Monday, July 12, 2010

Do You Know How to Handle Your Constraint?

A recent plant visit was a real lesson in constraint management.  What I have found is that we have a weak knowledge of how to identify a bottleneck and how to effectively manage this constraint.

To discuss this subject it is important to start with defining the distinction between a constraint and a bottleneck.  A constraint is a resource with the highest load.  A bottleneck is a resource that is unable to meet current demand. Manufacturing processes can have multiple bottlenecks.  The constraint is the bottleneck whose performance is directly related to the overall system performance.

The roots of a bottleneck focused approach in operations management can be traced back to the days of Henry Ford. He understood that the workstation with the maximum processing requirement, denoted as the bottleneck, would constrain the output of the system.

Eli Goldratt who authored "The Goal" was one of the first to write about constraint management in what he called "Theory of Constraints".  Goldratt defined 10 principles for managing a constraint.  Note that below the word "bottleneck" is used, but it is often a constraint rather than a bottleneck.

1. Balance flow, not capacity.  During the manufacturing of a product there will inevitably be faster and slower processes.  Therefore effort should be made to achieve continuous flow of materials. 

2. The utilization of a non-bottleneck is determined not by its own capacity but by some other constraint in the system.  It is the bottleneck operation that should govern flow, therefore the capacity and utilization of non-bottleneck resources is mostly irrelevant.

3. Utilization and activation are not synonymous.  A non-bottleneck machine should not be 'activated' all the time because overproduction will result.  Utilization of the machine occurs when the machine is activated to produce at a balanced rate.

4. An hour lost at a bottleneck is an hour lost for the whole system.  Since the bottleneck governs the amount of throughput for the factory, if the bottleneck stops it is equivalent to stopping the entire factory.

5. An hour saved at a non-bottleneck is merely a mirage.  In effect increased capacity at a non-bottleneck is worthless.

6. Bottlenecks govern both throughput and inventory systems.  A factory's output is the same as the bottleneck's output, and inventory should only enter the factory at a rate that the bottleneck is capable of handling.

7. The transfer batch may not, and many times should not, equal the process batch.  A transfer batch is the amount of work in process inventory that is moved along between workstations.  To maintain flow and minimize inventory costs this batch should not necessarily equal the production batch quantity.

8. The process batch should be variable, not fixed.  The system should not be constrained by the artificial requirements that product must be made in one large batch.  The batch size of a bottleneck machine should not necessarily equal the batch size of non-bottleneck machine.

9. Lead times are the result of a schedule, and can not be predetermined.  Pre-specified lead times of typical MRP solution will not reflect the true situation.

10. Schedules should be assembled by looking at all constraints simultaneously.  Constraints may be in the form of machines, labor, and material.  You must consider them all together when scheduling.

These 10 principles can help plan production effectively.  The plant schedule should be organized around the constraint since it determines the throughput of the plant.  It is critically important to learn how to identify bottlenecks, determine the system constraint, and learn to leverage the constraint to your advantage.  All systems have a constraint and this will not change but how you deal with them can.

If you enjoy this post you may want to connect with me on Linkedin or follow me on Twitter.  You can also subscribe to this feed or email to stay updated on all posts.  For those Facebook fans join A Lean Journey on our facebook fan page.


  1. Tim,

    This is an excellent post! Everyone who wants to make lean improvements in their plant should read this closely. Thanks for clarifying constrains and bottlenecks. Those terms are oftern used erroneously. The Goal is still a good read.

    Thanks again,

  2. It is great to see someone go back to "The Goal" and bring out the basics of it. It is a wonderful read and Herbie will stick with you forever. People still have a hard time realizing the product will not leave the plant any faster than the slowest process (#6). We still want to activate machines and not utilize them for flow.

    Thanks for bringing back "The Goal".

  3. Thank you for summarizing the constrain management. As "bottlenecks" occur too frequently across the globe, especially in SME sectors; I think it should be categorized as "one of the main causes" of human suffering in working population.