THE GENERIC TOC
The theory of constraints makes a clear distinction between
the decision taken at the planning stage and those
taken at the execution stage.
The decisions at the planning stage:
Decide upon the objectives of the planning—what do
we commit to our market?
Identify the critical elements that if impacted by
Murphy might disrupt the objectives. For instance,
any bottleneck is vulnerable to any delays, because it
doesn't have the capacity to catch up.
Plan the critical elements as needed.
Plan the buffers that protect the objectives and the
critical elements in the plan.
The decisions at the execution stage:
Follow the critical elements of the plan as is.
Monitor the state of the buffers.
Prioritize the activities according to state of the buff
ers they feed.
Expedite only those activities that feed buffers that
are at risk.
EXAMPLE NO. 1:
SCHEDULING THE SHOP
TOC has a methodology for planning the shop floor
called drum-buffer-rope (DBR).
The "drum" is the master production schedule plus
the detailed schedule of the capacity constraint resource
(CCR). All resources that have excess capacity are not
scheduled in detail, as they have enough excess capacity
to overcome Murphy.
"Buffer" is the protection mechanism. TOC uses mainly
time buffers to protect the on-time performance. The "time
buffer" is a similar concept to MRP's fixed lead time that
is used between levels in the bill-of-material (BOM), with
one critical difference. In DBR the "buffer" covers the whole
way from the raw materials release up to the constraint or
the shipping. Like "lead time," it allows enough time for
queue (wait time), setup, run, and move. But it does so for
much longer than MRP. The rationale is to protect only the single
capacity constraint and the shipping dates.
Rope is the part of the planning that schedules the
release of material.
TOC assumes that only one critical resource may truly
constrain the whole system. Hence, all the rest of the
resources need to have excess capacity. That excess capacity allows
the non-constraints to be much more flexible.
Hence, there is no need to detail the schedule for the
non-constraints. Certainly there is no need for protecting
The DBR methodology dictates the rules for the
planning, which includes the buffers at the critical
places. The planning objectives are shipping every firm
order on time. In order to ensure on-time delivery,
the capacity of the most loaded resource (called CCR—
capacity constraint resource) should be carefully
checked. When the market demand exceeds the CCR's
capacity, we like to load the CCR to its limits, and to
commit delivery based on the CCR schedule. We need
to allow enough time after the CCR for the order to
be completed SAFELY. That time, which is the estimated
lead time plus safety, is called the shipping
buffer. When we load the CCR to its limit and commit
accordingly to the market, the detailed schedule of the CCR
is a critical element of the planning. We
need to ensure that the CCR will
be able to follow the
schedule. That means buffering the CCR with a CCR buffer. The
CCR buffer allows enough time for the materials to reach the CCR
before it needs the parts. This
is done by the rope part of the planning. Of
course, we also need to ensure
the availability of materials
to support the release schedule.
Buffer management monitors the buffers. If the CCR
buffer in Figure 4 is 10 days, then 7 days after the release,
the parts are expected to be at the CCR's site. If not, meaning
70 percent of the buffer has been consumed and the
parts are not where they should be, expediting actions should push
the parts to the CCR. So, the schedule on
the CCR should be followed as is, except for probable
delays on the CCR itself. The shipping buffer should take
care of these delays—unless more than 70 percent of that
buffer is consumed, which also triggers expediting.
The combination of planning that incorporates buffers to protect
the objectives and the critical elements of
the planning produces a much less detailed planning
that strives to do the most the system can do (TOC terminology
of that is: exploiting the system constraint),
and yet it is a robust planning.
The part that is the execution leaves quite a lot of decisions
that either are not crucial for meeting the due dates
OR concentrate on the real threats to the plan as viewed
by the buffer management methodology. Every exception
is treated so the original plan will still be carried out.