Process Analysis, Capacity and Operations Design

Introduction

Industrial engineering improves the way people, materials, information, equipment, and energy are transformed into products and services. Process analysis provides the descriptive and analytical foundation for this work. It maps how work is performed, measures the resources available, identifies constraints, and supports decisions about process choice, capacity, layout, staffing, scheduling, and improvement.

A process is an ordered set of activities that transforms inputs into outputs for a customer. Manufacturing processes transform materials through operations such as machining, forming, assembly, inspection, and packaging. Service processes transform information, customer needs, or requests through activities such as registration, diagnosis, payment, and delivery. Although their outputs differ, both types of process can be analyzed through flow time, capacity, utilization, variability, quality, and cost.

The central managerial challenge is to supply the required output at acceptable cost, quality, speed, flexibility, safety, and sustainability. Excess capacity increases fixed cost, while insufficient capacity creates queues, delays, overtime, lost sales, and poor service. Effective operations design therefore links customer demand with process capability.

Learning Objectives

After studying this chapter, students should be able to:

1. Map and analyze manufacturing and service processes.
2. Distinguish capacity, throughput, cycle time, takt time, lead time, and work-in-process.
3. Identify bottlenecks and calculate process capacity.
4. Apply Little’s Law, break-even analysis, learning curves, line balancing, and overall equipment effectiveness.
5. Select an appropriate process type and evaluate its operational implications.
6. Recommend improvements using quantitative and qualitative evidence.

1. Process Structure and Process Mapping

A process map represents the sequence and relationship of activities. Common symbols include operations, inspections, transport, delay, and storage. In a high-level flowchart, rectangles represent activities, diamonds represent decisions, arrows represent flow, and terminators represent starts or ends. A value-stream map may additionally show information flow, inventory, takt time, and lead times.

A useful process map should identify:

• the customer and the required output;
• suppliers and inputs;
• activities, decisions, and rework loops;
• resources, operators, machines, and information systems;
• waiting, transport, inspection, and storage;
• quality checks and failure points; and
• measures such as time, quantity, cost, and defect rate.

The analyst should first map the process as it actually operates, often called the current state. Designing an ideal future state before observing reality can hide informal work, interruptions, workaround procedures, and undocumented queues.

A process may be analyzed at several levels. A macro process may include order receipt, production, dispatch, and invoicing. A detailed map may show every motion at one workstation. The appropriate level depends on the decision being made. Excessive detail can obscure the main constraint; insufficient detail can conceal rework and delay.

A process flow diagram should distinguish processing time from non-processing time. Processing time changes the product or fulfills the service requirement. Waiting, transport, inspection, and storage may be necessary, but they do not usually add customer value. This distinction supports waste reduction without encouraging unsafe removal of essential controls.

2. Demand, Takt Time, Cycle Time, and Throughput

Takt time

Takt time is the available production time per period divided by the customer demand required in that period:

Available time must exclude planned breaks, meetings, and scheduled maintenance when those periods cannot produce. Units must be consistent, such as seconds per unit or minutes per patient. Takt time is a demand-based pace; it is not the actual speed of a machine.

For example, a plant has 7.5 available hours per day and demand of 450 units per day. Available time is 27,000 seconds, so:

The line must complete one acceptable unit every 60 seconds on average to satisfy demand.

Cycle time

Cycle time is the elapsed time between successive completed units at a process or workstation. It may refer to the observed operating time of a machine, the effective time per unit, or the time between outputs of a line. The definition must be stated because organizations sometimes use “cycle time” to mean processing time for one unit.

If a workstation has an effective available time of 24,000 seconds per day and produces 400 units, its average output cycle time is:

To meet demand reliably, the effective cycle time of the line should generally be no greater than takt time. Capacity losses, changeovers, breakdowns, and variability mean that operating exactly at takt may be risky.

Throughput and capacity

Throughput is the actual rate of completed output. Capacity is the maximum sustainable output rate under stated conditions. Capacity depends on equipment, labor, materials, quality requirements, product mix, maintenance, and available time.

For one resource with effective available time and processing time per unit:

If several sequential resources form a line, the ideal line capacity is determined by the smallest station capacity:

This assumes adequate material availability, no blocking or starvation beyond the modeled effects, and a stable product mix. Actual throughput cannot exceed demand for long periods unless inventory accumulates.

Utilization

Utilization compares actual output or time used with available capacity:

Design capacity is an ideal maximum, whereas effective capacity accounts for realistic losses. A plant may have high utilization but poor delivery performance if demand exceeds effective capacity or if variability creates long queues. High utilization of every resource is not always desirable; a system with no protective capacity becomes sensitive to disruptions.

3. Bottlenecks and Flow Relationships

A bottleneck is a resource or policy that limits the output of the entire system. It may be a machine with the lowest capacity, a skilled employee, an approval step, a test facility, or a supply constraint. The bottleneck can move when product mix, staffing, maintenance, or demand changes.

A practical bottleneck analysis follows these steps:

1. Map the process and measure actual flows.
2. Calculate effective capacity at each activity.
3. Compare capacity with demand and observed queues.
4. Confirm the suspected constraint through direct observation.
5. Exploit existing bottleneck capacity by reducing downtime, defects, unnecessary setups, and interruptions.
6. Elevate capacity through additional labor, equipment, redesign, or subcontracting.
7. Recalculate the system because improvement may move the bottleneck.

Increasing a non-bottleneck’s capacity does not necessarily increase system throughput. For example, doubling the speed of a packaging operation cannot increase output when machining remains the constraint.

Little’s Law

Little’s Law relates average work-in-process, throughput, and average flow time:

where WIP is the average number of units in the system, throughput is average completed units per unit of time, and flow time is the average time a unit spends in the system. Units must be compatible: units, units per hour, and hours.

Rearrangements are:

Little’s Law applies to a stable system over a sufficiently long observation period, where average WIP and throughput are meaningful. It does not require a particular probability distribution, but the system should not be continuously accumulating or losing inventory.

If a repair center averages 120 jobs in progress and completes 30 jobs per day:

Reducing WIP while maintaining throughput reduces average flow time. However, simply restricting releases may reduce throughput if the system becomes starved; improvement must address the causes of delay.

4. Capacity Planning and Break-Even Analysis

Capacity decisions occur at strategic, tactical, and operational levels. Strategic decisions include facility size and technology. Tactical decisions include shifts, staffing, overtime, and subcontracting. Operational decisions include sequencing, dispatching, and preventive maintenance.

Capacity cushions protect against uncertainty:

As a percentage of expected demand:

The appropriate cushion depends on demand variability, supply risk, service expectations, cost of lost sales, and the possibility of flexible labor or equipment.

Break-even analysis

Break-even analysis compares fixed and variable costs. For one product, the break-even volume is:

The denominator is the contribution margin per unit. The model assumes constant selling price and variable cost, a fixed total cost within the relevant range, and all produced units are sold. It is less reliable when products have different margins, capacity constraints, nonlinear costs, or uncertain demand.

Suppose fixed cost is $120,000 per year, selling price is $50 per unit, and variable cost is $30 per unit:

At 8,000 units, operating profit is:

Break-even analysis should be supplemented by sensitivity analysis because small changes in price, demand, labor cost, or yield can materially affect the result.

5. Overall Equipment Effectiveness

Overall Equipment Effectiveness (OEE) evaluates how effectively equipment is used. The conventional components are availability, performance, and quality:

Each component is expressed as a decimal or percentage.

Thus:

The simplified form is valid when the definitions use the same period and counts. OEE is not identical to utilization, and it should not be used to rank departments without considering product mix, planned maintenance, safety, and strategic purpose.

For planned production time of 480 minutes, operating time of 420 minutes, ideal cycle time of 0.5 minutes per unit, total count of 780 units, and good count of 765 units:

• Availability = 420 ÷ 480 = 0.875.
• Performance = (0.5 × 780) ÷ 420 = 0.9286.
• Quality = 765 ÷ 780 = 0.9808.
• OEE = 0.875 × 0.9286 × 0.9808 ≈ 0.798, or 79.8%.

The loss categories guide improvement: breakdowns and setups reduce availability; speed losses and minor stops reduce performance; scrap and rework reduce quality.

6. Line Balancing

Assembly-line balancing assigns tasks to workstations while satisfying precedence relationships and a required cycle time. If daily available time is 27,000 seconds and demand is 450 units, the required line cycle time is 60 seconds per unit.

The theoretical minimum number of stations is:

where t_i is the time of task i, ∑ t_i is total task time, and C is the allowable cycle time. Since stations must be whole numbers, the result is rounded up.

Line efficiency is:

where N is the actual number of stations. Balance delay is:

A feasible assignment must respect precedence constraints and must not assign more than C seconds of work to a station. Common heuristics include ranked positional weight and largest-candidate rules. These methods provide practical solutions but do not always produce the mathematical optimum.

For task times of 20, 15, 25, 30, 10, and 20 seconds, total work is 120 seconds. With a cycle time of 60 seconds:

If precedence permits two stations each containing 60 seconds of work, efficiency is 100%. If precedence forces three stations, efficiency is 120 ÷ (3 × 60) = 66.7%.

Balancing should not be based only on arithmetic. Ergonomics, skill requirements, safety, quality risk, fatigue, model mix, and equipment reliability must also be considered.

7. Learning Curves

Learning curves describe reductions in average labor time as cumulative production increases. A commonly used cumulative-average model is:

where Y is average labor time per unit after producing X units, a is the time for the first unit, and b is the learning exponent:

Here r is the learning rate expressed as a decimal. A 90% learning rate means that average time per unit falls to 90% whenever cumulative output doubles. The model assumes that learning is systematic and that product, methods, workforce, and technology remain sufficiently comparable.

If the first-unit average time is 100 hours and the learning rate is 80%, then after four cumulative units, output has doubled twice, so average time is:

Total time for four units under the cumulative-average interpretation is 4 × 64 = 256 hours. This must not be confused with a unit-time learning curve, which predicts the time of a particular unit rather than the cumulative average. Organizations should state which model is being used.

Learning curves support labor planning, pricing, tender preparation, training plans, and capacity forecasts. They can mislead when learning plateaus, automation changes the process, workers are transferred, or design changes reset the learning process.

8. Process Selection and Operations Design

Process selection matches the volume and variety of demand with an appropriate transformation system. A useful continuum is:

The choice should consider volume, variety, demand stability, product life cycle, customization, capital, labor skills, quality requirements, regulatory controls, and environmental impact. A hybrid or cellular layout may combine flexibility with relatively smooth flow.

Product-process alignment is important. A highly customized product made on a dedicated line may create excessive changeovers and quality problems. Conversely, a stable high-volume product produced in a job shop may incur unnecessary handling and scheduling cost.

Service processes require additional attention to customer contact, demand arrival variability, perceived waiting, privacy, and the fact that the customer may participate in production. Service capacity is often perishable: an unused appointment slot cannot usually be stored for tomorrow.

9. Applied Case: Illustrative Machining Cell

Background

An illustrative manufacturer produces aluminum housings in a machining cell. Customer demand is 360 acceptable units per eight-hour day. The shift includes 30 minutes of planned breaks and 30 minutes of scheduled maintenance, leaving 420 planned production minutes, or 25,200 seconds.

The cell contains three sequential activities:

Problem and analysis

Takt time is:

Deburring is the bottleneck because its capacity is 360 units per day, lower than the other activities. The line has no capacity cushion at the demand level. Any absence, breakdown, scrap, or variation at deburring threatens delivery.

Suppose the cell initially holds an average of 140 units and completes 350 units per day. By Little’s Law, average flow time is:

If deburring is improved so that throughput reaches 360 units per day while WIP is reduced to 108 units, flow time becomes:

The improvement reduces average flow time by 0.1 day, or 24 minutes in a 240-minute operating day, if the day-equivalent convention is used consistently.

Proposed solution

The team studies tool wear, ergonomic motion, batch transfer, and rework. It introduces a deburring fixture, standardized work, preventive tool replacement, and smaller transfer batches. A second operator is trained to cover breaks, but an additional machine is not immediately purchased because the bottleneck’s nominal capacity equals demand and the improvement plan provides a modest protective margin.

The team also establishes daily measures for throughput, queue before deburring, first-pass yield, downtime, and OEE. If demand rises to 400 units per day, the cell must elevate deburring capacity through faster tooling, parallel equipment, overtime, or process redesign; improving CNC machining alone would not solve the constraint.

Outcome and learning points

The case demonstrates that takt is demand-driven, capacity is resource-driven, and throughput is system-limited. It also shows why a process map, capacity table, Little’s Law, and bottleneck confirmation should be used together. Quantitative analysis identifies the constraint, while observation and engineering judgment identify a practical remedy.

10. Implementation, Technology, and Governance

Digital work instructions, machine sensors, manufacturing execution systems, and statistical dashboards can improve visibility of cycle time, downtime, quality, and WIP. Artificial intelligence may assist with predictive maintenance, scheduling, anomaly detection, and demand forecasting. However, automated recommendations require reliable data, model validation, cybersecurity, human oversight, and clear accountability.

Process data may contain employee performance information or customer information. Access should follow legitimate business purposes, least-privilege principles, retention policies, and applicable privacy and labor requirements. Automated monitoring should not become an opaque disciplinary system. Measurements must distinguish system causes from individual behavior and must account for safety, training, product mix, and equipment condition.

Best practice includes:

• define measures and units before collecting data;
• validate timestamps, counts, and downtime codes;
• separate planned from unplanned losses;
• use first-hand observation to interpret dashboards;
• improve the bottleneck before non-constraints;
• standardize successful changes and audit them;
• evaluate safety, quality, cost, delivery, people, and environmental effects together; and
• review assumptions whenever demand, product, technology, or staffing changes.

Challenges and Limitations

Process models simplify reality. Average rates conceal variability, and nominal capacity may be unavailable because of breakdowns, absenteeism, changeovers, material shortages, or quality losses. Little’s Law describes averages and does not explain why a queue exists. Break-even analysis assumes stable costs and prices. OEE can encourage local optimization if managers pursue a high machine score while creating excess inventory. Learning curves may fail after design changes. Line-balancing heuristics may produce feasible but non-optimal assignments.

Consequently, formulas should support—not replace—engineering judgment. Analysts should report definitions, time periods, data quality, assumptions, confidence in measurements, and the consequences of alternative decisions.

Chapter Summary

Process analysis maps the transformation of inputs into outputs and distinguishes value-adding work from waiting, transport, inspection, storage, and rework. Takt time converts customer demand into a required pace, whereas cycle time describes operating performance. Capacity is resource-based, throughput is actual system output, and the bottleneck determines the maximum flow of a sequential system.

Little’s Law connects WIP, throughput, and flow time in a stable system. Break-even analysis links fixed cost, variable cost, price, and volume. OEE decomposes equipment effectiveness into availability, performance, and quality. Line balancing assigns precedence-constrained tasks to stations, while learning curves estimate improvement from cumulative experience. Process selection must align volume, variety, flexibility, technology, labor, and risk.

Effective operations design combines quantitative analysis with observation, safety, quality management, human factors, ethical data governance, and continuous review. The best decision is not necessarily the one that maximizes local utilization; it is the one that improves total system performance and customer value.

Key Terms

Process: An ordered set of activities that transforms inputs into outputs.

Process map: A visual representation of activities, decisions, and flows.

Takt time: Available production time divided by required customer demand.

Cycle time: Time between successive completed units or time required for a defined unit cycle.

Throughput: Actual rate of completed output.

Capacity: Maximum sustainable output under stated conditions.

Utilization: Actual output or time used relative to available capacity.

Bottleneck: Constraint that limits system throughput.

Work-in-process (WIP): Units currently within a process but not yet completed.

Flow time: Average time a unit spends in a process.

Little’s Law: Relationship stating WIP equals throughput multiplied by flow time.

Capacity cushion: Capacity available beyond expected demand.

Break-even volume: Output volume at which total revenue equals total cost.

Contribution margin: Selling price less variable cost per unit.

OEE: Product of availability, performance, and quality.

Line balancing: Assignment of tasks to stations subject to cycle-time and precedence constraints.

Learning curve: Model of improving performance as cumulative experience increases.

Process selection: Choice of a transformation system suited to volume, variety, and strategic requirements.

Standardized work: Documented best-known method for performing a task safely and consistently.

Value-adding activity: Activity that changes the output in a way the customer requires or values.

Review Questions

Short Answer Questions

1. Distinguish takt time from cycle time.
2. What is a bottleneck, and why can it move?
3. State Little’s Law and define its three variables.
4. Name the three components of OEE.
5. Why must a process-selection decision consider both volume and variety?

Long Answer Questions

1. Explain how a process map can reveal waiting, rework, and capacity constraints in a manufacturing process.
2. Describe the assumptions and limitations of break-even analysis.
3. Explain how line balancing is performed and why precedence constraints matter.
4. Compare job-shop, batch, repetitive, and continuous-flow processes.
5. Discuss how digital technologies and AI can improve process analysis while creating governance risks.

Analytical / Application Questions

1. A service center has 90 cases in progress and completes 15 cases per day. Calculate average flow time using Little’s Law and explain what the result means.
2. A production line has 28,800 available seconds per day and demand of 480 units. Three stations require 50, 65, and 55 seconds per unit. Calculate takt time, identify the bottleneck, and determine whether nominal capacity meets demand.
3. An equipment center has availability of 0.90, performance of 0.85, and quality of 0.98. Calculate OEE and recommend which loss category should receive initial attention if improvement resources are limited.

Multiple Choice Questions with Answers

Question 1

A factory has 28,800 available seconds per day and customer demand of 480 units per day. What is the takt time?

A. 45 s/unit

B. 60 s/unit

C. 65 s/unit

D. 75 s/unit

Correct Answer: B. 60 s/unit

Question 2

Which resource determines the capacity of a sequential production line under ideal flow assumptions?

A. The resource with the highest utilization

B. The resource with the lowest capacity

C. The resource with the shortest processing time

D. The final inspection resource only

Correct Answer: B. The resource with the lowest capacity

Question 3

A process has average WIP of 200 units and throughput of 50 units per hour. What is its average flow time?

A. 0.25 hour

B. 2 hours

C. 4 hours

D. 10,000 hours

Correct Answer: C. 4 hours

Question 4

Which expression correctly represents OEE?

A. Availability + Performance + Quality

B. Availability × Performance × Quality

C. Availability ÷ Performance × Quality

D. Performance − Availability − Quality

Correct Answer: B. Availability × Performance × Quality

Question 5

Fixed cost is $80,000, selling price is $40 per unit, and variable cost is $20 per unit. What is the break-even volume?

A. 2,000 units

B. 4,000 units

C. 6,000 units

D. 8,000 units

Correct Answer: B. 4,000 units

Question 6

If total task time is 180 seconds and allowable line cycle time is 60 seconds, what is the theoretical minimum number of stations?

A. 2

B. 3

C. 4

D. 6

Correct Answer: B. 3

Question 7

In a cumulative-average learning-curve model, an 80% learning rate means that average time per unit becomes what when cumulative output doubles?

A. 80% of its previous value

B. 20% of its previous value

C. 120% of its previous value

D. It remains unchanged

Correct Answer: A. 80% of its previous value

Question 8

Which statement about takt time is correct?

A. It is determined primarily by customer demand and available production time.

B. It is always equal to the bottleneck processing time.

C. It measures only machine breakdown duration.

D. It is the same as total lead time.

Correct Answer: A. It is determined primarily by customer demand and available production time.

Question 9

A machine has availability 0.80, performance 0.90, and quality 0.95. What is its OEE?

A. 68.4%

B. 72.0%

C. 76.0%

D. 85.5%

Correct Answer: A. 68.4%

Question 10

Which process type is generally most appropriate for very high volume and very low variety?

A. Project process

B. Job shop

C. Continuous-flow process

D. Custom workshop

Correct Answer: C. Continuous-flow process