As I said in my last posting, we will now discuss how policy constraints can have a negative impact on the throughput and profitability of organizations. And like the simple, common sense improvement ideas presented in Focus and Leverage Part 319, the elimination of policy constraints to gain additional throughput don’t cost anything to implement,
Perhaps the worst policy constraint of all is the general belief that all steps in the process must run at full capacity to drive up the performance metrics, operator efficiency or equipment utilization. By running all steps in the process at full capacity, work-in-process inventory levels will skyrocket and overall cycle times will be extended. In reality, the only process step that should be measured for efficiency or utilization is the system constraint and both metrics should be pushed higher and higher
Another of the more damaging policy constraints is the misguided effort to focus on cost reduction to improve profitability. What typically happens when managers are forced to continually reduce costs is that at some point in time, the constraint operation will be negatively impacted which automatically results in less throughput. When this happens, profitability will be negatively impacted.
Another policy constraint that can have a negative impact on throughput is the mistaken belief that line balancing is a good thing. Yes, it’s one of the basic teachings of Lean, but when you stop and consider what this policy can potentially do to throughput, you might change your mind. Let’s consider our imaginary 4-step process again. Proponents of line balancing focus on attempting to have each step in the process be close to takt time so that our process might look like the following.
Here we see the results of a line balancing effort where work has been distributed equally between all four steps in the process. This is great….right? Ask yourself what happens to the throughput if any of the process steps has downtime? Because of the equally distributed workload, there is no longer any possibility to “catch-up” if downtime occurs because we have removed our sprint capacity. That is, when one step goes down, the entire line stops! So while line balancing might appear on the surface to be a “good thing,” in my opinion, it really is not. What we actually see many times in balanced lines is that the constraint will move around the process numerous times during the day. It’s much harder to manage a balanced line.
There are other examples of policy constraints in Steven Bragg’s book, but as he points out, they all seem to deal with the same thing…..local optimization. By attempting to improve disparate parts of the process, we miss the true objective of system optimization. Let’s now turn our attention to Goldratt’s third step in his 5 focusing steps, subordination. Again, I like to present this concept using a visual example of the negative effect of not subordinating all other process steps to the constraint. To do so, let’s return to our original four-step process.
Because Step 3 has been identified as the system constraint, the concept of subordination means that we should effectively run all other steps at the same pace as the constraint. In our example process, this means that effectively each step should be running at one part/patient every 90 minutes. Why is this important you may be thinking? Let’s look at what would happen if every step in the process was running at its full capacity. That is, Step 1 would process 1 part/patient every 30 minutes and Step 2 would be processing 1 part or patient every 45 minutes and so forth. Graphically, the figure below demonstrates the impact of running every step at its full capacity limit.
In the figure below we see the unavoidable consequences of running steps 1 and 2 at their full capacity. Work-In-Process (WIP) inventory begins to grow immediately….first in front of Step 2 and then in front of Step 3. This move automatically results in excessive wait times for both products and patients and the sad truth is that the longer this process is permitted to run at full capacity, the WIP will grow proportionally. This of course happens because of the difference in individual processing times for each step. That is, because Step 1 is only 30 minutes and Step 2 is 45 minutes, WIP accumulates in front of Step 2. And because Step 2 is twice as fast as Step 3, WIP accumulates in front of Step 3 at an even faster rate. By contrast, even though Step 4 is faster than Step 3, it is forced to run at the same rate as Step 3 because it only receives parts or patients at a rate of 1 every 90 minutes.
So now we have used the first three steps of Goldratt’s 5 focusing steps, identify the system constraint, decide how to exploit the system constraint and subordinate all other steps to the system constraint. It’s important to remember that during the first three steps, usually no money is spent so effectively improving throughput here is free. But what happens if after the first three steps, the capacity of the constraint is still less than the market demand? In other words, if all of your improvements to the constraint are still not satisfying the demand that exists for your products or services, what is the next step? In my next posting we will look at Step 4 of Goldratt’s 5 focusing steps.