Hello TOC

TOC in Operations – Advanced · Lesson 9 of 12

A second rope

In a plant where stock and customer orders share components, customer orders lose their material to stock whenever stock demand is high, and get help only when they are already red. Tagging from release is impossible for heavy common parts. Add a second rope at the diversion point: on a set date, the common material is reserved for the order.

20 min · Not completed yet

The situation

A maker of physical security products (safes, strong-room doors, safe-deposit lockers and fire-resistant cabinets) sells in two markets. The domestic market expects short lead times; exports are regular make-to-order.

The plant has both shapes in it. A mother coil is slit, sheared and cut to length (divergent points); blanks are punched, notched and bent; components are pickled and drilled; everything converges at assembly, then paint and final fitting. Components are common across the products of a category; there are many end varieties; resources are shared.

The standard solution:

  • Standard products are made to availability, with a stock buffer of about six days at the warehouse.
  • Customised products are made to order, with a production buffer of 12 to 18 days.
  • Dispatch to branches takes four to five days.

What goes wrong

Material is common between MTO and MTA orders. Whenever the load of MTA orders is high, material meant for MTO orders is stolen for MTA. The MTO orders get priority at assembly only when they turn red. When they pile up, expediting every red order becomes impossible, and the on-time performance of MTO orders falls.

The obvious fix: tag the MTO material from release onwards. Here it can’t be done. The components are heavy, there are many of them, and up to assembly they are processed together because at that level they are identical. They become different only at assembly.

Pause and think: what does the rope at release actually protect against?

  • Stealing: releasing more of one order than the plant can absorb, which leaves others short
  • Late suppliers, by making sure the raw material is always in the plant before release
  • Quality problems at the first operation, by limiting how much is exposed to them

The essence of a rope is that an excess of one thing must not create a shortage of another. The first rope prevents overall capacity stealing. The problem here is a different kind of stealing, of material, at a different point.

The second rope

So add a second rope, at the diversion point, to stop material stealing.

  • When an MTO order reaches 50% buffer penetration, it is most likely at assembly, which in this plant is also the constraint.
  • On that date, the common material in the assembly area is identified and reserved for that order from the current work in process, and the order is completed.
  • If not all its components are there, the available ones are set aside, and the order is completed when the rest arrive.
  • Assembly still follows the normal colour priority, now based on the revised availability of material for MTA and MTO orders.

The second rope is a date, not a schedule. It appears as a separate buffer report, visible only on the additional-rope date, which shows the stage where the rope applies. Nobody has to plan assembly in detail; on that day, the material stops being anyone else’s.

Other rules from the same implementation

Several smaller rules in the same plant were written to stop stealing too:

  • No dispatch before the ex-factory date in the sales order. Shipping early looks good, but it is capacity stealing: it uses constraint time and material for an order that didn’t need it yet.
  • Feeder supplies only on instructions from planning, never on requests from the next department, and never more than the production order asks for.
  • Parts made by vendors for MTO orders get an assembly procurement buffer: the start date for the in-house shops that feed the vendor is the order’s release date minus that buffer, so the vendor’s components reach assembly by the release date.
  • Steady bought-out items run on a two-bin system; fluctuating ones get a stock buffer with buffer management.

Results

In the years after go-live, the business’s output tripled, work in process fell by more than half, monthly sales more than doubled, and finished stock fell by half. Before TOC, 80% of orders took up to 80 days from order to billing; after, 80% were completed in 40 days, with far less variation.

Check your understanding

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

1. MTO orders in the mixed plant lost their material mainly when…

  • The vendors delivered the common components late
  • The MTO orders were released too early
  • MTA load was high and common material went to MTA orders
  • Assembly worked overtime

Common material is taken by whatever is in front of assembly.

2. Why couldn’t the plant tag MTO material from release?

  • Tagging is against the rules of buffer management on the shop floor
  • The components were heavy, numerous, and identical until assembly
  • Customers didn’t want tags
  • The MTO orders had no release dates

They are processed together until assembly, where they diverge.

3. The second rope is triggered when the MTO order…

  • Is released to the plant
  • Turns black
  • Is booked by sales
  • Reaches 50% buffer penetration

By then it is most likely at assembly, and the common material is reserved for it.

4. The second rope is different from the first because it is…

  • A date with a reservation, not a release schedule for a resource
  • A schedule for the assembly line, set by planning each day
  • Managed by sales
  • Only for MTA orders

It adds protection without a second schedule.

5. Why is dispatching before the ex-factory date forbidden?

  • Transport is cheaper later in the month, when trucks return empty
  • Customers refuse early deliveries
  • It steals capacity and material for an order not yet needed
  • The invoice date would be wrong

Early shipping looks good locally and damages other orders.