Operations, as Taught by Eli Goldratt · Lesson 8 of 8

Only actions change reality

Goldratt ended every session the same way: knowledge is worth nothing until it is acted on. Three actions you can start this month, the obstacles you'll meet, the results to expect, and his answers to the questions managers always ask.

16 min · Not completed yet

Daniel: “I understand it all now. I’ll put a proposal to the board next quarter.”

Ruth: “Next quarter? You can start three things on Monday, within your own area of influence. And your area of influence is much larger than your area of control.”

Action 1: strengthen the bottleneck candidates

Identify the resources that might be the bottleneck and increase their effective capacity, without buying anything:

  1. Make sure they are constantly manned, through breaks and shift changes.
  2. Stagger breaks so the machine never stops for lunch.
  3. Make sure no tools or parts are ever missing when they’re needed.
  4. Offload work to other resources, even less efficient ones.

Then look for the results: more throughput, and due dates met.

Action 2: cut the batch sizes

Start by cutting batches in half, and no more than half. Cutting more may eat too much protective capacity in extra set-ups, and could even turn a non-bottleneck into a bottleneck. Check for the benefits: shorter lead times, less inventory, less expediting everywhere.

Action 3: implement DBR and buffer management in your area of influence

Get consensus, in this order:

  1. On the concept (use The Goal, the book or the film, or a presentation).
  2. On the drum.
  3. On the buffer sizes.
  4. On the mechanism for tying the rope.
  5. From everyone who will have to work under DBR and buffer management.

Your obstacles, and how you’ll overcome them

Goldratt’s programme asked participants, after each action, to type in what might stop them and what they would do about it, then print it and do it. Do the same here.

Write each obstacle as one line, followed by the action that overcomes it. If an obstacle feels like a wall, it usually hides a conflict: the conflict cloud template will help you take it apart.

What to expect

Based on experience in hundreds of plants and other operations, Goldratt listed the results to expect within about two months:

  • Inventory: work in progress down by at least 30%.
  • Lead time: down by at least 50%.
  • Due-date performance: late deliveries cut by at least half.
  • Capacity: at least 30% hidden capacity uncovered.

Questions managers ask

“We produce for stock, not to order. Does this apply?”

  • No: DBR is only for make-to-order plants with due dates on every order
  • Yes; shorter lead times often turn make-to-stock into make-to-order

We produce for stock when production lead time is much longer than the customer’s tolerance time. Then the customer seems unwilling to wait at all. But customers do have a real tolerance time: a week, two or three weeks, just not two months. DBR and buffer management often shrink lead time below that tolerance, and the plant can produce to order. (Where stock is held at remote locations, the distribution solution applies.)

“Which scheduling software supports DBR?”

  • Ask how it handles disturbances: planning and execution should be separate
  • Any software that can reschedule the whole plant in minutes after a disturbance

Many vendors claim TOC compatibility; most don’t. The ability to reschedule everything quickly is the wrong answer: rescheduling amplifies disturbances. Goldratt explored the deeper question, how to evaluate an investment in software, in his novel Necessary But Not Sufficient.

“We’re designing a new production line. Where should the bottleneck be?”

  • As far downstream as possible, so that everything before it is protected
  • As far upstream as possible
  • It’s a strategic choice; the middle needs the least protection

Every organisation has a constraint, so deciding where it should be is a strategic decision. Downstream: more certainty about due dates, but more inventory and lead time. Upstream: less inventory and lead time, but less certainty about due dates (and with scrap, throughput may suffer). In the middle: the least protective capacity is needed, so the least investment.

And the last question: which measurements must change, and what replaces them? Goldratt devoted an entire session to it: Finance and Measurements, the next course in this series.

The journey, in one page

  1. The undesirable effects of an operation are not separate problems.
  2. They grow from one mode of operation: local efficiency as the prime measurement.
  3. The measurement survives because of a conflict: fight waste versus fight for flow.
  4. The conflict rests on a false assumption: that an idle resource is a waste.
  5. Activating is not utilising; only the bottleneck’s time is the system’s time.
  6. Drum-Buffer-Rope plans the flow; the Roadrunner rule is its work ethic.
  7. Buffer management executes it, focuses improvement and protects growth.
  8. Only actions change reality.

Check your understanding

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

1. Which action increases bottleneck capacity without investment?

  • Buying a new machine
  • Adding a shift everywhere in the plant
  • Staggering breaks so it is always manned
  • Running bigger batches

Along with ready tools and parts, and offloading work.

2. Batch sizes should first be cut by…

  • 90%
  • Half, and no more
  • 10%
  • Nothing

More could eat the protective capacity.

3. Within two months, Goldratt expected lead time to fall by at least…

  • 5%
  • 10%
  • 25%
  • 50%

With WIP down at least 30% and late deliveries cut at least in half.

4. Make-to-stock is forced when…

  • Production lead time is much longer than the customer’s tolerance time
  • Customers are impatient by nature and won’t accept any wait
  • Machines are expensive
  • Forecasts are accurate

Shorten the lead time and the customer is willing to wait.

5. The first consensus to get when implementing DBR is on…

  • The buffer sizes
  • The software
  • The concept itself
  • The bonus scheme

Then the drum, the buffers, the rope, and everyone who works under it.