8.4 Lean Tools & Theory of Constraints

Process goals: the management context for improvement

Before choosing a lean or constraint-management tool, establish the process goal. A goal states the desired process result in customer, business, and quality terms; a measure tells leaders whether the process is approaching it. CMQ/OE questions often test the difference between a slogan ("improve service") and an operational goal ("reduce complete, accurate response lead time from five days to two days while maintaining at least 98% first-pass accuracy"). The second statement gives direction, a baseline, a target, and a guardrail.

A sound goal-setting cycle is:

1. Identify the customer requirement and the process output that affects it.
2. Establish a baseline using a stable definition and credible data.
3. Set a target, timeframe, owner, and review cadence.
4. Select leading measures that predict performance and lagging measures that confirm the result.
5. Monitor variation and side effects, then adjust the process—not merely the target.

The impact of a goal on quality is not automatically positive. A call center rewarded only for short handling time may transfer customers unnecessarily; a plant rewarded only for utilization may create excess inventory. Managers therefore use a balanced set of quality, delivery, cost, safety, and people measures. When a process measure moves, ask whether the customer-facing output moved in the intended direction.

Lean: create value and remove waste

Lean is a management approach for maximizing customer value with less waste, delay, effort, and variation. It is not a one-time cleanup event or a mandate to reduce headcount. Lean work begins by defining value from the customer’s perspective, mapping the value stream, creating flow, using pull where appropriate, and pursuing continuous improvement.

A useful exam distinction is between value-added work—work that transforms the product or service in a way the customer needs and would recognize as valuable—and necessary non-value-added work, such as a required regulatory record. Pure waste should be reduced or removed. Common waste categories include defects, overproduction, waiting, nonused talent, transportation, inventory, motion, and extra processing. The mnemonic can help recognition, but the manager’s task is to observe the work and verify the actual cause.

Applying the tools

5S makes a workplace self-explaining. After unneeded items are removed, necessary items have marked locations, cleaning doubles as inspection, visual standards establish normal, and sustaining routines make the condition durable. The best evidence of success is not a high audit score alone; it is quicker retrieval, fewer errors, safer work, and faster detection of abnormal conditions.

JIT and kanban support pull. In a pull system, downstream consumption triggers upstream replenishment. A kanban may be a card, bin, electronic signal, or other controlled authorization. The number of signals is deliberately related to demand, replenishment time, container size, and a justified buffer. If demand or lead time changes, the signal design must be reviewed. Pull does not eliminate the need for supplier risk management, capacity planning, or contingency inventory.

A value-stream map includes both material flow and information flow. It usually captures cycle time, changeover time, uptime, queue inventory, lead time, and information triggers across the whole path. A manager should distinguish process time from total lead time: a service may require ten minutes of touch time yet take ten days because it waits in queues and approvals. The future-state map proposes a better flow; an implementation plan assigns experiments, owners, and measures.

SMED means single-minute exchange of die, but its broader lesson is quick changeover. Observe the changeover in detail. Classify each activity as internal (must occur while the equipment is stopped) or external (can occur while it runs). Pre-stage materials, use quick-release devices, standardize settings, and verify the first good unit. Do not remove an inspection that protects the customer merely to claim a shorter changeover.

Poka-yoke is stronger than a warning because it either prevents the mistake or detects it at the point of occurrence. A barcode rule that rejects the wrong part number is preventive; an interlock that stops a cycle when a fixture is incomplete is preventive; an immediate sensor alarm is detection. Error-proofing should address predictable human-process interaction, not blame people for normal slips.

Standard work provides the baseline from which kaizen can occur. It specifies the best current sequence, expected time, work-in-process limit where relevant, key quality points, and safety requirements. Training Within Industry (TWI) job instruction breaks teaching into important steps, key points, and reasons so a supervisor can transfer the method consistently. A kaizen change becomes sustainable only when the revised standard, training, and follow-up are in place.

TPM seeks reliable, capable equipment through planned maintenance and the participation of operators, maintenance specialists, and managers. Overall equipment effectiveness (OEE) combines availability, performance, and quality: OEE = Availability × Performance × Quality. Availability reflects planned-production-time losses from downtime; performance reflects speed losses; quality reflects good units as a share of total units. Use OEE to identify the largest loss category, not as a universal target that justifies making inventory. Productivity is output relative to input; it can improve through less labor, material, time, energy, or capital per useful output, but it must be evaluated with quality and customer outcomes.

Theory of Constraints: improve the chain, not every link

The Theory of Constraints (TOC) views an organization as a system of dependent steps—like a chain. System throughput is limited by its current weakest link, the constraint. Improving a nonconstraint can be useful only if it supports the constraint or removes a different system problem; otherwise it can create excess inventory, cost, and confusion. This is the central contrast between local optimization (making one department look efficient) and system optimization (improving the end-to-end result).

TOC’s five focusing steps are:

1. Identify the system constraint.
2. Exploit it: use its existing capacity effectively; eliminate avoidable idle time, defects, and poor priorities.
3. Subordinate other activities to the constraint’s pace and needs.
4. Elevate the constraint when needed by adding capacity, redesigning work, outsourcing carefully, or changing policy.
5. When the constraint moves, return to step one and avoid inertia.

A physical constraint is tangible: a furnace, specialized analyst, test lab, supplier, limited room, or finite machine capacity. A policy constraint is a rule, measure, assumption, approval sequence, batch-size rule, or incentive that limits performance. For example, a rule requiring a weekly approval meeting can create a longer lead-time constraint than any production machine. Policy constraints are frequently missed because they are embedded in "the way we do things."

TOC uses cause-and-effect thinking to distinguish undesirable effects (UDEs) from a core problem. UDEs are visible symptoms: late deliveries, expediting, overtime, inventory growth, and customer complaints. A core problem is the common cause or conflict that produces multiple UDEs. Do not solve every symptom independently before testing whether one policy or assumption links them. A change may also cause solution deterioration: an initially helpful solution creates negative effects later. For example, adding a large protective queue may stabilize one department while hiding defects and extending customer lead time. Evaluate both immediate relief and downstream consequences.

The terminology above is TOC-specific. On an exam scenario, select the action that protects customer value and total system flow. If a nonconstraint team is fully utilized producing work the constraint cannot process, its local metric may improve while inventory and lead time worsen. The better response is to synchronize releases, protect the constraint from defects and starvation, and measure the whole chain.

Exam scenario

A laboratory has excess sample-preparation capacity but one calibrated instrument with a two-week queue. Managers propose maximizing preparation utilization by releasing every ready sample. This is local optimization: it increases work waiting for the instrument. A TOC response is to identify the instrument as the current constraint, ensure it processes only complete and correct high-priority samples, schedule upstream preparation to its pace, reduce avoidable downtime, and then assess whether capacity elevation is justified. A lean response may additionally map information delays, reduce setup time on the instrument, and error-proof sample preparation.

Use lean to make waste visible and flow smoother; use TOC to determine where system improvement matters most. Both require defined goals, trustworthy measures, employee knowledge, and attention to quality consequences.