LBMS before the Start of Construction

Because optimized production rates will be used in the buy-out of subcontractors, it is extremely difficult and costly to change the total duration after subcontractors have been bought out.  Therefore, schedule optimization and compression should happen before buy-out.  It is also extremely important that work is planned to be continuous because cascading delays during production happen because of discontinuous work - starts and stops of subcontractors.  Any discontinuity increases the probability that the schedule cannot be implemented as planned - subcontractors tend to balance their own work if the GC or CM cannot show that the plan is already balanced.

Owners should be interested in seeing the following deliverables for each major design release:

• Resource-loaded schedules with continuous flow
• Explanations for any discontinuous work patterns
• Clear identification of bottleneck trades by General Contractor or CM
• Detailed explanation why those bottlenecks cannot be accelerated
• Risk analysis of the schedule showing the risky areas and the probability of completing each major milestone
• Cash flow

LBMS during Construction

During construction, the schedule should function as an early warning system.  By tracking completed locations, the total quantity of work-in-place can be easily calculated.  Daily reporting of total manpower on site makes it possible to calculate actual productivity and actual production rates.  These rates can be used to forecast progress and identify problems much earlier than in CPM systems.  For a more complete explanation of production control exercises performed on-site, please refer to the BIM 401 Webinar.

Weekly or monthly Owner reports should include the following items:

• Comparison of completed tasks and locations to plan
• Explanations for not implementing continuous work as planned
• Schedule forecast before control actions (if everything continues on the same production rate)
• Schedule forecast after control actions and documentation of those control actions (for example, adding 5 carpenters to the drywall crew, working on Saturdays, 2-shift work for selected tasks etc)
• Problem tasks where production rate is < 80% of planned
• Percentage of production completed compared to the plan (for the whole project and each major construction phase, such as foundations)
• Resource forecast for the next month for each main subcontractor to reveal unrealistic resource assumptions
• Updated cash flow forecast based on schedule forecast

Parallel LBMS and CPM

Trying to run parallel schedules in a project does not work too well.  CPM schedules have traditionally worked only as Owner reporting tools and have had limited relevance in field.  The information content of a CPM schedule is extremely limited, typically lacking quantities and resources and showing only durations without revealing the manpower and quantity assumptions behind the duration estimates.  Critical path and float are faulty as concepts because they do not consider the requirement to have continuous, balanced flow of resources through the project. Only continuous workflow makes it possible to forecast production and give early warnings of problems.

Schedule optimization using location-based planning typically results in 10% compression without the need for additional resources or increase in risk levels. This compression can be achieved so that most of the tasks in the project have continuous flow. However, all optimization is theoretical unless the schedule is actually followed on-site.  Having parallel schedules undermines this goal unless the schedules match exactly.  Location-based scheduling typically results in better quality, optimized, resource-based schedules in 70% of the time required to build a CPM schedule.  Optimizing a schedule from scratch takes about 50% of the time of converting a CPM schedule to LBMS and using that as the starting point of optimization.  All of these facts speak to abandoning the practice of running parallel schedules.  Additionally, CPM is included in LBMS, but only about 20% of LBMS information content is included in CPM (specifically logic layer 5: random CPM links).  Why would the Owner want to pay for two schedules and often for two teams of parallel schedulers?

If you are an Owner interested in exploring these ideas, please contact us.  We are more than happy to help you!

Case study 1: Kamppi Center, Finland

Kamppi Center building complex is located in the center of Helsinki.  The total budget was €500M.  The center consists of a central bus station, a metro terminal, an internal parking area, a six-floor shopping center, and three combined office and residential towers. The total duration of the project was planned to be four years.

LBMS was adopted as the sole schedule planning and control system. The implementation started after earthworks phase and lasted until the hand-off of the building.  The goal of LBMS use was to optimize the schedule, to compress duration, and to control production so that the optimized duration could be met.  Two consultants from my team tracked production and analyzed information and generated reports and recommendations for production management.  We were also present in subcontractor meetings showing production status.  The LBMS room came to function as the nerve center of the project because the subcontractors, superintendents, and project managers knew that we had the best information.

Our goal was to compress the original four year schedule by six months. Through proactive production control, the GC was able to achieve this and saved millions in overhead costs.  The project was completed in June 2005.

Case study 2: Opus Business Park, Finland

Opus Business Park was a case study in my PhD research.  It was a simple office building of 14,528 m2 in Helsinki.  The contract duration was from May 2004 to February 2006.  However, the General Contractor wanted to hand the building over two months earlier than planned.

LBMS was again the only schedule planning and control system that was used in the project.  The implementation started from earthworks and continued until the hand-off.  The project engineer used the software, and I helped by having weekly meetings on site where we analyzed progress and planned control actions.  The project engineer communicated the required actions to the project manager and superintendents and to subcontractors in weekly meetings.  Initial schedule optimization and risk analysis revealed that it was possible to compress the duration by two months without adding too many resources.  Location-based milestones were given in RFPs to subcontractors and the project finished according to the compressed schedule.

Company -level implementations

In addition to individual case studies, five large GCs in Finland have implemented LBMS to be their only allowed planning and control system. All GCs have reported good results - better overall quality of schedules, consistent duration compression, and better project data for use in company -level systems, such as procurement systems.  An example implementation case study can be found in (Soini, Leskelä, Seppänen 2004).

Why are all of the successful case studies in Finland?

The Finnish construction industry has practiced location-based techniques since the 1980s, starting with manual techniques (pen and paper), evolving to simple flowline drawing software in the 1990s, and finally to Vico Control (previously DynaProject) in 2002.  Already in the 1980s, location-based management was shown to compress schedules and result in better control of production.  Gantt charts resurfaced in the 1990s because of better availability of automated software for CPM applications.  Because of this history, it was easy to convince people to throw everything else aside and do the project completely with LBMS from start to finish.

Although the earliest known implementation of LBMS in construction is in the US (the Empire State Building), US contractors do not have the same familiarity with LBMS.  The prevalent scheduling method is CPM, and production control is practiced informally in the field by experienced superintendents, who each have their own systems of tracking data.  In 2007 and 2008, the GCs have started to experiment with production control but most of the projects are only now entering construction.  Because typical large scale construction projects take years, we do not have case studies which have been carried out from beginning to end using LBMS.  Many projects are now getting close to production, but it will be years before the first case studies will be completed.

Similar results to those in Finland - 10-20% duration compression and productivity improvement - can be achieved in US projects.  The most innovative contractors are already doing this.  Definitive bottom-line end results will not be known until the first projects are finished, but all the contractors already appreciate the better quality of information for decision making.  It is best to start now - we have the experience required to make it work for you.

References:

Soini, M., Leskelä, I. & Seppänen, O. (2004) Implementation of Line-of-Balance based scheduling and project control system in a large construction company. Proceedings of the 12th Annual Conference of the International Group for Lean Construction, Helsingør, Denmark

Big Owners are not satisfied with the ROI their GCs are getting from BIM.  They feel they have already paid for coordinated drawings and model, so clash detection ROI is not enough. Instead, Owners are looking for true ROI with schedule compression. If the model has been coordinated, and all elements can be quantified, why can't buildings be built faster? Using model-based quantities and optimal productivity rates to drive the Location-Based Management System is the solution.

Model-based quantities

GCs in the US are increasingly using 3D models to perform quantity takeoffs. To enable schedule compression by the use of LBMS, the quantity takeoff needs to produce the following information:

~ Quantities for each task

~ Quantities for each location

Quantities for each task

Each line item in the schedule should have at least one associated quantity item. A task in LBMS is defined as "work which can be completely finished by one crew in a location before moving to the next location."  Therefore, tasks should only contain scope which can be performed by one subcontractor. Any task which hands work off to another subcontractor should be included in the schedule.

Because this means that, for example, formwork and rebar need to be quantified, either a very detailed model is required or specialized quantity takeoff software, such as Vico Office Takeoff Manager, needs to be used.

Quantities for each location

LBMS requires quantities by location. Location Breakdown Structure is a critical planning decision. We have seen projects where a change in project's Location Breakdown Structure resulted in schedule compression of 10% with no increase in resource requirements or risk. To achieve this, Location Breakdown Structure needs to be an automatic, real-time feature of the construction management system, allowing different scenarios to be tried and quantities to be recalculated on click of a button.  Vico Office has a dedicated LBS Manager to achieve these functions.

Optimal productivity rates

The US construction industry typically schedules using historical durations which are not based on quantities, productivity rates, or resource assumptions. Therefore, the goals and total durations of projects are defined based on historical performance. Production rate requirements or resource needs are not known because they are thought to be the subcontractor's problem. This is, in my opinion, the main reason why the ROI of BIM has not been consistently achieved so far.

Historical productivity databases, such as RS Means, do exist. They have similar problems. Productivity has been measured based on average productivity in projects which did not use methods to improve productivity. Because the resulting standards include a lot of waste, using these productivity rates will bloat durations or result in excessive resource needs.

Instead of using historical productivity rates, optimal productivity rates should be used for planning. The goal is to provide optimal circumstances for the subcontractor to do work. When the subcontractor starts, the location should be clean, all RFIs resolved, as-built data of previous trades available, model and drawings should be coordinated, materials available and all prerequisite work completed. In addition, there should be continuous flow from one location to the next to remove starts and stops and to maximize learning benefits. If all this has been done before the subcontractor starts, workers will be able to achieve optimal productivity. They are only constrained by their skill level and production rates of predecessor tasks. This optimal productivity has been tracked separately from average, historical productivity in Finland and has been shown to be 10-40% higher, depending on trade.

The ROI of BIM can be achieved by creating optimal conditions for subcontractors, quantifying those optimal conditions, and adjusting subcontract agreements to reward the GC for providing optimal conditions.  The subcontractors need to produce according to agreed-on production rates when optimal conditions are available and to completely finish one location before moving to the next location.  Together with schedule optimization and prevention of cascading delays, these productivity gains should allow reliable schedule compression of at least 20%, concentrating on finishes and MEP phase. Location-based Management System, together with software tools such as Vico Office, makes this schedule compression possible.

There is a common misconception that a prerequisite of location-based management is using IPD or some other collaborative contract form. Although these new contract forms provide better transparency and a better way to share the savings, over 90% of projects using LBMS (Location-based management system) in Europe and in the US have the traditional hard bid as the contract form. However, there are great differences in how LBMS is used in different contract forms and where the ROI is going to come from.  This article discusses the use of LBMS in projects where subcontractors are working on a hard bid basis.

Master Schedule

The goal of the master schedule is to define "need times" for procurement, the Location Breakdown Structure for the project, overall strategy of construction (multiple crews/single crews), milestones for construction phases, total project duration, and production rate requirements for all the main subcontractors.  Because the master schedule is typically planned before subcontractors are selected, productivity rates may be unavailable for subcontracted work unless the GC has been using LBMS in previous projects and collected these productivity rates.

The master schedule should be planned by using model-based quantities and productivity rates from RS Means or company's own database.  The actual number of resources is relevant only for checking the feasibility of the production rate. It will change when the subcontractor is selected. The critical thing is to establish a production rate requirement (for example, sf/week) which is used in the Request for Proposal.

Procurement Schedule

In location-based scheduling, procurement events (such as design needed, RFP, bid evaluation, contract) are pulled based on the production schedule. Production gets the first priority and latest possible dates are used for procurement to minimize changes to bid documentation.

Bid Evaluation

Bids are evaluated based on location-based quantities. Quantities can be sent to subcontractors as indicative quantities and subcontractors should be asked to provide unit rates. This helps the GC to validate the total price and find out the reasons for differences in unit rates. Contracts can still be done as total price. Bids should include a production rate that the subcontractor is able to achieve in good conditions (units / day or / week assuming that the drawings are available, the location is clean, materials are available, RFIs answered, etc.).

Detailed Planning

After the subcontractor has been selected, his/her work should be planned in detail using the productivity and resource information. Subcontractors should be required to give this information. If they do not, the GC will be able to calculate it anyway during production. The result of integrating the subcontractor's information with the schedule is an accurate resource graph. The subcontractor should commit to the general resource loading although specific mobilization dates and crew sizes will be adjusted based on schedule forecast information and actual productivity. Typically a resource loaded schedule will show different resource loading than expected by the subcontractor. This is typical because the subcontractor does not know about requirements of other trades.

Production Control

Actual start and finish dates in each location are tracked. Subcontractors should be required to provide this information weekly (daily for tasks with production problems). Actual resources should be reported daily but they do not need to be allocated to locations or tasks because that can be automated based on ongoing tasks and locations on that date. By combining actual production with resource number, actual labor consumption (manhours / unit) will be calculated and can be used for forecasting future progress. If a subcontractor is not achieving their production rate targets, it is a simple calculation to determine how much productivity needs to be improved or how many more workers need to be added to achieve the required production rate.

Alarms are used to prevent subcontractors from causing problems for the other trades.  When an alarm happens, the GC needs to take immediate action to prevent cascading delays. In typical projects, cascading delays increase project duration by 10% (see my earlier article about cascading delays).

Claims Support

Because the location-based management system contains planned quantities, production rates, resources and productivity and actual values for the same variables, it is straightforward to negotiate any claims. All the required information is contained in the BIM model or LBMS, supplemented by meeting memos, RFIs and other information. Typically the objective data contained in LBMS can be used to prevent litigation because facts are not subject to interpretation.

General Contractors in the US are increasingly using Control as a tool to monitor production. The implementation has typically followed the following steps:

• Analysis of a CPM schedule - discovery of huge inefficiencies caused by starts and stops and resource conflicts
• Optimization of baseline using location-based methods - continuous workflow, quantities, resources, often cash flow as well
• Monitoring production on site using control charts - comparisons to original baseline

There is a big, required additional step to achieve the real benefits of location-based management. Location-based controlling (as opposed to monitoring) needs to be implemented to maximize productivity and prevent cascading delays. The rich information content in a location-based plan needs to be used in management decision making.

There are two types of control systems: push and pull control systems. Push control means starting work according to predetermined plan, without considering the current state of the production system. This typically manifests as contract brokering and forcing the subcontractors to start according to their contract, or to increase their production rate just because they are delayed without considering the real value of that production rate increase. It can be said that most construction companies today use push controlling as their control methodology.

Lean Construction is based on pull controlling. A location-based version of pull controlling is to use the forecasts to evaluate location availability and only start new crews when production data shows that there are free locations for them. This will increase productivity of individual operations and prevent cascading delay chains from happening (refer to my earlier post about location conflict and productivity loss and cascading delay chains). However, it requires much more data about production than available in activity-based systems.

Location-based management (and Control software) implements pull controlling by recognizing that there are four levels of information: baseline plan, current plan, progress and forecast.

Baseline

Baseline tasks are the original strategy for building the project and form the basis of commitment between the GC and the Owner. However, they are based on assumptions of productivity, original assumptions of means and methods and often missing subcontractor input. Baseline tasks can only be updated based on Owner-approved changes and delays. Their main use from production control perspective is to establish production rate requirements and milestones for most critical tasks. This production rate information can be used in subcontractor negotiations to ensure that selecting the cheapest subcontractor does not result in buying a process bottleneck. Required production rate (units / week or / day) can be specified in bid documents.

All users of Control start their journey by analyzing and optimizing their baseline.

Current

Current level of information is used to plan the current best understanding about how to build the project. Current information is based on up-to-date quantities, subcontractor productivity and resource information and results from logic workshops. Current plans should be planned together with the team building a phase of the project (for example Foundations or Superstructure). Current plans reflect commitments between General Contractor and subcontractors and should be updated when the basis of that commitment changes (for example, a better way to do work is found, error in quantities is discovered etc.)

The current level of information is not generally used at the moment even though it is one of the most important parts of location-based management system. Instead, the GCs update progress information (see below) and baseline tasks.

Progress

Progress information records actual progress on site and compares it to Baseline and Current plans. Location-based progress information includes start and finish dates of tasks and locations (or % completed in a location) and actual resources of each subcontractor. This information is used to calculate actual production rate (units / day) and actual productivity (manhours / unit). Progress information can be shown in Flowline figures or as a Control chart.

Although progress information is useful in explaining historical production problems, using it for control is essentially like trying to control the project by looking at rear-view mirror. Most companies are trying to use progress information, together with baseline, to control production. This is dangerous as it typically leads to push controlling (red colors in control chart initiate action to catch up by increasing resources).

Forecast

Forecast is the most important part of Location based management because it makes possible pull controlling and associated productivity benefits. Forecast is calculated based on combination of progress and current plans and can be edited to enable look-ahead planning. Tasks, resources, logic and quantities are taken from current information (or baseline if current is not available) but productivity and production rates are taken from progress information. Critically, forecasts start where progress information ends. (see figure below) The initial forecast shows what will happen if the production continues in the same way as previously. Pull controlling is using the forecast information (= the situation on site) to plan control actions. Increase in resources, or suspending a task is shown by adjusting the forecast - not the original plan. Alarms are generated if location conflicts occur.

Although forecasts can be used to successfully prevent cascading delays, there is very little evidence that this critical part of location-based of management is being implemented.

It is time to move from push controlling to pull controlling.

Location-based management is being implemented in many US projects, especially in hospital construction. Because the management system is new in the market, many contractors have hard time figuring out what is the best way to get started. By starting, I do not mean training of Control software, because that is the easiest part of implementation. Getting the real value out of location-based management requires a lean approach to production - a decision to fight waste in all its forms and increase productivity of all stakeholders in the project. Successful implementation requires certain transparency and open discussion about production problems. Subcontractors need to share information about their production rates and productivity, General Contractor needs to accept the coordination role of subcontractors, instead of working just as a contract broker with a hands-off approach. General Contractor should not hide or accept production problems but take ownership and push for early resolution. Because of this transparency, location-based management works best with new contract forms, such as IPD. However, almost all projects completed with LBMS have used traditional contract forms and still received a lot of benefits from greater transparency.

Because implementing location-based management means that the management philosophy of the project will change, all participants in the project should understand it and be able to contribute to its success. Especially drywall and MEP subcontractor buy-in is critical. To gain acceptance, the best way in my experience had been to organize a series of workshops where subcontractors and other team members learn what is LBMS, and are able to contribute to critical decisions about the production system.

The goal of the first workshop is to demonstrate the planning and control functions of LBMS and discuss how each party will be affected by implementation: what are the information requirements and benefits for each subcontractor.

The goal of the second workshop is to decide the location breakdown structure of the project which works for most subcontractors. Typically, each project has at least three different location breakdown structures: structure, exterior and interior. Structural and exterior breakdowns typically affect fewer subcontractors, so most of the time is typically spent planning the interior location breakdown. The best way to run the workshop is to use drawings and have each subcontractor describe how they see the project – which areas they can complete before moving to the next area. There are always some tasks which do not follow other breakdowns (such as wiring of electrical systems) but it is typically possible to find a breakdown which is a good fit for most subcontractors. The result of the workshop is a marked-up set of drawings which will be distributed to participants to form a common understanding of locations used in the project.

Next workshops concentrate on a given construction phase or locations – such as Foundations or Rough-in and finishes in corridors. These are logic workshops where the task list of the project is finalized and dependencies between tasks are decided. Subcontractors typically say that GC’s try to enforce sub-optimal logic on them based on their previous template projects. When I have ran logic workshops, I typically see that different teams of subcontractors and General Contractor will decide on different sequences. Often innovations emerge: things can be done in a different sequence if design is changed. The goal is to find and commit to an optimal sequence of tasks in each main location type (for example, corridors, lobbies, office rooms, operation rooms etc.).

Once list of tasks and locations is known and quantities are available from the 3D model, it is then easy to build the schedule. Subcontractor productivity rates (manhours / unit) and resource constraints are needed at this point. My approach has been that the General Contractor puts together the first draft which meets all the contractual milestone dates and then the schedule is iterated and optimized in workshops by using Flowline visualization and resource graphs as communication tools. Typically resources required to achieve the deadline based on productivity rates are different than what subcontractors anticipated. This will often lead to more innovation – how to increase productivity or to change sequence so that resource constraints are achieved.

Planning should not be left to the scheduler or superintendent alone. To get maximized benefits, the knowledge of the whole team should be utilized.

In my PhD research I found a lot of problems with the current practice of running subcontractor meetings. Much of the time was spent describing what had happened in the past with other subcontractors dozing off when one was delivering his narrative. Narratives concentrated on what the subcontractor had been doing and typically tended to ignore production problems on site before they became serious. There was no attempt to tie the individual narratives together into a coherent story. Discussions about the future were similarly problematic. Each subcontractor described what they were going to do but there were no concrete goals or commitments.

The General Contractor practiced push control by making sure that the scheduled dates were achieved. If there were delays from the schedule, the standard response was to make the subcontractor increase the crew size or start new tasks in new locations. Most of the decisions were made in isolation considering one subcontractor at the time. The complex interrelated nature of construction was ignored.  This typically led to overcrowding, multiple subcontractors in the same location and consequent loss of productivity and cascading delay chains.

There have been attempts in Lean Construction to solve these problems before. The key idea in Lean Construction is to discuss and remove constraints of production to ensure that commitments can be achieved. Using location-based management system to run subcontractor meetings can provide further improvement. Location-based subcontractor meeting process includes the following components:

• Previous week's status is sent to participants before the meeting in the control chart format. Everyone comes prepared to the meeting and knows the status. Discussion about the past focuses on delays and production problems.
• Numerical production data calculated using quantities, production rates and actual crew size on site drives the discussion about problems
• Future is discussed using the flowline forecast to identify free locations and risk of location congestion
• Forecasts are changed according to the decisions and they form the new look-ahead plan
• Resources and crew sizes are discussed using the resource forecast. Forecast needs to be adjusted to take into account any resource constraints.

By including the resource discussion in every subcontractor meeting and concentrating on preventing future production problems, cascading delay chains can be stopped

One year of data collection and three-and-a-half years of analyzing and summarizing the results during weekends, evenings, and holidays has finally paid off: a full manuscript of my PhD thesis is ready! It will be publicly defended in Helsinki, Finland in June 2009 and will be made available online after that.

When I started the work on the PhD, I had no idea about just how broken construction projects really are. I thought that construction project problems are primarily those of planning - if you plan a better plan based on locations, quantities and productivity rates, things should start to work better. My plan was to introduce three projects to location-based planning, optimize the plan with the project teams and then record progress against the optimized plan and try to find out why actual progress deviates from plans. Based on earlier research results and experience, I expected to see some deviations based and then to see control actions to restore the planned schedule.

I spent a year collecting data from three construction projects every week. Although there were some deviations from plans which were discussed in subcontractor meetings in each project, there was no sense of great urgency or big problems. In reality, based on data analysis, each project had hundreds of production problems (start-up delays, discontinuities or slowdowns) which were never discussed in subcontractor meetings!

It seemed that management was unaware that these problems were happening. It also turned out that many of these problems had actually generated an alarm in Vico's Control software but the alarm had been ignored with the hope that production somehow automatically will be restored to planned production rate.  Delays suffered by one subcontractor in one area rippled down to other subcontractors in other areas because of shared resources, shared locations, and unanalyzed decisions. Each project had a clearly identifiable chain of cascading delays which started in the beginning of interior finishes and MEP phase and continued to end of commissioning. The total effect of these delays was 10% increase in project duration in each project and huge productivity losses to subcontractors because of location congestion. The loss of productivity associated with labor flow issues has been estimated to be 20% (Thomas et al. 2003), not to mention the General Contractor management time required to continuously update the plans and communicate different plans to subcontractors.

Figure 1 below shows a typical cascading delay chain (red circles indicate production problems). In future blog articles, I will discuss the root causes of these delays and how to prevent them from happening. The main conclusions of my thesis are currently being tested in construction projects both in Finland and in the US. I will keep you posted on the results.

References:

Thomas, H.R., Horman, M.J., Minchin, R.E. Jr. & Dong, C. (2003). Improving Labor Flow Reliability for Better Productivity as Lean Construction Principle. Journal of Construction Engineering and Management, Vol. 129, No. 3, pp 251-261.

February 21, 2008

The basic assumption of location-based planning is that working continuously results in higher productivity. Multiple crews working in the same location also result in decreased productivity and increased risk because resources might leave the site and return late.

In my PhD research I collected progress data from three projects weekly. All the projects showed the same result. If multiple crews are working in the same location, work either slows down or stops altogether. Slowdowns and stoppages happened even if the tasks did not have technical logic link with each other. Many tasks can be done in free sequence but they can not be done at the same time in the same location. For example floor covering work or plasterwork interferes with any other trade working in the same area.

Figure below shows a typical example from one of the case projects showing interference. Notice that the actual slopes of all of the tasks are very close to planned slope but only if the tasks can keep the complete location to themselves. Immediately when tasks happen at the same time in the same location, work either stops or slows down considerably. Average productivity loss is 20 to 50 % depending on the particular mix of tasks. In a project of 15 000 m2, I identified over 350 issues where productivity was lost because of these conflicts. It was also scary to notice from meeting memos that subcontractors had the same crew on site but achieved 50 % of production rate when there were conflicts. Think about the effects on subcontractor bottom line...

What are the implications for planning?

• CPM is not a valid planning technique because slowdowns and stoppages happening in locations even without technical precedence have larger effect than a delays on critical path (will be shown in my PhD)
• Buffers are necessary to keep trades far enough from each other
• Weekly location-based controlling is required to achieve productivity benefits and to prevent trades from clashing

For subcontractors, this will affect the bottom line directly. For General Contractors, project contractual milestones become easier to achieve because each task can be produced faster. Schedule risk of the project will decrease because slowdowns and stoppages are by nature unpredictable. Some of the time saving is required for buffers but at least 10 % duration reduction is possible without increases in risk level. Alternatively the same duration can be achieved with much lower risk.

November 26, 2007 

A couple of weeks ago, I was visiting California to present and discuss location-based planning at Stanford University and some companies. In almost all discussions I had, the question about critical path, float and buffers came up. Although I argued that float and criticality do not have much relevance as tools for production control, I will address them in this article because it is a critical issue for contractors working in a CPM-based contractual climate.

Location-based planning vs. CPM

Location-based scheduling is a special form of CPM which automatically generates and updates logic based on locations. Therefore, any CPM schedule can be translated to a location-based schedule (with one location), and any location-based schedule can be changed to a CPM schedule (with location names as the activity names). The logic is exactly the same.

The only difference between location-based planning and CPM in float calculations is the powerful continuity constraint. If the continuity constraint is on (and it should be for most trades), the start date of the task is pulled forward, so that all locations of the task happen continuously. This allows the planner to optimize the schedule by finding the optimal production rates for tasks spanning several locations. By doing so, the planner is effectively minimizing the float of the project, but planning work continuity. To preserve the critical path, the float of all locations must be made equal during the backward pass.

Figure 1, below, shows as early as possible schedule of two tasks going through five locations. The successor task is faster than the predecessor, so the task becomes discontinuous. In this case, the total float will be calculated as in normal CPM.

In contrast, the following example delays the start of Task 2 and forces continuity. In this case, the total float of all locations becomes the same (zero) and all locations are critical.

Buffers vs. lags

Basic CPM has just lags. Location-based scheduling has two similar concepts: the normal CPM lag and the buffer.

Lags are used only when it is technically mandatory to wait a number of days after finishing (starting) a predecessor and starting (finishing) a successor task. A good example of this is concrete curing, or a location lag of two floors between structure and finishes for safety reasons.

Buffers are a way of reducing the schedule risks of a project. Location-based schedules use quantities and productivity rates to define durations, which assume that crews can work with their optimal productivity. In the process of location-based schedule optimization, the total float of a project is decreased. Because of the combination of lesser total float and smaller durations, the total duration of a project is greatly compressed. However, this duration compression is impossible to achieve in practice - unless buffers are planned between the tasks.

Buffers are used to protect the continuous production of a successor task from the possible production deviations of its predecessor. Therefore, they should be owned by the General Contractor who is responsible for production. All the durations of a location-based schedule assume the best possible productivity for the crew in question. In effect, all materials should be available; all design complete and other crews should not be working in the same work area. Location-based planning aims to achieve these optimally productive conditions by isolating the crew using buffers. In the event that something goes wrong, the buffer is absorbed before the next trade suffers.

Visually, buffers and lags look the same in the Flowline. They both force an empty space horizontally between the tasks. Also, float calculations are exactly the same - lags and buffers both give the same critical activities. However, in progress forecasting, buffers are absorbed before the next trade's forecast is changed.

The figure below shows two tasks with a Finish- to- Start relationship in each location, with a buffer of 5 days. The Task 1 start-up delay of up to 5 days will be totally absorbed by the buffer with no consequences to Task 2.

The matter is different if there is any production rate deviation. Any size of buffer will run out if production is going too slowly; as shown in the figure below. The location of the red dot shows that if Task 1 continues at the same rate, Task 2 will run out of space in the second location. In this case, the buffer allows Management to detect the problem and react before the alarm becomes reality. 

Criticality and float as management tools

Location-based management puts the greatest emphasis on the management of resource flows, and isolating crews from each other using buffers. Any break in work continuity will result in additional costs for the associated subcontractor. Therefore, alarms showing up in a location-based schedule should be the main management tool, because if they realize both time and money will be lost. Buffers allow management to notice the deviation before it happens and give them time to react.

From a project perspective, production rate and continuity is more important for some tasks. These tasks can be defined by criticality and float. Remember that if any location of the task is critical, the production of all locations becomes critical. This makes controlling easier, because criticality is defined by the type of work - not the location where the work happens. Even if the critical path transfers from one task to another in a later location, it is critical to achieve the correct production rate well before that location. It is crucial to preserve the work continuity of all critical and near-critical tasks, because if a subcontractor has to leave the site because of a lack of work, the task will very likely be delayed by the return delay of the subcontractor. Let the critical subcontractor do his work enjoying optimal conditions!

Location-based management is inherently collaborative in nature. The General Contractor commits to providing optimal work conditions, and the subcontractors commit to providing the required production rate. Subcontractors benefit from higher productivity and cost savings, and the GC benefits from a higher probability of achieving the schedule. Ask your subcontractors - you will find that everyone wants to work this way!

I will write more on the topic of criticality, and how location-based management should be used for delay analysis and in the support of site meetings in later articles. Next time, I will return to the MEP (Mechanical, Electrical, and Plumbing) theme and describe the location-based planning process which includes MEP contractors. MEP is somewhat of a special case in location-based management, because their resources can work in multiple tasks. When one task runs out of work, there is usually something else that can be done to balance the resources. This will be addressed in later articles.