International Group for Lean Construction
11th Annual Conference
Meeting at Virginia Tech
Blacksburg, Virginia
July 22 thru 24, 2003
IGLC-11, Day One
First of three days of 57 presentations of accepted papers.
Theory
08:23 7/22/03
Linguistic Action Foundation of Lean Construction, Hal Macomber
Premise: The explanation of the Last Planner System™ effectiveness can be enhanced by applying the linguistic action perspective.
Well-conceived projects conducted on the Last Planner System™ (LPS) were not always successful. Project leadership appeared to be a central issue. Hal and Greg Howell began looking for an explanation beyond the usual one found in the lean construction literature.
They started with Fernando Flores' description of management:
Management is that process of openness, listening, and eliciting commitments, which includes a concern for the articulation and activation of the network of commitments, primarily produced through promises and requests, allowing for the autonomy of the productive units.
The authors began to reconstruct the LPS as a particular kind of linguistic system for providing management.
The linguistic action perspective and the LPS addressed five key concerns for all projects:
- Coordinating action among project participants
- Learning
- Geting help
- Innovation
- Project control
All five concerns require a foundation of trust.
The authors claim the new role of a project manager, particularly on constuction projects, is to turn strangers into friends and friends into team mates. That happens by producing trust. The engine of trust is a well-functioning network of commitment.
09:09 7/22/03
Complexity Theory, Sven Bertelsen
Great lecture on the fundamentals of complex systems. The dynamics of large project chaos was presented as "The Battleground."
- Order is going out of fashion.
- Complexity is part of nature.
- Chaos is here to stay.
- Life is an unruly mess.
Paulo: There's a difference between chaos and complexity. Chaos emerges when we no longer understand what is going on. Chaos is established by people...they don't talk to each other. Without consensual autonomy of teams with live in chaos everyday. People talking is the answer.
Sven: I agree mainly with you. The term chaos is hard to define. Chaos emerges within a project. Yes, chaos is something we generate ourselves. Complexity is more prevalent than we recognize.
Audience: COmplexity is dealt with risk management on projects. COmplexity is not ignored. Within current complexity theory the system is trying to balance itself. How do you get control of a system going out of control?
Sven: I am not sure of my own thinking here. Too much order is not good. Self-organizing autonomous people is good. The farther away from chaos you get the worse the system will perform.
09:30 7/22/03
Achieving Change in Construction, Lauri Koskela
Premise: Current change initiatives are not producing desired results in the industry.
Design-build, partnering, and IT solutions, are all falling short of expectations. Only marginal benefits.
Paradigmatic change is required rejecting the transformation model as the basis of construction project management. Change to post-modern approach.
Start the process up-stream with the owner. Alternatively, start with operational design, prefabrication, and site production processes.
We might consider a small wins (Weick) strategy. The usual LPS implementation could be considered a small wins approach to introducing change.
Policy advice:
Can be nurtured
Need new understanding -- basic research on implementation
Discussion:
How do you see the role of universities as agents of change? Many professors are agents of change. While it's possible it is not occurring in all universities. Unions offer hope.
Academics are not rewarded for interacting with the real world.
09:51 7/22/03
Six-Sigma in Lean Construction, Tariq Abdelhamid
Variability is the universal enemy.
Schonberger (1986)
Six-Sigma is guided by DMAIC (define, measure, analyze, improve, and control) and DFSS (design for six sigma).
Comments:
Good introduction to six sigma and the opportunity for construction projects.
10:11 7/22/03
Measuring Lean Conformance, Lou Troendle
Premise: Looking at field work to develop an approach for assessing and benchmarking.
CII initiated effort.
Five principles:
- Standardization
- Culture/People
- Continuous Improvement-Quality of Product
- Customer Focus-Understanding the Needs
- Eliminate Waste-Identify and Measure
Developed a lean wheel for displaying assessments based on a proposed set of questionnaires.
Discussion:
Are there plans to evaluate the data collected? Good question. We haven't approached that. Looking at coming up with metrics.
Are you connecting this to percent complete? We are looking at other issues in the questionnaire. Putting an emphasis on visual management.
Did the people filling out the questionnaire have instruction in lean? No. The questions are designed so they can be answered
Have you looked at projects claiming to be lean? Yes, one. Don't know the specifics.
Are you looking at outcomes? This is really process maturity. What is the fundamental methodology? We are trying to determine the state of lean in the industry. Are people who are claiming to be lean really lean. And, what is lean. Finally, what are the outcomes of the project? How are those outcomes related to lean practices?
Comments:
Taking direction from lean manufacturing. Not spending enough time with the people who are lean.
General Questions on Theory Track
Audience: Have you considered the effect of business models on the rate of change in the construction industry?
Lauri: Yes. Different models...(sorry, didn't get his response)
Audience: Are we getting chaos because we have bad plans?
Sven: The nature of the project must be considered before assessing good plan or bad plan. We have chaos even when we have good plans.
Production Control
11:06 7/22/03
Production Planning Support Systems, Luis Alarcon
Premise: New support system is required to assure implementation success.
Barriers to implementation
- Time
- Training
- Organization
- Lack of self-criticism
Continuous improvement in the project setting is rarely implemented.
The support system is designed as a collaborative effort for designing a project-specific approach for fully implementing LPS. Call this Plan Control.
Did a number of cases showing positive results: improving PPC, cost savings, zero accidents, and timely completion. Other cases showed similar results.
Plan Control is an approach that emphasizes integration. Includes: integration tools, learning, communication and transparency, and flexibility.
Attribute implementation success to the collaborative approach in establishing an implementation support system.
11:18 7/22/03
Update on The Last Planner System™, Glenn Ballard
Premise: The LPS has changed dramatically in 11 years. Time for reflection with an emphasis on the relationship between scheduling and production control.
Neglected the goal-setting aspects of good scheduling. In the traditional contract mentality goal-setting occurs independent of other scope of the project. The WBS is the usual approach for doing the segmentation of goals. We must provide a focal point for control in the hand-off between specialists.
Tool for designing schedules based on phase or pull schedules produced in reverse. Attention is given to the hand-offs. Production of the phase plans occur as the phase of the project approaches. This has the effect of reducing uncertainty based on the delay in doing the detail scheduling.
The process of reverse phase scheduling is the essence of planning. It is a conversation that takes place among the principal performers of the project.
Innovations and Issues
- Make customer acceptance the measure of task completion
- Promote learning through sharing success stories on rate of improvement in PPC
- Incorporate reliable promising
- Measure PPC against planned day vs. week for hand off
Questions:
- Daily meetings?
- Daily commitments?
- Multiple levels of commitment?
- How to avoid the misuse of LPS? It is not a tool for micro-management.
- Can we measure reduction in duration variation resulting from LPS?
11:36 7/22/03
Line-of-Balance Based Schedule Planning and Control, Jouko Kankainen
Premise: Line-of-Balance (LoB) can be the control tool for achieving the desired schedule.
Comments:
Frankly, I didn't understand the presentation. While the claim is the project can be controlled using LoB, it appears to be a better re-planning tool. The paper was also a pitch for DYNAProject™, a software tool for using the LoB approach on construction projects. I wonder if a LoB approach is better used in the goal-setting.
11:50 7/22/03
Comparing Process Improvement Based on PPC, Tariq Abdelhamid
Premise: Labor utilization will improve with increases in PPC.
Believe those who are seeking the truth; doubt those that find it.
Andre Gide
Did an empirical analysis of a manufactured home company in Texas. Stepped through the production steps and the process. In this case it is a push system. Used PPC to measure the quality of the assignment process.
Discussion of Production Control
Audience: If you apply the LUF at the system level rather than the operations does it give desired results.
Glenn: Must put your attention on balance.
Tariq: It is an issue of sub optimization based on the metrics.
Audience: Did you find a relationship between 1st and last station LUF.
Tariq: No. But we don't have enough data to say that with confidence. What we saw makes sense. There was no coordination of hand-offs. There was no space for buffers between stations.
Audience: There seems to be cross directions in this movement at this moment. Is 75% PPC good enough and why?
Glenn: No. We should be aspiring to 100% PPC one week in advance. We are not taking reasons analysis and improvement seriously.
Tariq: The PPC was 78% for all the stations. When we first measured PPC we got 100%, but it was hiding the rework that was occurring and how the measurement was calculated.
Luis: I agree with the previous opinions. Sometimes the way you measure PPC doesn't put enough independence on the activity. We tend to shift companies who are doing well on a weekly PPC measure we encourage them to go to daily.
Jouko: Haven't tried the LPS on our project. The MPS must be implemented as planned.
Audience: After using LPS we got to 100% PPC in a few weeks. If you don't get to 100% the project is managing you. What do you think?
Glenn: As you approach 100% the next challenge is to extend your reliability horizon. If you can extend it and increase you reliability.
Audience: Sometimes you get to 100% by using resources you didn't plan or want to use.
Audience: On LoB are you generating random numbers based on construction experience?
: We are using the CPM to inform the approach.
Audience: What do you use for phasing? Did you notice that the people working in the hidden factory were they the experienced people? Have you seen people incenting 100% PPC?
Glenn: There are a variety of ways for determining the phases: building systems (Uniformat II). This groups specialists in 'natural' teams. Client milestones often determine phasing.
Tariq: We did notice they pulled out the experience workers for doing the rework. They would bring in swing workers as replacements. While less experienced they could work in different stations. Or, they would work the experienced people overtime. On the last question, we only used PPC as a diagnostic. However, once people finished their 10 units for the day they went home.
Luis: I'm seeing the use of free time as an incentive. I've only seen one case of using PPC for basis of incentive. PPC increased from 50-80%.
Jouko: Keeping the line in balance is the objective. So, it is desired for people to go home rather than advancing work.
Product Development
13:36 7/22/03
Client Requirements Management, Carlos Formosa
Premise: The success of construction projects depends more and more on the management of customer requirements.
Did representative case studies for one residential builder. Surveyed customers and visited sites. Examined the current process and the decision-making process.
Guidelines for Managing Client Requirements
- ID the main clients
- ID client requirements
- ID the decision-makers
- Formalize client-supplier relationships
- Define and communicate project objectives
- Consider production system requirements
- Monitor the degree of design maturity
- Must be able to trace design solutions to client objectives
- Process coordination
- Multi-discipline teams
- Tools and practices for capturing and updating client requirements
- Use of IT systems
Discussion:
Glenn: Please comment on the use of "client requirements" as traditional architects' design brief.
Carlos: We fight the expression "client requirement".
Comments:
Force field analysis might be useful for investigating competing client requirements. A workflow
14:00 7/22/03
Dynamic States of Project Purpose, Michael Whelton
Premise: See a difference between needs and requirements and the implications for project design. Management behavior adjusts to the level of uncertainty and complexity of client needs. As uncertainty increases an emergent style of managing the process is called for.
Project definition is a social adaptive process. Transitioning from needs to requirements involves facilitating dialogs.
Discussion:
Audience: What are your thoughts on measurements?
Michael: Measure the client's satisfaction with the process.
Hal:
Michael: Shift in style from an administrative approach to one that is more fluid and adaptive.
Comments:
Confused by a repeated use of the phrase 'management actions.' The use of BP workflow offers an opportunity for enhancing this process.
14:13 7/22/03
Enablers for Concurrent Engineering, John Kamara
Premise: Missing the opportunity of concurrent engineering (CE) in construction.
Integration and concurrency are key principles of CE. Includes the scheduling of activities and the interaction of team members and their tools.
Support requirements:
- Distribution
- Heterogeneity
- Autonomy
Challenges:
- Linkages between firm-level and project-level support
- Getting clients on board
- Achieving 'true' concurrency
Enablers:
- Integrated procurement strategies
- Process protocol
- BEACON: model for assessing readiness for CE
- Groupware
- Intranets and extranets
- PlanWeaver
Inconclusive results in the paper.
Discussion:
Lauri: How are the goals related to CE? Are there other tools related to project goals? I'll answer the 2nd question. Yes. I am concerned you only focussed on the advanced CE principles.
: Designing in parallel is bound to reduce costs just due to the advantage of shorter schedules. The 2nd question...CE is defined more by the principles. There are many tools that support project goals.
Comments:
The attention is on CE as an end in itself.
14:31 7/22/03
Design for Manufacture and Assembly, Christine Pasquire
Premise: A drive for competitiveness in the industry is showing the benefits of taking a DFMA approach.
Just because construction projects are engineered to order doesn't mean we can't adopt standards in the materials that are used throughout the project, e.g., the selection of standard fastening devices.
Barriers to improving design
- Risk is passed on
- Design roles are vague
- Supply chain involvement is restricted
- Excess of information
- Learning is prohibited by project focus
New Philosophy
Start with client business needs
Pre-assembly and standard components are not concepts that may be addressed 'at some point'.
14:47 7/22/03
Open Building Strategy for Converting Obsolete Office Buildings to Residential Uses, Steohen Kendall
Premise: Report on the state of the strategy.
Requires
New design methods for architects
Comments:
Fascinating description of a paradigm shift underway for the design and delivery of multi purpose buildings.
15:05 7/22/03
, Mike Kagioglou
Premise: What are the main barriers to implementing product development process modelling
Barriers found here are consistent with barriers throughout the industry:
- Integration
- Poor attention to human issues
- Problems in the process content
- Lack of leadership
- Barren organizational context
Wrap-Up
Glenn: Special client requirement management?
Carlos: This is much broader that the usual architects' brief. It goes beyond the users and owners. It covers other constituencies. Also, as solutions are developed client requirements change.
Glenn: Where do you see it is appropriate and useful to bring the production team into the front-end process?
Carlos: As early as possible. 'Project definition' is broader than client requirements.
Audience: DFMA in manufacturing has led to radical redesign of products. It seems that the flow of information needs to change for that to happen in construction. Please comment.
Christine: Some manufacturers are thinking about this. Need to understand how the owner will make money out of the building.
Audience: Open building is very exciting. There's a viable market to pull on you to find value. You say the stock of old buildings is huge. How many are adaptable for this?
Stephen: There are certain types of buildings that lend themselves to this. It has to do with depth from the facade, the unimpeded areas. Other than that, if we get smarter we can do this with residential buildings. You have more units per floor with residential that with office. Financing and later turnover is more complex. The building structure is not the main hindrance.
Audience: One of the advantages of fitting out residential buildings in advance is separating construction decisions from the sales process. It seems it would increase variability in the flow rate. Why not keep a stock of different floor plans?
Stephen: We worked with a developer who selected a mix of units. By the time the project finished they had a bunch of dog units. The approach allows you to make decisions when there is less uncertainty. Delivering buildings as an integrated whole is bad for everyone.
Audience: It seems the arguments for this have been around for a long time. What will create a tipping point?
Stephen: There's supply and demand. The Urban Land Institute points to a tremendous demand for flexible building solutions. Professionals are holding up the supply. Architects are the key impediment stuck in old paradigms. We need some demonstration projects in the U.S. It's happening elsewhere in the world.
Safety
15:50 7/22/03
Workers at the Edge, Takis Mitropoulos
Premise: Current thinking on safety is not keeping workers safe.
Rasmussen's theory has us to increase workers' ability to cope at the edge of safety. Requires situational awareness and effective error management. Learning from aviation investigating crew resource management.
Errors are unavoidable. Error countermeasures:
- Error avoidance
- Error trapping
- Error mitigation
Error management behaviors
- Cross checking
- Informing others of problematic situations
- Facilitating input from all members
- Preparation and contingency planning
Training in non-technical skills
- Briefing strategies
- Communications
- Assertiveness
- Decision-making
- Workload management
Conclusions
Construction work does not occur in a cockpit
16:03 7/22/03
Signal Detection Theory: Enabling WOrk Near the Edge, Tariq Abdelhamid
Premise:
COnstruction workers represent 6.4% of the US workforce. 23% of all occupational deaths are construction related. The conventional approach is to follow rules. That doesn't keep people from drinking and driving. New approach is to tell people how much they can drink to still be in control. Training is now available to help people understand when they are moving from zone 1 to zone 2. The hazard zone for one person is larger for one person that it is for others (just like the tolerance to alcohol).
Must look at the effectiveness of training on a person-by-person basis.
Signal detection theory is used in situations where two discrete states of the world cannot be easily discriminated. The world can be segmented into signal and noise. One can respond with yes or no to the signal and the noise. Looking for 'hits' and 'correct rejections' rather than missing the signal or accepting a false condition.
16:18 7/22/03
New Paradigm in Concrete Products Production, John Dean
Premise: Wet Cast industry has changed very little. THis is a look at ergonomics in the industry. Accident rates have been unacceptable. After changing practices they got a 75% reduction in accidents.
They automated tasks that put large strains on bodies.
16:30 7/22/03
Learning to Think and Detail from First Principles, Stephen Emmitt
Premise: Focusing on the detail, joint solution, allows people to reappraise their approach to detailing. Includes sustainability and disassembly.
Positive results:
- Attention to joints and minimizing waste
- Attention to process issues
- Solutions were lean
Negative issues:
- Some traditional solutions
- Health and safety not well-addressed
16:37 7/22/03
Managing Construction Process Risks, Martin
Premise: Looking at the ability of people to make the right decisions.
Postulates:
- The future is uncertain and unknowable
- We shape the future by our action
- We need continuous learning
Robust planning:
- Improve foresight
- Progressively reduce uncertainty/increase the level of detail
- Assure commitment
- Allow for uncontrollable events
The process of planning is a convergent process (like navigating a boat to a destination). We must produce alignment in the interests and intent of the parties. Alliance projects (JVs) support this where the traditional projects don't.
PPC is complex. People set stretch goals trying to get more out of the crews. PPC is negatively impacted.
LPS is not it. There is so much more to address. It is so much more complex than that. CII's lean assessment approach is superficial.
Comments:
Robust planning as presented misses organic evolving nature of projects.
Safety Wrap-Up
Audience: Can you apply what you've learned on safety to quality?
Takis: If we substitute failures for accidents, this would be a good area for investigation. Signal theory would apply to accepting assignments.
Sven: The real issue in Rasmussen's model is that there is an edge. The edge needs to be correctly identified. The edge doesn't happen gradually. It happens instantly.
Martin: People's perception of the edge is quite different
Audience: How do you make money if PPC is below 75%?
Audience: What is the role of safety management in an environment that is not rich in safety devices?
Audience: Health and safety problems are caused by high pressure, untrained workers,
Audience: The lead men like last planner.
Audience: Need to address the complexity and chaos situation. How can we take positive advantage of chaos?
Audience: I don't see that the designers are taking safety considerations into their design. Can they look at it in a different way?
Sven: Has it been easy to find the "first principles"? Which ones are you using?
