logistics—with fewer simultaneous work packages in progress there is less need
for labor and equipment to support the planned pace of production. There was more ease in
controlling the end result of the finished work, concentrating a bigger number of simultaneous
activities in a single site. A cost reduction with sanitary and safety procedures on the jobsite
occurred.
The Gerona Building began with 43 planned work packages for the project as a whole.
We ended with 29 work packages, assembled using the above-mentioned technique or
techniques of superimposition of independent tasks.
Phase 3: Control of the
Demand/Consumption
of Raw Material, Labor, Equipment,
and Tools and Production
Cost Control
With the production schedule under control, the third phase of Hernandez’s lean
construction program was to implement a management model providing the immediate
control of the demand for raw material, labor, equipment, tools, and production costs at the
jobsite. The idea was to provide more agility and quality in the decision-making process and
ensure the lowest possible production cost, as well as the shortest construction time and the
maximum end quality.
In Brazil, follow-up and control activities are traditionally conducted by a technical staff
installed at the jobsite specifically for that purpose. They represent a significant part of the
fixed over - head allocated to the work, which does not vary monthly according to the
production volume. Fixed direct costs correspond to 5 to 7 percent of the total project cost.
At the Gerona Building, we tried something different from the traditional model. Instead
of using professionals to collect data concerning the demand for raw material, labor,
equipment, tools, and their respective cost in the field, we used the process designs developed
that defined the expected demand for raw material, labor, equipment, and tools for each
work package, allowing a direct comparison with the original budget.
When a given cycle allocated to a specific work package began production, the
amount of raw material, labor, equipment, and tools required had already been defined,
as well as the moment when they should be available at the tasks fronts. Thus, at the end of
each cycle, the jobsite team simply had to confirm if the real demand had complied with the
standard defined. If the answer were yes, the production cost of that particular work pack- age
cycle was duly identified, ensuring the precision of information for the next cycles to be
generated. If the answer were no, it was necessary to change the original process design to
reflect the actual consumptions. This process made the consolidation of a jobsite learning
process feasible. The latter, in turn, ensures the company’s competitive advantage over time.
Phase 4: Current Situation and Subsequent Targets
The results attained have surpassed the company directors’ expectations. Now they intend
to adopt the same rationale for production phases without repetitive work cycles, like
foundations and underground work and roofing, that represent a significant part of the time
and total production cost.
The company’s procurement sector adapted very well to the pulled production strategy,
significantly reducing the inventory of material at the jobsite. In addition, the certainty that
the production cycles will be met makes it possible to contract a global amount of
material from suppliers, with better end prices and the freedom to schedule partial delivery
according to the expected progress of the work. The company’s design and technical specs
sector held meetings with its designers to develop production-driven designs, reducing the
time it takes for the production teams to understand the plans and reducing eventual mistakes
and rework during construction. The company’s engineering team worked with the project
designers, adding lean production concepts learned on job sites to new projects. After the
conclusion of the Gerona Building, within the fore- seen time frame and with the expected
cost reduction, the company adopted the same strategy for other ventures, attaining important
results like the reduction of 20 percent (in addition to the initial 30 percent) in direct labor
costs.
In 2001, Construtora Hernandez has begun an ISO 9000 total quality management
certification process that involves all quality-related issues according to lean construction
principles. This is a unique experience in Brazil so far and represents a major challenge for the
company’s technical teams. Within this scenario the idea is to emphasize sanitary and job
safety processes and strengthen a production environment driven by competitiveness, not
losing sight of the end customer and the production teams at the jobsite.
The deployment of
production-managed
models based on lean production principles and
techniques is
feasi
ble and can be applied to any type of construction venture, regardless of
the execution technology
employed.
It must be clear that this philosophy intends to
optimize
the
production
management
process, above all. That does not
mean
that it will
ignore technological developments underlying
the
improvement of the actual process.
Our personal experience has shown that we can attain
an
average reduction of the
expected construction time between 20 and 30 percent of the initial estimate, and a
reduction of the
pro
duction cost between 5 and 12 percent of the total amount
in
totally
different projects, like the construction of
McDonald’s
stores or churches (a 40 percent
reduction of the foreseen global construction time, using the same production team), the
execut
ion of horizontal residential condos (a 20 percent reduction of the production
time), commercial buildings, and shopping malls. We are certain that lean construction
in Brazil is feasible because of the recognition of the importance of planning
and
design functions by the companies in the sector. This has
consol
idated new types of
sustained competitiveness in
environments
marked by uncertainty and risk.
REFERENCES
1. Ballard G., and G. Howell. Implementing Lean Construction: Stabilizing Work Flow.
International Group for Lean Construction Meeting Procedures. Chile, 1994.
2. Goldratt, E. Critical Chain. North River Press, 1997.
3. Koskela, L. Application of the New Production
Philosophy
to Construction.
Technical
Report #72. Center for Integrated Facility
Engineering, Department
of
Civil
Engineering.
Stanford University, 1992.
4. Ohno, T. Toyota Production
System
. Productivity Press, 1988.
5. Womack, J., and D. Jones. Lean Thinking—Banish Waste and Create Wealth in
Your Corporation. New
York:
Simon
and Schuster, 1996.
