In this instalment of the series on solving any problem, we continue on from the initial point of view that we are attempting to optimize the flow of a pipeline.
Define the Goal
The first thing we need to do is to define what the goal is. In a sense we are using a filter to view our system. For most for-profit organisation the goal is to maximise profits. If there is product already flowing through then we should be able to observe a relatively stable flow from inputs to outputs through a pipeline. The flow rate through the pipeline is most affected by the constraint, the bottleneck that allows the least amount of work-in-progress (WIP) to pass.
Five Focusing Steps for the Constraint
Once we have identified the goal and the pipeline we can apply the five focusing steps of TOC. They are:
- Find the Constraint
- Exploit the Constraint
- Subordinate Everything Else
- Elevate the Constraint
- Repeat
Find the Constraint
Rhe first thing we need to do is to locate this constraint. Depending on the configuration of the pipeline, there can be more than one bottleneck. If we consider a simple straight line value stream, there is but ONE bottleneck, that we call a constraint, similar to the analogy of a chain, that is the weakest link in the pipeline. If we use the simple production line this would be the work center with the lowest flow rate. The same will go for a multi-stream pipeline. We just have to identify first the critical chain and then using flow rates identify the constraint, the work center in the chain with the lowest flow rate.
Exploit The Constraint
Essentially this is the only work center whose efficiency we care about, so we need to ensure that it doesn’t stop working to meet our flow rate goal. For example, We can ensure the downtime is reduced by eliminating special causes or simply removing excessive work orders that reduce the flow of work-in-progress to the constraint. If we have a very balanced line we need to take care that the constraint does not shift to another work center while we are not prepared.
Subordinate Everything Else
Sometimes the previous step may result in the constraint moving to another work center. If not, we subordinate the other work centers to ensure the constraint’s performance is maximised.
Up to now we are not required to expend much if any OPEX and no CAPEX. We are rearranging our resources and addressing special causes when they arise. This is typical day-to-day work in a manufacturing plant and is nothing special. The tools we can use at this stage include Theory Of Constraint’s Thinking Processes and Toyota Production System (TPS) A3 Thinking. So up to this point even the smallest companies can apply this principle.
Elevate the Constraint
This fourth step is when we improve the flow rate of the constraint, and needs to be approached with caution. We need to consider where the constraint will be should our improvements be successful. The tool to use is Design of Experiments as allow us to deal with common causes in process management.
How does this compare with 6 Sigma or Toyota Production System?
6 Sigma does not have any concept of flow except for the SIPOC.
Lean does address this as part of the value stream made up of a linear series of work centers. The objective in Lean is to meet the takt time of sales, and this is done by tuning the flow rate of every work center in the value stream to share the exact same takt time.
There are three main issues I observed when I had to implement this for my improvement programs. The first is the tuning of all the stations meant that instead of one constraint, we now have a family of possible constraints. This is because we cannot always remove the fluctuations in velocity through each work center, and so from time to time we will see the constraint moving about the production line simply due to random fluctuations. This can be a problem when the fluctuations are large.
The second issue is the cost of making those changes of flow rate are often quite high, since we are making changes in flow rate. Instead of a quick win, this represents an outflow of revenue from the beginning. This can be viewed as unattractive to organisations that are already suffering from poor productivity. And we might want to change the flow rate in the future, incurring yet more costs each time.
Finally, as part of all projects we have to justify the changes. Since there is significant expenses this drives cost reduction often via headcount reduction. I recalled that the accountant that was part of my projects always wanted to know how many FTEs were to be removed and when those savings could be booked. This bothered me because I knew this was not how Toyota did it, and I knew that this would affect morale issues both for the direct and indirect labor involved.
Using TOC methodology we are able to enjoy quick wins by our focus on the constraint, without much outlay of resources. In addition once the changes were made and credibility was built it was much easier to make bigger and more rewarding changes that elevated the constraint. This is unlike Lean where I experienced the need for front loading of costs from Day One, including the need for training of the workforce. For TOC we need only to educate those involved in defining the goal, and those involved in the focusing steps, because the efficiencies for non-constraint work centers would no longer be attractive. For those used to high efficiency levels, it can be unsettling to be out of work, and some assurances that this is going to be the norm moving forward would be needed.