Hello TOC

TOC in Operations · Lesson 5 of 7

Choke the release

Most of an order’s lead time is spent waiting. Limit how far ahead work is released, and watch work in progress, lead time and lateness fall together. Then find the sweet spot yourself.

18 min · Not completed yet

Where does lead time go?

Follow one order through a typical plant and time it. You will usually find that the time anyone actually spends working on it, the touch time, is less than 10% of its total lead time. The rest is waiting: in queues, in front of machines that are busy with other orders, at assembly for a missing part.

Two things follow from that:

  • Lead time is mostly queue. And queues are made of work in progress. If you halve the work in progress at the same output, you roughly halve the lead time.
  • In an emergency, an order can move very fast. Everyone has seen an urgent order go through in a fraction of the normal time. It doesn’t need more capacity, only a clear road.

Pause and think: a plant’s lead time is six weeks, and it is often late. What do most managers do?

  • Release work later, so less of it is on the floor
  • Release work earlier, to give it more time
  • Ask customers to order less
  • Stop accepting rush orders

“Start earlier” feels like protection. But releasing earlier puts more work on the floor, lengthens every queue, and gives every work centre more choice to cherry-pick from. The lead time grows to fill the new release time, and lateness doesn’t improve. More time is not better.

Limit the visibility

The TOC remedy is simple enough to fit in three lines:

  1. Identify the clean orders: those whose materials and information are complete, so they can actually run.
  2. Prioritise the clean orders by dispatch date.
  3. Limit visibility to about half of the current lead time. If the current lead time is six weeks, release only work due within the next three.

Choking the release brings the excess WIP down, and with it, two things happen. Queues form only at the real constraint, so the constraint becomes visible where before there seemed to be many. And with fewer orders on each work centre, there is little to cherry-pick from, so the batching instinct from lesson 3 stops doing damage.

Why half? Because touch time is still a small part of the remaining lead time. Half the old lead time leaves plenty of buffer for normal disruptions.

Play: Choke the Release

Before you play

The plant. A small A-shaped plant. Every order needs two parts that meet at Assembly: a P part (Cutting, then Machining) and a Q part (Cutting, then Welding). Machining is the constraint; the other machines have spare capacity. Machines batch the same set-ups whenever they can, just as people do.

The customers. Most order 20 days before they need delivery. A quarter are short-notice: they need it in 5 days.

What you control. Only one thing: when an order is released to the floor.

  • Weekly plan (the usual way): every Monday, release everything due in the next 14 days.
  • Daily rope: every day, release only orders due within your chosen window, from 1 to 14 days.

What to watch. The queue bars during the run, then four numbers: orders delivered on time, short-notice orders on time, orders tied up on the floor or waiting to ship, and days an order spends on the floor.

Your goal. Find the window that gets nearly everything on time, short-notice orders included, with the fewest orders tied up. Every run uses the same customers and orders, so only your rule changes. Start with the weekly plan, then try at least three rope windows.

What the curve shows

If you tried several windows, you drew a U-curve:

  • Release too late and the plant starves: orders arrive on the floor with less time than they need.
  • Release too early and the floor floods: queues grow, machines batch across far-off orders, parts fall out of step, and the near orders wait. Short-notice orders suffer first.
  • In between there is a wide, comfortable range where nearly everything is on time with a fraction of the inventory. That range starts near half the old lead time.

Notice what didn’t change: the machines, the people, the set-up times and the customers. Only the release rule.

Check your understanding

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

1. In a typical plant, touch time is…

  • About half of the lead time
  • Most of the lead time
  • Less than 10% of the lead time
  • Equal to the lead time

The rest is waiting in queues.

2. The TOC rule of thumb for limiting visibility is to release work due within about…

  • Half of the current lead time
  • Twice the current lead time
  • One day
  • The full order book

Enough buffer for disruptions, without flooding the floor.

3. Releasing work earlier to “give it more time” usually…

  • Shortens the lead time
  • Improves on-time delivery
  • Reduces inventory
  • Lengthens queues and lead time without improving delivery

More time is not better: the lead time grows to fill it.

4. Why does choking the release make the constraint visible?

  • Because it adds a sign to the constraint
  • Because with less WIP, queues form only in front of the real constraint
  • Because the constraint speeds up
  • Because other machines are removed

High WIP creates queues everywhere, which looks like many constraints.

5. Release work too late, and…

  • Inventory rises
  • Nothing changes
  • The plant starves and orders can’t finish on time
  • The constraint moves to the market

That’s the left side of the U-curve. Choke, but don’t strangle.