Are You Rolling Sixes and Ones or Consistent Threes?

The parade of trades simulation. Red team has normal six-sided die rolling numbers one through six representing variation. Blue team has modified die rolling only twos, threes, and fours representing reduced variation. Results after running simulation: Red team finishes in 26 weeks with 380 workers on site costing $21.7 million. Blue team finishes in 21 weeks with 280 workers on site costing $12.9 million. That’s an $8.8 million difference. Same project. Different approach. Red team tried rolling sixes to get out fast creating variation. Week looks like 6552211 representing attempt to move quickly at first followed by slowdown because of interruption eventually ending up stuck. Blue team maintained flow. Week looks like 4434322 representing steady consistent work without overwhelming rush from rolling six or painful crawl of one. When variation is eliminated, work flows from one end to the other without getting held up. The lesson: flow has very little variation. Movement doesn’t equal production. All this movement is actually waste. People start pushing and creating variation because it makes them feel good. Gives impression of progress. But requires twice as much material, 100 more people, and five more units of material inventory. Everything slows down because part of workforce is dedicated to managing, moving, inventorying, fixing, replacing, reordering, and organizing materials instead of installing work. Don’t roll sixes and ones when you can roll threes.

Here’s what most teams miss. They think pushing equals progress. Getting out of ground fast. Being aggressive with complex areas. Advancing schedule whenever possible. Bringing all materials now. Keep pushing everything for schedule. These people are trying to roll sixes in parade of trades which will later cause mess of ones. What if you told superintendent to slow down a little and keep steady flow and even pace? He might say you know absolutely nothing about construction and should get another job. And yet the data shows he will finish in 26 weeks with 380 people on site with material inventory of 10 like red team. This happens because people think movement equals production which is not the case. All this movement is actually waste. It gives impression of progress but requires twice as much material, 100 more people, and more inventory. The blue team rolling consistent threes and fours finishes five weeks earlier with 100 fewer people and $8.8 million less cost. Flow beats pushing. Data proves it.

The challenge is most superintendents were taught to push. Get work done fast. Start as early as possible. Bring materials now. Don’t slow down anything. Have workers working everywhere on site with no empty areas. But that’s exactly what destroys flow creating variation extending duration increasing costs. When superintendent won’t keep flow on schedule and hold dates, trades keep more people on site and more materials on site. When materials pile up, production slows down. Workforce dedicated to managing materials instead of installing work. CPM pushes as much work to start as early as possible putting too much work in process that must be managed, maintained, checked, monitored. When work in process increases, capacity of team and project resources reduced causing rework and eight wastes. But if you finish as you go, you have capacity to focus on work in process and complete it right first time. Don’t roll sixes and ones. Roll consistent threes.

Leveling Out Around Bottlenecks as Strategy

Some may be very concerned about our comment to slow down certain resources. But consider what happens when after we have optimized and sped up all bottlenecks, we continue to let the faster trades continue to go fast.

What happens? People are stacked in certain areas without flow, without geographical control, burying certain scopes, installing too early, which increases the amount of defects and use of resources such as the project management team’s time. This affects the trades that really need the help. There’s little merit to going faster than the general throughput.

The law of bottlenecks states that throughput time is primarily affected by the process that has the longest cycle time. In construction, the overall throughput of the phase is mostly affected by the process with the longest duration within the system.

Best practice is to first optimize the bottlenecks for the slower installations and then to even out the throughput of the remaining work. Therefore, the entire system. Because CPM does not allow us to see our bottlenecks, we cannot optimize them so the system ends up with a longer overall project duration, overproduced areas, fluctuating worker and material inventory levels, and a number of detrimental starts and stops.

The tricky thing with bottlenecks is that new ones will show up when you optimize the first ones or the largest ones so it’s a continual game of increasing flow by adjusting the throughput of the system. This is the key to achieving the shortest overall duration with the smallest crew sizes with the most minimal inventory levels and a visual system that identifies problems when they happen in a continuous flow that allows an evenness the team can use to focus their attention on the removal of roadblocks.

Finish as You Go

For Takt to work well, we have to finish as we go. Historically, superintendents have attempted to accelerate the project to gain time at the end to fix all the mistakes from going too fast and not installing it right the first time. With Takt, we cannot have and should not need to have that excessive buffer time for an excessive punch list.

With Takt, and while working to prevent roadblocks, the team, and especially the engineers, will be inspecting work, identifying punch list items, and working with trades to correct as they go. Meaning, finish as they go. Before a trade leaves the area or the project, they should not only install the production work, but they should finish the scope in its entirety.

This will take a mindset shift for most of our leaders in construction, but when this path is taken, we will see a better quality product delivered to the customer sooner.

Limiting Work in Process

This inherently ensures work in process is limited. CPM pushes as much work to the start as early as possible. That is a normal thought and concept, but it does something that is detrimental to our construction operations. It puts too much work in process.

Work in process that must be managed, maintained, checked, monitored, supported, and which utilizes precious resources. When work in process increases needlessly, the capacity of the team and the project resources are reduced, which causes rework and a number of the other eight wastes. It also causes unevenness and overburden.

The goal is to keep the project management team balanced, healthy, with workers going at a reasonable speed, and trade resources working in a flow and not over capacity, so the team can continue preparing and executing work with quality at the source.

When we increase work in process, the team is so focused on handling the overburden on the project and the project resources, they do not see roadblocks and remove them ahead of the work. Their ability to prepare work diminishes, and more defective work is passed through the system.

But if we finish as we go, we will have the capacity to focus on the work that is in process and complete it right the first time. Then, quality at the source becomes a culture.

Quality at the Source

Quality at the source is a concept that states a culture and condition in which each employee and worker is held responsible for ensuring the quality of products at the place of work or when the work is produced. To do this, the culture and project systems must include standardized work, self checks, peer checks, visual management, mistake proofing, and continuous improvement of the system.

The Law of the Effect of Variation

The law of the effect of variation is that throughput time and, in effect, the overall project duration is affected as variation increases. Variation has a higher effect on throughput time as the process gets closer to 100% utilization.

For construction, this means that variation is detrimental to flow. Variation comes in two forms:

• Inevitable variation: This type of variation comes from sources that are not under the control of the project team or the contractor. Examples of this would be weather, acts of God, or customer change requests.
• Non-inevitable variation: This type of variation comes from sources that are under the control of the project team or the contractor. Examples of these would be defects that cause delays, RFIs that slow the work, and interruptions in trade flow.

When we assert that Takt planning creates stability and reduces variation, the frequent response is: things change on a construction site. How can you reduce variation? The answer is we reduce non-inevitable variation. We don’t want everything changing, especially if we are unable to respond. What do we do? We stabilize what we can so we can focus where we need to.

The Parade of Trades Simulation Explained

The parade of trades simulation proves flow beats pushing. The tricky secret is that the red team had a regular six-sided die with numbers ranging from 1 to 6. The blue team had a die that could only roll 4s, 3s, and 2s.

A week with a normal die would be something like 6552211 and would represent an attempt to move quickly at first, followed by a slowdown because of an interruption and eventually ending up stuck. A week with the blue team’s die might look like 4434322.

This is synonymous with the concept of maintaining the flow of work. A flow has very little variation. When that variation is eliminated, the chips, or the work, could flow from one end to the other without getting held up by the overwhelming rush from rolling a 6 or the painful crawl of a 1.

The Current State of Our Industry

We often hear people say: we need to get out of the ground fast when we can influence a fewer number of contractors. This is correct. We also hear that we need to be aggressive with complex and unknown areas and scopes on our projects. This is also correct.

Sometimes, though, well-meaning folks will apply both of those concepts to the entire project and say things like: advance the schedule whenever possible, or I want all my materials here now, or just bring it, or I’m a pusher, or keep pushing everything that you can for the schedule. These people are trying to roll 6s in the parade of trades, which will later cause a mess of 1s.

Consider this: What if we told a superintendent to slow down a little and keep a steady flow and an even pace? What would he say? He might say we know absolutely nothing about construction and should go get another job, and yet the data shows he will finish in 26 weeks with 380 people on site with a material inventory of 10 like the red team parade of trades.

This happens because people think movement equals production, which is not the case. All this movement is actually a waste. People start pushing and creating variation because it makes them feel good. It gives an impression of progress, but it requires twice as much material, 100 more people, and 5 more units of material inventory in a week.

What Happens When You Don’t Keep Flow

What happens when a superintendent won’t keep a flow on the schedule and hold dates? If you’re a trade partner, how would you react? You would possibly keep more people on site and most definitely keep more materials on site.

What happens when materials pile up on site? You guessed it. Production slows down. Everything slows down because part of our workforce is dedicated to managing, moving, material inventorying, fixing, replacing, reordering and organizing materials.

The Factory Analogy

Imagine a factory with an assembly line that produces a certain number of finished items every hour. The throughput is the rate at which the factory can process the raw materials into finished items. Say you have 4 machines in the factory that work together to produce the final finished product.

The first machine can work on 4 items per hour, the second at 2 items per hour, and both the third and fourth machines produce 4 items per hour. What is the throughput of the system per hour? The answer is 2 because of the bottleneck.

But here is the moment. The throughput in the example where we had all machines working at full efficiency, even though they were going different speeds, is likely going to be 1.25 or 1.5 finished items per hour, not 2.

Consider what happens between the first and second machines. An inventory of materials begins to pile up. Manpower is then allocated to manage the material inventory. Space in the factory diminishes. People who would otherwise be running machines are now managing the machines and the material inventory. Workers down the line are waiting, and more resources are needed to manage materials on the third and fourth machines.

Waste increases and the speed of the system decreases. Therefore, the throughput is 1.5 or fewer finished items per hour instead of 2. They would have been better off to increase the 2 part per hour machine to a 4 part per hour machine or slow everything down to 2.

The Cost Difference

The cost associated with variation is staggering. Compare just this little example the cost difference between the red team and the blue team:

• Red team: 26 weeks with 380 workers = 380 people × $55 per hour × 40 hours per week × 26 weeks = $21.7 million
• Blue team: 21 weeks with 280 workers = 280 people × $55 per hour × 40 hours per week × 21 weeks = $12.9 million

This is an $8.8 million difference. This ratio can be scaled for whatever size project we are talking about. The difference is in the man hours spent in waste because of a lack of flow. Pushing literally destroys projects, and our industry superintendents must protect us from this.

The System Failed You

Let’s be clear. When superintendents push instead of creating flow, it’s not entirely their fault. The system failed by teaching that movement equals production when actually movement is waste. Nobody showed the parade of trades simulation proving flow beats pushing. Nobody explained that red team rolling sixes and ones finishes in 26 weeks with 380 workers costing $21.7 million while blue team rolling consistent threes and fours finishes in 21 weeks with 280 workers costing $12.9 million. That’s $8.8 million difference on same project. The system taught push when actually flow beats push every time.

The system also failed by not teaching to finish as you go. If your project needs superintendent coaching, project support, or leadership development, Elevate Construction can help your field teams stabilize, schedule, and flow. Historically, superintendents accelerate project to gain time at end to fix mistakes from going too fast and not installing right first time. With Takt, cannot and should not need excessive buffer time for excessive punch list. Team inspects work, identifies punch, corrects as they go. Before trade leaves area, should install production work AND finish scope in entirety. Mindset shift but delivers better quality product sooner. The system taught create buffer for rework when actually finish as you go prevents rework.

The system fails by teaching to push work to start as early as possible putting too much work in process. Work in process must be managed, maintained, checked, monitored, supported utilizing precious resources. When work in process increases, capacity of team and project resources reduced causing rework and eight wastes. Team so focused on handling overburden they don’t see roadblocks and remove them ahead of work. Ability to prepare work diminishes. More defective work passed through system. The system taught maximize work in process when actually limiting work in process creates capacity to complete work right first time.

The Challenge

Here’s your assignment. Stop rolling sixes and ones. Start rolling consistent threes and fours.

Understand the data. Red team (variation) finishes in 26 weeks with 380 workers costing $21.7 million. Blue team (flow) finishes in 21 weeks with 280 workers costing $12.9 million. $8.8 million difference. Flow beats pushing. Movement doesn’t equal production. Movement is waste.

Level out around bottlenecks. First optimize bottlenecks for slower installations. Then even out throughput of remaining work. Little merit to going faster than general throughput. When faster trades continue going fast, people stacked in areas without flow, burying scopes, installing too early, increasing defects.

Finish as you go. Don’t accelerate project to gain time at end to fix mistakes. Inspect work, identify punch, correct as you go. Before trade leaves area, install production work AND finish scope in entirety. Delivers better quality product sooner.

Limit work in process. Don’t push as much work to start as early as possible. Too much work in process must be managed, maintained, checked, monitored. When work in process increases, capacity of team and project resources reduced. If you finish as you go, have capacity to focus on work in process and complete right first time.

Reduce non-inevitable variation. Stabilize what you can so you can focus where you need to. Don’t want everything changing especially when unable to respond. Variation is detrimental to flow. Variation has higher effect on throughput time as process gets closer to 100% utilization.

Stop pushing. If you hear someone ask to create variation in schedules and flow, be skeptical. If you hear folks ask for increase of material inventory, give it second look. Want to create stable environments, keep workers installing work they plan for day or week, have plan for everything.

Don’t roll sixes and ones when you can roll threes. The answer is not to push. Flow will always reduce materials, manpower, mistakes, and time it takes to do something.

On we go.

FAQ

What is the parade of trades simulation?

Red team has normal die (1-6) representing variation. Blue team has modified die (2s, 3s, 4s) representing flow. Red team finishes 26 weeks, 380 workers, $21.7 million. Blue team finishes 21 weeks, 280 workers, $12.9 million. $8.8 million difference. Flow beats pushing. Consistent threes and fours beat sixes and ones.

Why does pushing slow down production?

When materials pile up on site, part of workforce dedicated to managing, moving, inventorying, fixing, replacing, reordering, organizing materials instead of installing work. Waste increases and speed of system decreases. Movement doesn’t equal production. Movement is waste creating impression of progress but requiring twice as much material, 100 more people.

What does “finish as you go” mean?

Before trade leaves area, should install production work AND finish scope in entirety. Don’t accelerate project to gain time at end to fix mistakes. Inspect work, identify punch, correct as you go. Delivers better quality product sooner. No excessive buffer time for excessive punch list.

Why limit work in process?

CPM pushes work to start as early as possible putting too much work in process that must be managed, maintained, checked, monitored. When work in process increases, capacity of team and project resources reduced causing rework and eight wastes. Team focused on handling overburden doesn’t see roadblocks. If you finish as you go, have capacity to complete work right first time.

What’s the difference between inevitable and non-inevitable variation?

Inevitable variation: not under control of project team (weather, acts of God, customer changes). Non-inevitable variation: under control of project team (defects causing delays, RFIs slowing work, interruptions in trade flow). Reduce non-inevitable variation. Stabilize what you can so you can focus where you need to.