As a recent article in APICS Magazine points out, “while containing an issue and eliminating an issue are different things, containment does help keep the problem from getting worse.”
And that pretty well sums up the simple concept behind the long-touted Drum-Buffer-Rope (or DBR) theory first espoused by Dr. Eliyahu Goldratt many years ago, and brought most notably and literally to life in his best-selling book, The Goal.
DBR, as APICS points out, has held its own for a good while now as a method for scheduling and managing operations that have internal constraints or a capacity-constrained resource. That constrained resource with the least capacity will always set the pace for the entire production line. Thus, providing it with a “buffer” of inventory is the simplest way to prevent starving that constraint and thus creating a full stoppage due to lack of work. Simply put, make sure your “constraint” has plenty of inventory in front of it and you have effectively eliminated it as your worst constraint.
From there, the Theory of Constraints tells us, one moves on to the next most critical constraint… and so on, gradually working one’s way through a cascading series of constraints in the never-ending search for continuous improvement.
The idea of DBR then is to put a single information link – the “rope” – between the constraint and production starts. Because it’s a single link, a DBR production control system becomes simpler even than kanban, which requires communication across many workstations, whether that takes the form of production cards, empty containers or empty spaces for inbound materials.
Of course, variations in processing and material transfer can dash the best of plans. Here again, DBR can help. As APICS points out (in an article by William Levinson titled “Completing the Link, Mar/Apr ’15 issue) “If no containment action is taken, inventory accumulates the same way.” The author gives the example of a traffic jam in a highway operating near capacity when a driver slows down. Everyone has to slow down, and no one can make up the resulting loss.
Likewise, in a “balanced factory” where each workstation has the same capacity, variation can cause inventory to accumulate and reduce throughput to less than the expected rate, and time lost at any work station theoretically can never be made up.
If all workstations but one (the “constraint”) have excess capacity, “then the problem is limited to that one workstation. It therefore is necessary to keep a buffer of protective inventory to shield that workstation from starvation.”
The size of that buffer depends on the variation in the system prior to the constraint, notes Levinson. So while this form of containment may not be a genuine solution, it does eliminate the effect of most of the variation. Simple, but effective.