Theory of Constraints (TOC): improving by attacking the bottleneck

In one line

The Theory of Constraints (TOC) says that in any chain of interdependent resources there are a few (often just one) that limit the output of the whole system, and that this is where improvement should be concentrated.

What it is

The paper from the Universidad Tecnológica Equinoccial describes Goldratt's TOC as a continuous improvement process based on systems thinking. A constraint (or bottleneck) is any element that limits the system in achieving its goal; it can be a person, a machine, a policy, the lack of a tool. Constraints "are neither negative nor positive, they are a reality." The paper distinguishes two types of resources: the capacity-constrained resource, whose capacity is less than or equal to the demand placed on it, and the non-constraint resource, whose capacity exceeds the demand.

According to Goldratt, what determines the plant's capacity is the capacity of the constraint resource, so you have to balance the flow of all resources to the pace of the bottleneck and make the most of its time, because an hour lost there is an hour lost for the whole system.

The five steps

1. Identify the system's constraint.
2. Exploit it: get the maximum output from it without wasting time.
3. Subordinate everything else to the pace of the constraint.
4. Elevate it: increase its capacity (with investment, if necessary).
5. Go back to step 1: when a constraint is broken, another one appears.

The goal, in a for-profit company, is to make money now and in the future, and it is measured with three indicators: throughput (the rate at which the system generates money through sales), operating expenses and inventories.

Real example

An exercise about a croissant factory in the UNGS book shows the logic with a physical bottleneck. The process goes through mixing (20 dozen per hour), shaping (30), baking (10), cooling (20), tray change (60) and serving (60). The system's cycle time is set by the slowest workstation: baking, at 6 minutes per dozen (10 dozen per hour). The factory's capacity is then 10 dozen per hour, 80 per day with an 8-hour shift, no matter that shaping can do 30. Adding a second oven would double baking capacity to 20 dozen per hour, but the constraint would move to mixing and cooling, which also produce 20: the three stages would end up tied. That is exactly the idea of step 5: once the constraint is broken, another one appears. (A note: the book's printed answer swaps the baking values, giving 10 min per dozen and 6 dozen per hour; we use those from the problem statement, which are consistent with the book's own solution for the second oven.)

The real case in the Ecuadorian paper is a snack factory in Quito, "Productos Alexander", with a customer service index of 89.58%: it was not able to fill all orders. Its diagnosis identified frying as the capacity-constrained resource. It had a capacity of 1.20 kg per minute, which over an 8-hour shift would give 576 kg, but 363.3 kg per day were being processed: frying was producing only about 5 of the 8 hours, a loss of 3 hours of production, which was the cause of the unfilled orders.

To exploit the constraint, the paper proposed reducing setup time (preheating the oil before the raw material arrives, using the logic of SMED), shrinking the transfer batch to 9 kg and scheduling the work so that frying was never starved: 51 batches of 9 kg repeated every 8 minutes. Result projected by the paper (it is an improvement proposal, not a measurement taken after implementing it): production about 21% higher and 100% of orders fulfilled in quantity and on time, with no investment. In TOC terms, net profit would go from 15,333 to 17,392 dollars and return on investment from 33% to 36%. And it anticipates step 5: with frying solved, the constraint would "move to the market."

How to apply it

1. Measure the capacity of each workstation (cycle time or kg per minute) to see which one is the slowest, using the tool from the capacity article.
2. Confirm it by looking at where material piles up before the workstation.
3. Make sure the constraint never waits: material always available, quick changeovers, maintenance first on that equipment.
4. Adjust the pace of the other workstations to that of the constraint (do not produce more than it can absorb).
5. Only if that is not enough, invest in more capacity.

Benefits

• It concentrates effort where the impact is real, avoiding improvements that do not move the result.
• The constraint can often be exploited without investment, as the snack case proposes.
• It gives a simple logic for prioritizing maintenance: a stoppage at the constraint costs more than at any other piece of equipment.

Limitations to keep in mind

• Finding the constraint is not always obvious: in the snack case, packaging looked like the candidate based on nominal capacity, and raw material yield had to be considered to determine that it was frying.
• When a physical constraint is broken, the next one may be a policy or the market, which are harder to move.
• A system with many changing constraints (highly variable demand) requires reviewing the analysis frequently.

In summary

TOC is a prioritization rule: improve where the system is limited. It is complemented by Lean (which eliminates waste) and by OEE (which measures how much of each piece of equipment is used), and its most direct tool is measuring the capacity of the bottleneck.