Operations, Capacity and Quality Management

Introduction and context

Operations management converts resources into products and services. In supply chains, this conversion cannot be managed in isolation: purchasing, production, warehousing, transport, sales and customer service are linked through shared capacity, inventories, schedules and quality requirements. A decision that improves one activity may damage total supply-chain performance. For example, producing large batches may reduce unit setup cost but increase inventory, obsolescence and response time.

Advanced supply-chain managers therefore coordinate four related decisions:

1. Capacity: how much output the system can produce and when.
2. Production and scheduling: which jobs are performed, in what sequence and at what time.
3. Process improvement: how waste, variation and delays are reduced.
4. Quality management: how conformance, reliability and customer value are designed and controlled.

This chapter integrates bottleneck analysis, lean systems, the Theory of Constraints (TOC), Six Sigma, total quality management (TQM) and continuous-improvement methods. The central principle is that performance should be improved across the whole flow, not merely within individual departments.

Learning objectives

By the end of this chapter, the learner should be able to:

• distinguish design, effective and actual capacity;
• calculate utilisation, efficiency, throughput and cycle time;
• identify bottlenecks and explain their supply-chain consequences;
• apply the five focusing steps of TOC;
• use lean principles and value-stream thinking to remove waste;
• explain Six Sigma, DMAIC, process capability and control charts;
• coordinate aggregate planning, scheduling and quality decisions;
• evaluate improvement proposals using operational and financial measures;
• recognise implementation, ethical, technological and governance risks.

1. Capacity management

1.1 Capacity concepts

Design capacity is the theoretical maximum output under ideal conditions. Effective capacity is the realistic maximum after allowing for planned maintenance, breaks, product mix, changeovers, staffing and policy constraints. Actual output is what the operation really produces.

Two useful measures are:

Output and capacity must use the same unit and time period, such as units per shift or tonnes per week. Utilisation may be below 100% because of demand shortages or downtime. Efficiency can exceed 100% only if the stated effective-capacity standard is outdated or the calculation is inconsistent; it should normally be investigated rather than accepted automatically.

Capacity is also classified by planning horizon:

1.2 Capacity planning alternatives

A level strategy maintains relatively stable production and uses inventory or backlogs to absorb demand variation. It supports workforce stability but may create holding cost or customer delay. A chase strategy adjusts production to demand through hiring, layoffs, overtime, temporary labour or outsourcing. It reduces inventory but can increase cost, quality risk and workforce instability. A mixed strategy combines these methods.

Capacity decisions should consider more than average demand. Managers should assess demand variability, product mix, setup time, absenteeism, maintenance, supplier reliability, safety stock and the cost of lost sales. A resource that appears to have adequate average capacity may fail during peaks.

1.3 Capacity cushions and yield

A capacity cushion is capacity deliberately kept above expected demand to absorb uncertainty. It is particularly important where demand is volatile, replenishment lead time is long, or service failures are costly. Excess capacity, however, creates fixed cost and may encourage inefficient processes.

When output is affected by scrap or rework, required input is greater than saleable output. If the first-pass yield is 95%, producing 1,000 good units requires:

Thus, required input = 1,000 ÷ 0.95 = 1,052.63, so at least 1,053 units must enter the process, assuming units are indivisible. This calculation assumes the yield is stable and that rework is not counted as first-pass output.

2. Flow, bottlenecks and throughput

2.1 Bottleneck analysis

A bottleneck is a resource whose available capacity is less than, or closest to, the workload required by the system. It controls the maximum flow rate. Bottlenecks may be machines, labour skills, inspection stations, warehouse docks, transport lanes, information approvals or suppliers.

For each resource, calculate the workload required:

The summation covers all products using the resource. Compare required workload with available time:

A ratio above 1 indicates overload. The largest ratio is a strong bottleneck candidate, although variability and queueing may make a resource with a slightly lower ratio operationally more restrictive.

Worked example: identifying a bottleneck

A weekly demand mix consists of 400 units of Product A and 200 units of Product B. Their processing times are:

Required workloads are:

• Cutting: (400 × 3) + (200 × 5) = 2,200 minutes; load ratio = 2,200 ÷ 2,400 = 0.917.
• Assembly: (400 × 6) + (200 × 8) = 4,000 minutes; load ratio = 4,000 ÷ 2,400 = 1.667.
• Inspection: (400 × 2) + (200 × 3) = 1,400 minutes; load ratio = 1,400 ÷ 1,200 = 1.167.

Assembly is the primary bottleneck. Inspection is also overloaded, but increasing inspection capacity alone will not raise total output while assembly remains more severely constrained.

2.2 Little’s Law and flow relationships

For a stable system, Little’s Law relates work in process, throughput and flow time:

where L is average work in process (units), λ is average throughput (units per time), and W is average time in the system (time per unit). The relationship assumes a stable long-run system in which average arrival and departure rates are equal.

For example, if a process averages 120 units per day and each unit spends 2.5 days in the system, average work in process is 120 × 2.5 = 300 units. Reducing work in process or improving flow time is not merely an inventory exercise; it can improve responsiveness and expose process problems.

2.3 Theory of Constraints

TOC views every system as constrained by at least one limiting factor. Its five focusing steps are:

1. Identify the constraint.
2. Exploit it by ensuring that its time is used on valuable productive work.
3. Subordinate other activities to the constraint’s rhythm.
4. Elevate the constraint by adding capacity or changing the process.
5. If the constraint moves, return to step one and prevent inertia from becoming the next constraint.

Examples of exploitation include reducing setups at the bottleneck, preventing quality defects from reaching it, ensuring materials are available before work begins and assigning skilled operators to critical tasks. Subordination may require upstream processes to slow down rather than create excess inventory.

TOC distinguishes throughput (the rate at which the system generates money through sales), inventory (money invested in items intended for sale) and operating expense (money spent to convert inventory into throughput). This perspective discourages local efficiency measures that increase production without increasing sales.

3. Production planning and scheduling

Aggregate planning converts demand forecasts into medium-term decisions about production rate, workforce, inventory, overtime and subcontracting. The master production schedule translates this plan into end-item quantities and dates. Material requirements planning then determines component requirements, planned orders and release dates, subject to lead times and lot-sizing rules.

Scheduling allocates scarce resources to specific jobs. Common priority rules include:

• First come, first served (FCFS): perceived as fair but not necessarily efficient.
• Shortest processing time (SPT): often reduces average flow time, but may delay long jobs.
• Earliest due date (EDD): prioritises urgent delivery dates.
• Critical ratio: compares remaining time to required processing time.

A ratio below 1 indicates that a job is behind the time required for on-time completion, assuming no interruptions. Rules should not be applied mechanically: customer priorities, setup families, safety, contractual commitments and bottleneck protection may override them.

Finite-capacity scheduling is preferable when resource limits are binding. Infinite-capacity plans may promise work that the plant, warehouse or supplier cannot actually perform. Schedule stability is also important: excessive expediting creates nervousness, overtime and quality risk.

4. Lean operations and flow improvement

Lean management seeks to deliver customer value with minimal waste. Value is defined from the customer’s perspective, not merely by internal activity. The classic waste categories are transportation, inventory, motion, waiting, overproduction, overprocessing, defects and underused human capability.

Lean principles include:

1. specify value;
2. map the value stream;
3. create continuous flow;
4. establish pull based on demand;
5. pursue perfection through continuous improvement.

A value-stream map distinguishes value-adding time from waiting, inspection, movement and storage. Takt time expresses the pace required to meet demand:

If a shift has 420 available minutes and demand is 140 units, takt time is 3 minutes per unit. Takt is a demand-based planning rate, not necessarily the actual cycle time. To meet demand, the effective cycle time at each required workstation should generally be no greater than takt time, after considering losses and variability.

Kanban is a pull-control mechanism that authorises replenishment when downstream consumption occurs. A simplified card calculation is:

The result is rounded up to a whole container. This formula assumes relatively stable demand and lead time; highly variable or unreliable processes require careful parameter review. Excessive kanban cards conceal problems by increasing inventory.

Lean tools include 5S, standardised work, visual management, quick changeover (SMED), mistake-proofing (poka-yoke), preventive maintenance, cellular layouts and kaizen. Lean is not simply a cost-reduction programme. Removing buffers without improving reliability can make the system fragile.

5. Quality management

Quality has two complementary meanings: conformance to requirements and fitness for customer use. A product can conform to an internal specification yet fail to satisfy customer needs if the specification is wrong.

5.1 Prevention, appraisal and failure costs

The cost of quality is commonly grouped as:

Quality management emphasises prevention and process capability rather than relying only on final inspection. Inspection detects defects; it does not by itself remove their causes.

TQM treats quality as a responsibility of leadership and every process. Its principles include customer focus, process management, employee involvement, evidence-based decisions, supplier relationships and continual improvement. Deming’s systems perspective is especially relevant to supply chains: variation may be created by purchasing policies, forecast changes, incentive systems or poorly designed specifications rather than by operators alone.

5.2 Six Sigma and DMAIC

Six Sigma uses structured problem solving and statistical thinking to reduce defects and variation. For existing processes, DMAIC means Define, Measure, Analyse, Improve and Control.

• Define: clarify the customer, problem, scope, goal and critical-to-quality requirements.
• Measure: establish operational definitions, measurement-system reliability and baseline performance.
• Analyse: identify and verify root causes using data.
• Improve: design, test and implement countermeasures.
• Control: standardise the improved process and monitor its performance.

A useful capability index is:

where USL and LSL are the upper and lower specification limits, and σ is the within-process standard deviation. Cₚ measures potential capability if the process is centred. It does not measure centring.

where μ is the process mean. These indices assume a stable, approximately normal process and a meaningful estimate of within-process variation. A commonly used organisational benchmark is Cₚₖ ≥ 1.33, but the appropriate threshold depends on risk, industry and customer requirements. Capability indices should not be calculated for an unstable process without first addressing special causes.

5.3 Statistical process control

A control chart separates common-cause variation, inherent to the process, from special-cause variation, associated with an identifiable change. For a mean chart with known or appropriately estimated standard deviation:

For sample means, the standard error is σ ÷ √n, so limits are based on μ ± 3σ ÷ √n. Control limits are not specification limits: control limits describe process behaviour, whereas specifications describe customer or engineering requirements. A stable process can consistently produce the wrong output, and an unstable process may sometimes fall within specifications.

Root-cause tools include the Pareto chart, cause-and-effect diagram, 5 Whys, process mapping, fault-tree analysis and failure mode and effects analysis (FMEA). FMEA prioritises potential failures by severity, occurrence and detectability; numerical risk-priority scores are screening aids, not substitutes for engineering judgement.

6. Coordinating operations, capacity and quality

Local optimisation is a major source of supply-chain failure. Purchasing may seek low unit prices through large lots; production may seek long runs; warehousing may seek full truckloads; sales may promise short lead times. The result can be high inventory and poor service.

Coordination requires common metrics, including:

• on-time-in-full delivery;
• schedule adherence;
• throughput and flow time;
• forecast error and demand bias;
• capacity utilisation and bottleneck load;
• first-pass yield and defects per unit;
• overall equipment effectiveness (OEE);
• inventory turns and days of supply;
• total landed cost and cost of poor quality.

OEE is often represented as:

Availability is actual running time relative to planned production time; performance compares ideal cycle output with actual running output; quality is good output divided by total output. The three factors must be expressed as proportions, not percentages, during multiplication. OEE is useful for diagnosing equipment losses, but it can encourage local optimisation if maximising machine utilisation conflicts with flow or customer demand.

A practical improvement cycle is: establish the customer requirement, map the end-to-end process, identify the system constraint, measure baseline performance, verify causes, test a focused countermeasure, standardise the result and review whether the constraint has shifted.

7. Applied case: Northstar Appliances

The following is a fictional illustrative case.

Background

Northstar Appliances assembles 600 refrigerators per week. Customers complain about late deliveries, while finished-goods inventory is rising. Management plans to purchase another assembly line because assembly is reported to be “busy all the time.”

Problem and analysis

A cross-functional team measures the process. Cutting has 2,800 minutes of weekly demand, assembly has 3,500 minutes, testing has 1,450 minutes, and dispatch has 2,100 minutes. Available times are respectively 3,200, 3,600, 1,600 and 2,800 minutes. The load ratios are 0.875, 0.972, 0.906 and 0.750. Assembly is the closest constraint, but testing has high variability and frequent rework. A value-stream map shows that units wait two days before testing because batches are released from assembly.

Quality data show that 6% of units fail first-pass testing. The failures are concentrated in a wiring operation performed before assembly. The existing performance measure rewards assembly for completing large batches, even when testing is not ready.

Approach and solution

The team applies TOC and lean methods. It protects assembly time by ensuring complete kits are available, reduces changeover through standardised tooling and schedules small release batches matched to testing capacity. It introduces a point-of-use wiring check, a simple poka-yoke connector and a daily Pareto review of defects. A cross-functional schedule gives testing priority for the product mix that has the highest contribution per constrained minute, subject to customer commitments.

A pilot uses a control chart for first-pass yield and measures flow time, on-time delivery and work in process. The organisation does not immediately buy a new line; it first exploits the existing constraint and verifies demand.

Outcome

After eight weeks, first-pass yield rises from 94% to 98%, average work in process falls, and testing no longer receives large unpredictable batches. On-time delivery improves. The assembly load remains high, but its effective throughput increases because less time is consumed by rework and waiting. The next constraint becomes a specialised compressor test, so the improvement cycle begins again.

Learning points

The case illustrates five principles. First, the busiest resource is not automatically the only constraint. Second, quality losses consume capacity. Third, batch release can create delay even when processing time is short. Fourth, local utilisation targets can conflict with total flow. Finally, capital investment should follow verified constraint analysis rather than precede it.

8. Technology, governance and contemporary issues

Advanced planning and scheduling systems, manufacturing execution systems, warehouse-management systems, sensors and digital twins can improve visibility and synchronisation. Machine-learning forecasts may identify demand patterns, while predictive maintenance models can anticipate equipment failure. However, technology does not correct poor master data, unstable processes or unclear accountability.

AI-supported decisions require governance. Managers should validate training data, monitor forecast bias, explain important recommendations, protect confidential supplier and employee information, and retain human authority over safety, quality release and employment decisions. Automated scheduling may unintentionally disadvantage certain customers or workers if its objectives are poorly defined.

Environmental and social quality are increasingly operational requirements. Capacity decisions should consider energy, water, emissions, hazardous materials, worker safety and product end-of-life. Supplier audits must be risk-based and respectful of labour rights; passing an audit should not be treated as proof that all ethical risks have disappeared.

Challenges and limitations

Operations models simplify reality. Demand is uncertain, processing times vary, workers learn, machines fail and suppliers may not respond as planned. Utilisation measures can hide queues; average data can hide volatility; capability indices can mislead when processes are unstable; and lean techniques can increase exposure to disruption when buffers are removed without resilience measures.

Managers should therefore combine quantitative analysis with observation, worker knowledge and scenario testing. Improvement targets must avoid encouraging defect concealment, unsafe speed, excessive overtime or manipulation of data. The best system balances flow, quality, cost, resilience, sustainability and employee wellbeing.

Best-practice checklist

• Define capacity in common units and realistic time periods.
• Separate design, effective and actual capacity.
• Identify constraints using workload, variability and observed queues.
• Protect the bottleneck from starvation, defects and unnecessary setups.
• Schedule with finite capacity and communicate changes clearly.
• Use pull systems only when processes are sufficiently stable.
• Measure quality at the source and investigate root causes.
• Confirm that measurement systems are reliable before analysing variation.
• Distinguish control limits from customer specifications.
• Evaluate improvements using end-to-end outcomes, not departmental utilisation alone.
• Standardise successful changes and reassess the constraint.
• Govern digital and AI tools through validation, security, transparency and human oversight.

Chapter Summary

Capacity management aligns resources with demand across strategic, tactical and operational horizons. Utilisation and efficiency must be interpreted carefully because effective capacity includes realistic losses. Bottleneck analysis identifies the resource that limits system flow, while Little’s Law explains the relationship between work in process, throughput and time. TOC provides a disciplined sequence for exploiting and elevating constraints.

Lean management removes activities that do not create customer value and uses flow, pull, takt, standard work and mistake-proofing to improve responsiveness. Quality management moves beyond final inspection toward prevention, process capability and organisation-wide responsibility. Six Sigma’s DMAIC cycle structures evidence-based improvement, while control charts distinguish common and special causes. Capacity, scheduling and quality must be managed as an integrated system, supported by appropriate technology and governed with attention to safety, ethics, resilience and sustainability.

Key Terms

Actual capacity: Output actually achieved in a specified period.

Bottleneck: Resource that constrains total system throughput.

Capacity cushion: Intentional capacity above expected demand for uncertainty.

Control chart: Time-ordered chart used to distinguish process variation causes.

Design capacity: Theoretical maximum output under ideal conditions.

DMAIC: Define, Measure, Analyse, Improve and Control.

Effective capacity: Realistic capacity after planned losses and operating conditions.

First-pass yield: Proportion of units completed correctly without rework.

Lean management: Systematic pursuit of customer value with minimal waste.

Little’s Law: Relationship L = λW among work in process, throughput and flow time.

OEE: Product of availability, performance and quality.

Poka-yoke: Mistake-proofing method that prevents or detects error.

Process capability: Ability of a stable process to meet specifications.

Pull system: Control system in which downstream consumption authorises replenishment.

Six Sigma: Data-based approach to reducing process variation and defects.

Takt time: Available production time divided by customer demand.

Theory of Constraints: Management approach focused on the system’s limiting constraint.

Throughput: Rate at which the system generates saleable output.

Total quality management: Organisation-wide approach to customer-focused quality and continual improvement.

Value-stream mapping: Visual analysis of material and information flow from demand to delivery.

Review Questions

Short Answer Questions

1. Distinguish design capacity, effective capacity and actual output.
2. What is a bottleneck, and why does it influence total system throughput?
3. State the five focusing steps of the Theory of Constraints.
4. Explain the difference between control limits and specification limits.
5. What is the purpose of takt time in a lean system?

Long Answer Questions

1. Explain how aggregate capacity planning, master scheduling and finite-capacity scheduling should be coordinated in a supply chain.
2. Compare a level strategy, a chase strategy and a mixed capacity strategy, including their operational risks.
3. Discuss how lean management and Six Sigma complement and differ from one another.
4. Explain the DMAIC method and the conditions under which process capability indices are meaningful.
5. Evaluate the argument that maximising equipment utilisation always improves supply-chain performance.

Analytical / Application Questions

1. A process has 2,000 minutes of available time. Product A requires 4 minutes and demand is 300 units; Product B requires 6 minutes and demand is 180 units. Calculate the required workload and load ratio. Is the resource overloaded?
2. A plant has 360 available minutes per shift and demand of 120 units. Calculate takt time. Explain two reasons why actual cycle time may still fail to meet customer demand even if the average cycle time equals takt time.
3. A process has a specification range of 90 to 110, a mean of 104 and a standard deviation of 2. Calculate Cₚ and Cₚₖ, state the assumptions required, and interpret the difference between the two indices.

Multiple Choice Questions with Answers

Question 1

Which measure compares actual output with effective capacity?

A. Utilisation

B. Efficiency

C. Takt time

D. First-pass yield

Correct Answer: B. Efficiency

Question 2

A resource requires 4,000 minutes of work and has 2,400 minutes available. What is its load ratio?

A. 0.60

B. 1.20

C. 1.67

D. 2.40

Correct Answer: C. 1.67

Question 3

Which TOC step follows identification of the constraint?

A. Elevate it immediately

B. Exploit it

C. Subordinate the customer

D. Eliminate all inventory

Correct Answer: B. Exploit it

Question 4

If throughput is 120 units per day and average flow time is 2.5 days, Little’s Law gives average work in process of:

A. 48 units

B. 122.5 units

C. 300 units

D. 480 units

Correct Answer: C. 300 units

Question 5

If available production time is 420 minutes and demand is 140 units, takt time is:

A. 2 minutes per unit

B. 3 minutes per unit

C. 140 minutes per unit

D. 560 minutes per unit

Correct Answer: B. 3 minutes per unit

Question 6

Which activity is an example of prevention cost?

A. Warranty replacement

B. Final inspection

C. Supplier-quality training

D. Scrapping defective output

Correct Answer: C. Supplier-quality training

Question 7

In DMAIC, which phase verifies the causes of a performance problem?

A. Define

B. Measure

C. Analyse

D. Control

Correct Answer: C. Analyse

Question 8

A stable process has USL = 110, LSL = 90 and σ = 2. What is Cp?

A. 0.83

B. 1.00

C. 1.67

D. 3.33

Correct Answer: C. 1.67

Question 9

Which statement correctly distinguishes control limits from specification limits?

A. Both are set only by customers

B. Control limits describe process behaviour; specifications describe requirements

C. Specifications change automatically when variation changes

D. Control limits determine product design requirements

Correct Answer: B. Control limits describe process behaviour; specifications describe requirements

Question 10

What is the principal purpose of a pull system?

A. To authorise replenishment from downstream consumption

B. To maximise upstream batch size

C. To keep every machine continuously busy

D. To eliminate all safety stock regardless of risk

Correct Answer: A. To authorise replenishment from downstream consumption