Hello TOC

TOC in Operations · Lesson 4 of 7

The TOC way of production

Flow as the first objective of operations. Why local efficiencies have to go, how Drum-Buffer-Rope controls work in progress for the whole plant, and how it differs from controlling it everywhere.

14 min · Not completed yet

The TOC way of production starts from one statement: flow is the most important objective of operations. Everything else follows from it:

  1. A practical mechanism to avoid over-production: work enters the floor only when it’s needed.
  2. Local efficiencies are abolished: non-constraints are not judged on how busy they are.
  3. A focusing process to balance flow: one simple priority signal tells everyone what to work on next.

Why local efficiencies have to go

Suppose machine C is the constraint and you want it busy 100% of the time.

  • C needs a continuous bank of work in front of it, or a hiccup upstream will starve it.
  • When machines A or B upstream are disrupted, they must not only keep feeding C but also rebuild the bank. So A and B must have more capacity than C.
  • The same goes for D after C: it must clear C’s output and recover from its own stoppages.

So every resource except C must have spare capacity just to keep C fully busy. If you then also demand that A, B and D be 100% busy, the only place their extra output can go is inventory, and the inventory hits the roof. You can’t have both. Keeping the constraint busy requires the others not to be.

Pause and think: a supervisor proudly reports that every machine in the plant ran at 95% utilisation last month. What does that tell you?

  • The plant is perfectly balanced
  • Output must have been at its maximum
  • Nothing: utilisation is always good news
  • Either inventory went up a lot, or the constraint had no protection when things went wrong

Non-constraints at 95% have almost no protective capacity. Either they produced ahead of need (inventory), or the first breakdown will ripple straight through to the constraint.

Drum, buffer, rope

  • Drum. The constraint’s schedule sets the pace for the whole plant: what to produce, in what sequence. Plan it first.
  • Buffer. Time, not stock: material is released early enough that it reaches the constraint, and the shipping dock, with a margin for normal disruptions. A single buffer placed where it matters protects far better than padding every operation.
  • Rope. Material release is tied to the drum. Nothing goes onto the floor before the drum will need it, minus the buffer time.

The rest of the plant is not scheduled. Non-constraints simply work on whatever has arrived, in priority order, and stop when there is nothing to do.

In most plants today the market, not a machine, is the real constraint, and the busiest internal resource has a little spare capacity. The simplified version of DBR used most often takes that into account: it keeps one buffer, before shipping, ties the rope to the order’s due date, and watches the most loaded resource so it isn’t overloaded.

Controlling work in progress: space or time?

There are two ways to stop over-production.

By space, as in conveyors, trolleys or kanban squares: when the space between two work centres is full, the upstream one stops. It’s simple, but it limits WIP between every pair of work centres. Since the WIP needed in front of the constraint depends on the normal disruptions upstream and downstream, the squares are usually kept to a bare minimum, and the constraint loses output whenever a disruption is bigger than the square allows.

By time, as in DBR: WIP is not restrained between every pair of work centres. It is controlled for the whole system, through the rope, so the effects of disruptions are absorbed across the system, and extra work is held only where it’s needed: in front of the constraint and before shipping.

Exercise: old way or TOC way?

  • Every work centre has a monthly utilisation target
  • Only the constraint has a detailed schedule
  • Material is released when the drum will need it, minus the buffer time
  • Material is released as soon as it arrives, so nobody runs out of work
  • Non-constraints are expected to stand idle at times
  • Small fixed WIP limits between every pair of machines, whatever the disruptions

Check your understanding

Answer all five. Get four right to complete the lesson.

1. In the TOC way of production, the most important objective of operations is…

  • Flow
  • Utilisation
  • Cost per unit
  • Headcount

Flow comes first; cost is protected by protecting the constraint.

2. To keep the constraint busy 100% of the time, the resources feeding it must…

  • Also be busy 100% of the time
  • Have less capacity than the constraint
  • Have more capacity than the constraint, so they can rebuild its buffer after a disruption
  • Be scheduled in detail

That spare capacity is protective capacity, not waste.

3. What does the rope do?

  • Schedules every machine
  • Ties material release to the constraint’s needs
  • Measures utilisation
  • Holds stock in front of every machine

Nothing enters the floor before the drum will need it, minus the buffer.

4. Compared with controlling WIP by space between every pair of machines, DBR…

  • Holds more WIP everywhere
  • Removes all WIP
  • Has no buffers
  • Controls WIP for the whole system and holds extra only where it’s needed

Disruptions are absorbed across the system, and the constraint stays protected.

5. In DBR, how are non-constraints scheduled?

  • In detail, hour by hour
  • To maximise their utilisation
  • They aren’t: they work on what has arrived, in priority order, and stop when there is nothing to do
  • By the sales team

Only the drum is scheduled. That is what keeps planning simple.