Glossary

Theory of Constraints glossary

52 terms in plain words, each linked to the free lesson, guide or game that explains it.

Core ideas

Theory of Constraints (TOC)

A management approach developed by Dr. Eliyahu M. Goldratt. Every system has very few constraints that limit its performance; managing around them is the fastest way to improve the whole.

Goal

What the system exists to achieve. For a business: make more money now and in the future, while keeping customers satisfied and people secure. A constraint only makes sense in relation to a goal.

Constraint

Anything that limits a system from getting more of its goal. It can be physical (a machine, a team), the market (not enough demand), or a policy (a rule or measure).

Bottleneck

A resource whose capacity is less than the demand placed on it. Not every busy resource is a bottleneck; many only look overloaded because work is released too early.

Capacity constrained resource (CCR)

A resource that is not a bottleneck on average but can become one if it is badly scheduled. It needs watching and protecting like a constraint.

Policy constraint

A rule, measure or habit that makes the system work against its own goal, such as rewarding machine utilisation or releasing every order on arrival. The most common kind of constraint, and the hardest to see.

Five Focusing Steps

TOC’s process of ongoing improvement: identify the constraint, exploit it, subordinate everything else to it, elevate it, and when it moves, go back to step one without letting inertia become the constraint.

Exploit

Step 2: get the most out of the constraint as it is, before spending money. Stop it losing time to waiting, wrong priorities, breaks, rework and work others could do.

Subordinate

Step 3: make every other part of the system support the decision on how to exploit the constraint, even if that means idle time elsewhere.

Elevate

Step 4: add capacity to the constraint (a machine, a shift, people, a supplier) once exploiting and subordinating have taken it as far as they can.

Inertia

The rules and habits built for an old constraint that stay in place after it has moved. The warning in step 5 of the Five Focusing Steps.

Local optimum

Improving one part of a system on its own measure (efficiency, cost, utilisation) in a way that does not help, or even hurts, the whole.

Operations

Drum-Buffer-Rope (DBR)

TOC’s scheduling method for production. The drum is the constraint’s schedule, the buffer is time that protects it and the shipping dates, and the rope ties material release to the drum.

Drum

The pace-setter in DBR: the constraint’s schedule, or simply the market’s due dates when no internal resource is a constraint (simplified DBR).

Time buffer

Extra lead time placed before the constraint or before shipping, so normal variation is absorbed and orders still arrive on time.

Rope

The rule that releases material to the floor only a buffer’s length before it is needed, so work in progress stays low and the floor is not flooded.

Buffer management

Using how much of each buffer has been consumed to set priorities: green, yellow and red. Red orders get attention first; frequent causes of red show where to improve.

Buffer status

The share of a buffer already used up. Below a third is green, up to two thirds yellow, above that red.

Plant types (V, A, T, I)

A way to classify plants by how material flows: V plants split few materials into many products, A plants assemble many parts into few products, T plants assemble common parts into many variants late, and I plants send every product along the same single route, like a dedicated line.

Work in progress (WIP)

Work that has been started but not finished. Too much WIP lengthens lead times, hides priorities and starves the constraint of the right work.

Supply chain

Make to order / make to stock

Making only against customer orders versus making ahead to hold stock. Which one fits depends on whether customers will wait as long as it takes to make and deliver.

Replenishment (pull)

Refilling stock according to what was actually consumed, usually daily, instead of pushing stock out according to a forecast.

Stock buffer

The target stock at a location: enough to cover the “maximum” demand reasonably expected within one replenishment time, adjusted for how reliable supply is.

Replenishment time

The time from consumption to the replacement arriving: order lead time, production lead time and transport. The factor that drives how much stock is needed.

Dynamic buffer management

Adjusting stock buffers up or down when they stay too long in red or in green, so targets follow real demand instead of a forecast.

Aggregation

Holding stock centrally, closer to the source, because the combined swings of many locations are much smaller than the sum of their separate swings.

Projects

Critical Chain Project Management (CCPM)

TOC’s method for projects. Tasks are planned without padding, safety is pooled into buffers that protect the whole project, and work is controlled by buffer consumption.

Critical chain

The longest chain of dependent tasks in a project, taking both task dependencies and shared resources into account. It sets the project’s length.

Project buffer

One shared block of safety time at the end of the critical chain, usually about half its length, instead of padding inside every task.

Feeding buffer

A smaller buffer where a side chain of tasks joins the critical chain, so delays on the side chain don’t hold up the critical chain.

Bad multitasking

Switching between open tasks before finishing them, usually because too many projects are open. Work time barely changes, but every task’s lead time balloons.

Student syndrome

Starting a task late because there seems to be plenty of time, so the safety is used up before the work begins.

Parkinson’s law

Work expands to fill the time allowed. A task finished early is rarely reported early, so gains are lost while delays are passed on.

Full kit

Everything a task or project needs to be completed without interruption (drawings, approvals, materials, decisions), checked before it starts.

Fever chart

A chart of buffer consumed against critical chain completed. It shows at a glance whether a project is green, yellow or red, and where it is heading.

Finance and measures

Throughput (T)

The rate at which a system makes money through sales: sales revenue minus truly variable costs (mostly raw material).

Investment (I)

Money tied up in the system: inventory, machines, buildings and other things bought to be used to make money.

Operating expense (OE)

All the money spent to turn investment into throughput: wages, salaries, rent, power, maintenance.

Throughput Accounting

Decision-making with T, I and OE instead of allocated product costs. Net profit = T − OE; return on investment = (T − OE) ÷ I.

T per constraint unit (T/CU)

Throughput earned per minute (or hour) of the constraint. When capacity is limited, it ranks products and orders better than margin per unit.

Cost world / throughput world

Two ways of managing: the cost world judges every part by its own cost and efficiency; the throughput world judges every decision by its effect on the whole system’s throughput.

Thinking Process

Thinking Processes

TOC’s set of logic tools for answering three questions: what to change, what to change to, and how to cause the change.

Undesirable effect (UDE)

A severe, recurring problem, defined against the system’s goal. A symptom of a deeper cause, and the starting point of an analysis.

Conflict cloud

A five-box diagram of a dilemma: a common objective, two needs, and two opposing actions. Surfacing the assumptions behind its arrows shows how to break the conflict without compromise.

Assumption

The reason we believe an arrow in a cloud or tree holds. Challenging a false assumption is how a conflict is broken.

Injection

A new idea or action that makes a false assumption untrue, and so dissolves a conflict or produces a desired effect.

Core conflict

The one conflict behind many undesirable effects. Found by building clouds for several UDEs and generalising what they share.

Current reality tree (CRT)

A cause-and-effect map, read “if… then…”, that connects the undesirable effects to the few root causes behind them.

Future reality tree (FRT)

A cause-and-effect map that tests whether the proposed injections will really produce the desired effects, before anything is changed.

Negative branch

A chain of cause and effect showing how a proposed solution could lead to something bad, so it can be trimmed with an extra injection before it happens.

Prerequisite tree (PRT)

A map of the obstacles to implementing a solution and the intermediate objectives that overcome them, in the order they must be reached.

Layers of resistance

The stages people go through before accepting a change, from disagreeing about the problem to unspoken fears. Addressed in order, they turn resistance into buy-in.

The Theory of Constraints and its tools were developed by Dr. Eliyahu M. Goldratt. These definitions are in my own words.