Where Is Your Current Constraint?

We've offered many specific suggestions for improving your CNC environment in this column. Indeed, our primary focus is to help you find ways to improve. However, we've pretty much left it up to you to determine whether a given technique should be employed. We'd like to present a proven approach determining where to start improving, as well as where you should concentrate your efforts in the future.

In our setup and cycle time reduction courses, we simply state that if a machine is in setup for a greater percentage of available production time, then start with setup reduction for that machine. If it is running production for the greater percentage of available production time, start with cycle time reduction. While this may be a reasonably good rule of thumb for a specific machine that you wish to improve, it still begs the question "Which machines should you start with?"

We're going to assume that your company needs to increase its output without increasing operating expenses. This translates into getting more output with your current resources. While there can be other reasons why a company must improve, our discussions apply only if your goal is to increase output.

You must begin by locating your company's current constraint, which is the entity that limits maximum output at the present time.

Think of constraints as being like bottlenecks, and they should be pretty easy to spot. If you're talking about the current company-wide constraint, then determine what department in the company is currently falling behind, holding up production in other areas. Find the department that has the largest backlog of work, and you've probably found the company-wide constraint.

If you have correctly located the current constraint, improving it will increase total output (from the company, department and/or machine).

While improving an area that is not the current constraint may have certain benefits, it will not improve total output.

If you improve enough, the improved entity will no longer be the current constraint. We call this process "eliminating the current constraint." Once the current constraint is eliminated, another will appear. When you find it, your goal will be to eliminate it. When you do, yet another constraint will appear. This process of constraint elimination will be repeated until you're satisfied with the current output level.

So eliminating constraints is an on-going battle. It doesn't end until everyone is satisfied with the current output level (of the company, department and/or machine).

An example of finding the current constraint. Our example product-producing company has three departments: the mill department, the lathe department and the grinder department. The general flow of workpieces includes some work in each department.

To improve on a company-wide basis (increase the company's total output), we must first locate the department that is the current constraint. Say, for example, that throughput time per workpiece is 3minutes in lathe department. In mill department, throughput time per workpiece is 4 minutes and 15 seconds. In the grinder department, throughput time per workpiece is 5 minutes and 20 seconds.

In this case, it's likely that the grinder department is the current constraint. If all departments are working the same amount of time, workpiece production lots are probably stacked up in the grinder department, waiting to be worked on.

Maybe an excessive amount of scrap is being run on a given machine, making it the current constraint for the department. If you find a way to eliminate the scrap, you can eliminate this machine as the current constraint. Of course, some other entity in the department will then become the current constraint. If this machine is truly the current constraint for the grinder department, any improvements you make for it will have an immediate impact on the company's total output.

After eliminating the next current constraint, throughput time per workpiece in the grinding department drops to 3 minutes and 50 seconds. Now the mill department becomes the current company-wide constraint.

Of course, the next goal will be to eliminate the mill department as the current company-wide constraint. When this is done, the grinding department may again become the constraint. This leapfrogging between the grinder department and the mill department will continue until the lathe department eventually becomes the current company-wide constraint.