Abolish Efficiency?

Henry Ford abolished efficiency in his assembly lines. He limited the space between workstations. There is no space in assembly lines to build WIP inventory. Every workstation must work at the constraint’s production rate. Non-constraint workstations stop or slow down to match the speed of the constraint.

Taichi Ohno (Toyota) did something similar with his Toyota Production System. He introduced Kanban cards. A workstation that consumes one unit of inventory sends a Kanban card to signal the prior station should produce 1 unit. The result is the same as Ford’s. Workstations stop or slow down to match the constraints rate.

Ohno gradually removed Kanban cards to reduce inventory more and more. He did this until problems (rocks disturbing the flow) appear. ‘Rocks’ are removed (solved). A good flow resumes with less stock. As inventory continues to reduce ‘rocks’ appear again and again. The production process improves continually.

What is Ohno’s ultimate result? Is it that only the constraint resource has a significant queue? Do other resources have no queue or only a very small one?

Eli Goldratt’s Drum-Buffer-Rope process abolishes efficiency similarly. The constraint is the drum that calls the beat and the production rate. The rope signals to order release that the constraint has finished another unit. The rope causes the release of the next work order. It restricts the work in the process to the drumbeat. Workstations produce at the constraint’s rate.

The buffer in the Drum-Buffer-Rope process protects the constraint. The constraint never stops due to a lack of work orders. (A lack of customer orders can cause it to stop. Such a situation indicates the constraint may not be the factory. It is the market.)

The buffer in the Drum-Buffer-Rope process protects the constraint. The constraint never stops due to a lack of work orders. (A lack of customer orders can cause it to stop. Such a situation indicates the constraint may not be the factory. It is the market.)

Ford’s assembly lines are flawed. They do not protect the constraint because there is no buffer. This is not a problem if their quality is perfect. The Japanese automobile factories all have excellent quality. This high quality has been transferred to all relevant factories.

However, even excellent quality does not exclude the occasional breakdown. The number of workstations in an assembly line increases the probability a line will sometimes be stopped. How much do automobile manufacturers lose due to equipment downtime? According to Forbes (Feb 22, 2022) …

The average manufacturer?confronts 800 hours of equipment downtime per year — more than 15 hours per week. That downtime comes at a cost, and it isn’t cheap. For example, the average automotive manufacturer loses $22,000 per minute when the production line stops. That quickly adds up. Overall, unplanned downtime costs industrial manufacturers as much as?$50 billion a year.

Forbes does not tell us how they calculate these numbers. Let’s assume they are correct or at least very, very large. The 800 hours are a bit more than 1 month of 24-hour days of production. The 800 hours are about 8.5% of capacity.

What should Ford have done to protect an assembly line’s constraint? Eli Goldratt suggested the following:

Split the assembly line before and after the constraint. Do this to keep the constraint running when a work centre breaks down. Before the constraint, put 3 unfinished cars. After the constraint, put a space big enough to hold 3 cars. (3 is enough because of the excellent process and product quality.

These buffers ensure the constraint continues to work during breakdowns. With this measure, Goldratt achieved a 10% increase in production.

How much is 8.5% worth? Forbes told us the average manufacturer loses 800 hours annually due to equipment downtime. How much is this 8.5% worth?

Let’s assume a 100’000’000€ business, with 50% variable cost and a 6% return on sales. Fixed cost will then be 44% of 100’000’000€. If we implement the recommendation, there is no added cost (bar variable cost and the line rearrangement). Sales increase to 108’500’000 and profit increases from 6’000’000€ to 10’250’000€. That is a 70% jump.

Forbes recommended 6 actions to improve downtime. They did not recommend the buffers.

If you have such an environment, isn’t it better to implement buffers to get an immediate return? Improve your quality but how close to 6-Sigma can you get and at what cost for that?

Important is to minimize work-in-process inventory but not at the expense of constraint capacity. Less WIP speeds up production. Lead time is shorter. Less WIP makes operators’ jobs easier – fewer orders to deal with. Continual improvement at your constraint will increase your output and bottom even more.

NB. The strategy outlined above works in automated lines. Sometimes these are designed and built without an appropriate buffer and space at the constraint. It may be expensive to change an existing production line. However, determine the size of the potential benefit.