Resource Data, Standards and Capacity Modeling

Introduction

Capacity Requirement Planning (CRP) converts the production plan, master production schedule, and material requirements plan into a time-phased estimate of the load placed on work centres. Its quality depends less on sophisticated software than on the reliability of the resource data supplied to it. A routing that omits setup time, a work centre calendar that ignores planned maintenance, or a labor standard that assumes an unavailable skill can produce a mathematically precise but operationally false plan.

This chapter develops a resource model for CRP. It examines the structure of operation times, labor and machine standards, calendars and shifts, skills, tooling, subcontracting, alternative routings, data accuracy, and the limitations of standard time. The central principle is that capacity is not merely the number of machines or employees; it is the usable capability of a resource over a defined time horizon.

Learning Objectives

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

1. Distinguish setup, run, queue, move, wait, and maintenance times.
2. Construct labor and machine capacity standards with appropriate units and assumptions.
3. Model shifts, breaks, calendars, skills, tooling, subcontracting, and alternate routings.
4. Calculate gross and effective capacity and compare it with workload.
5. Diagnose data-quality problems and assess the limitations of standard-time systems.
6. Recommend governance and improvement practices for reliable CRP data.

1. The Resource Model in CRP

A resource model represents the relationships among products, operations, work centres, people, equipment, tools, calendars, and time standards. A typical routing operation contains:

• an operation identifier and sequence;
• a work centre or resource group;
• setup time;
• run time or production rate;
• labor requirements and skill qualifications;
• machine and tooling requirements;
• queue, move, and wait assumptions;
• yield, scrap, and rework information;
• possible alternate resources or routings.

The model is usually maintained in an enterprise resource planning or manufacturing execution system. However, the system does not remove the need for engineering judgment. It stores assumptions and applies rules; it cannot determine whether an observed cycle-time change represents process improvement, product-mix change, operator learning, or poor data collection.

A useful conceptual distinction is:

• Required capacity: the time demanded by planned work.
• Available capacity: the scheduled time made available by the resource calendar.
• Effective capacity: available capacity after realistic losses such as breaks, planned maintenance, minor stoppages, and expected efficiency.
• Demonstrated capacity: capacity supported by historical evidence under defined conditions.

The distinction matters because a nominal eight-hour shift is not normally eight hours of productive processing.

Figure: How resource data and calendars are transformed into a CRP load-versus-capacity view.

2. Components of Manufacturing Time

Setup time

Setup time is the elapsed time required to prepare a resource for a job. It may include cleaning, fixture installation, tool change, program loading, first-piece adjustment, inspection, and documentation. Setup may be job-dependent, such as changing from one product family to another, or resource-dependent, such as a routine warm-up.

If a batch contains Q units, setup time is incurred once for the batch, whereas run time is normally incurred for each unit. For a batch with no scrap, operation time is:

where S is setup time in hours per batch, Q is batch quantity, and r is run time in hours per unit. The average time per unit is:

These expressions assume one setup per batch and a constant run time. If multiple setups, learning effects, or product-dependent rates exist, the routing must represent them explicitly.

Run or processing time

Run time is the time during which the operation is actively performed. It can be expressed as a time per unit or as a rate. If the production rate is R units per hour, then:

This relationship applies only when the rate is stable and measured for the same product, method, and quality condition. A machine cycle time may not equal labor time: a machine can process automatically while an operator performs another task or services another machine.

Queue, move, wait, and inspection time

Queue time is time spent waiting before processing. Move time is time spent transporting material between locations. Wait time may arise from material availability, approval, batching, or downstream synchronization. Inspection time can be labor time, machine time, or both. These times influence lead time and work-in-process, but not all consume the bottleneck resource in the same way. CRP should therefore identify the resource charged by each time category rather than simply adding all elapsed time to machine load.

Maintenance and downtime

Planned preventive maintenance, calibration, cleaning, and changeover losses reduce resource availability. Corrective maintenance is uncertain and should generally be represented through historical availability assumptions or scenario analysis rather than hidden inside an arbitrary processing standard.

3. Labor and Machine Standards

A standard is a documented expectation for time, method, output, or resource usage under specified conditions. It is not automatically a target for every individual worker. A credible standard states its scope: product, method, equipment, batch size, quality requirement, operator qualification, and date of approval.

Labor standards

Labor time may be divided into direct setup labor, direct run labor, indirect support, inspection, material handling, and normal allowances. A labor standard must specify whether it represents:

• touch time by one person;
• elapsed operation time;
• total person-hours;
• or time charged to a particular skill pool.

If three operators work for 20 minutes, elapsed labor time is 20 minutes but labor content is 60 person-minutes. Confusing these measures causes major CRP errors.

Machine standards

Machine standards include cycle time, setup occupancy, warm-up, tool replacement, cleaning, and planned downtime where appropriate. A machine may be assigned to one operation while its operator serves several machines. The labor and machine loads should consequently be calculated separately.

Suppose a batch of 120 units requires 0.08 machine-hours per unit and 1.5 setup hours. The machine load is:

If two qualified operators each work for the full setup and run period, the direct labor load is 22.2 person-hours, not 11.1 person-hours. If one operator supervises two identical machines simultaneously, the labor model must use the actual attendance pattern rather than multiplying machine time blindly.

Allowances and performance ratings

Traditional work measurement may begin with an observed time, adjust it by a performance rating, and add an allowance. In abstract form:

where T_observed is observed time, P is a performance-rating factor expressed relative to normal performance, and A is an allowance fraction. This formulation assumes the rating and allowance are defined consistently. Some organizations instead use a factor based on available working time; the method must not be mixed with the additive allowance convention.

For example, if observed time is 4.0 minutes, the rating factor is 1.10, and allowance is 15%:

The result is a standard under the stated method, not a universal truth. Ratings are vulnerable to observer bias, and allowances must reflect fatigue, unavoidable delay, personal needs, and the operating environment without becoming a concealed inefficiency factor.

4. Capacity Equations and Calendars

For a resource with n identical units, h scheduled hours per unit per period, and d planned downtime hours per unit, gross scheduled capacity is:

Scheduled usable capacity before performance losses is:

If availability A_v, performance P_v, and quality Q_v are expressed as fractions between 0 and 1, effective output-equivalent capacity can be represented as:

This product is appropriate when the three factors are defined independently and sequentially. If planned downtime has already been removed from h, it must not be removed again. Availability and performance definitions also vary across organizations; governance must prevent double counting.

For workload L and available capacity C, utilization is:

Utilization is dimensionless, although it is often expressed as a percentage. A value above 1 indicates overload under the model assumptions. A utilization near 1 may also be risky because variability produces queues even when average load does not exceed capacity.

Calendars and shifts

A resource calendar should identify working days, shift start and end times, breaks, holidays, planned maintenance, training, and exceptional closures. It should also distinguish the calendars of people and machines. A machine may be available continuously while a certified operator is available only on one shift.

Consider two machines operating one 8-hour shift for five days. Gross weekly capacity is 2 × 8 × 5 = 80 machine-hours. If each machine has 0.5 hours of planned maintenance per day and a 0.5-hour unpaid break does not consume machine time, machine availability is:

2 × (8 − 0.5) × 5 = 75 machine-hours. The labor calendar must be calculated independently if the break applies to operators.

5. Skills, Tooling and Constrained Resources

Capacity is constrained by more than equipment. A work centre may have adequate machine hours but insufficient welders, programmers, inspectors, or licensed technicians. Skill matrices should record qualification level, authorization expiry, training status, and permitted operations. A generic “labor” category is inadequate for specialist CRP.

Tooling can be modeled as:

• an unlimited shared resource, when availability is not restrictive;
• a finite concurrent resource, when only a fixed number of tools can be used at one time;
• a consumable, when usage reduces inventory;
• or a setup-dependent requirement, when a tool is needed only during preparation.

Tool conflicts are often invisible in traditional work-centre reports. A plan may show sufficient machine capacity while two operations require the same special fixture simultaneously. Tool calendars, preventive maintenance, calibration status, and location should be integrated where the tool is a genuine constraint.

6. Subcontracting and Alternative Routings

Subcontracting transfers some processing to an external supplier. The model should include supplier capacity, minimum batch size, transport time, queue time, quality release, purchase lead time, and cost. Subcontracting is not infinite capacity; it is capacity with contractual, logistical, and quality constraints.

Alternative routings provide different feasible paths for the same operation. A routing decision can be based on priority, cost, capability, available capacity, due date, or a hybrid rule. Alternatives must be technically valid: a faster machine may not have the required tolerance, certification, material capability, or tooling.

A routing model should distinguish:

• parallel alternatives, where one of several resources performs the operation;
• sequential alternatives, where all listed operations are required;
• conditional alternatives, selected by product, order, quality, or capacity condition;
• rework routings, activated only after a nonconformance.

When alternative resources have different standards, CRP should calculate the load on the selected resource rather than averaging all alternatives. Average rates conceal bottlenecks and make accountability difficult.

7. Data Accuracy and Standard-Time Limitations

Resource data has several dimensions of quality:

A practical validation process combines document review, shop-floor observation, time studies, machine data, operator interviews, and comparison of planned versus actual results. The purpose is not to replace judgment with averages but to understand variation and its causes.

Standard time has important limitations:

1. Variation: material, temperature, product mix, tool wear, and operator experience change actual time.
2. Learning: newly introduced products may improve rapidly, making an early standard obsolete.
3. Method dependence: a standard is valid only for the method and equipment studied.
4. Behavioral effects: standards used as punitive targets can encourage under-reporting, rushing, or quality avoidance.
5. Aggregation error: family averages may be useful for rough planning but misleading for detailed scheduling.
6. Rare events: breakdowns, engineering changes, and exceptional rework may be absent from a small sample.
7. Metric ambiguity: efficiency, utilization, availability, and productivity are not interchangeable.

A statistically informed approach records sample size, dispersion, product family, and operating context. For observed times t₁, t₂, …, t_n, the sample mean is:

and the sample standard deviation is:

These statistics describe the sample; they do not prove that future times follow the same distribution. Where variability is material, planners may use percentile estimates, scenario ranges, or simulation rather than a single deterministic time.

8. Illustrative Case: Capacity Model for an Electronics Assembly Cell

Background

An illustrative electronics manufacturer assembles a controller in batches of 100 units. The final test operation can be performed on either Testers A or B. Both testers are technically qualified, but Tester B is slower and requires a specialized technician.

Problem

The weekly schedule contains four batches. Engineering data lists a 2-hour setup and 0.12 machine-hours per unit. The production supervisor reports that the first-piece approval adds 0.5 hour and that Tester A requires 1 hour of preventive maintenance each week. The original CRP report excludes both items.

Analysis

The corrected load per batch is:

For four batches:

Tester A has one 10-hour shift for five days, with 1 hour of planned maintenance weekly:

Tester B has 45 available hours, but its processing rate is 0.15 machine-hours per unit. If all four batches use B, its load is:

The 2.5-hour setup includes the same first-piece approval. Thus B alone cannot meet the requirement. A mixed plan is feasible only if the routing and technician calendars permit it. For example, three batches on A require 43.5 hours, leaving 5.5 hours of A capacity. One batch on B requires 17.5 hours, leaving substantial B capacity. The mix therefore fits machine capacity, but the planner must confirm that the specialized technician is available for the selected B shift and that transfer, test-program, and quality-release times are included.

Approach and outcome

The company revised the routing to separate setup, first-piece approval, and run time; linked tester capability to product revision; and introduced weekly maintenance calendars. It also created a skill requirement for the technician. The revised CRP report showed that the original plan was feasible only through an alternate routing, not through Tester A alone.

Learning points

The case demonstrates that small omitted times can materially change feasibility. It also shows why a capacity model must represent machine hours, labor skills, tooling or programs, and routing alternatives together. The correct managerial response is not simply to increase the nominal standard; it is to identify the actual resource and calendar responsible for each time.

9. Governance, Technology and Best Practice

Resource data should have an owner, approval date, revision history, source evidence, and review frequency. Engineering normally owns methods and standards; operations validates practical applicability; maintenance owns equipment availability assumptions; human resources or operations training owns qualification records; planning governs use in CRP. Cross-functional ownership prevents one department from optimizing its local measure at the expense of system accuracy.

Modern systems can collect machine states, operator confirmations, sensor signals, and execution timestamps. Analytics and AI can identify abnormal cycle times, classify downtime, estimate remaining maintenance windows, and flag standards that diverge from actual performance. These tools should support—not silently replace—engineering review. Automated recommendations require explainability, data lineage, cybersecurity controls, and protection against feedback loops in which a biased historical schedule becomes the basis for future standards.

Recommended practice is to:

1. define every time element and unit;
2. separate machine, labor, and elapsed-time measures;
3. maintain effective-dated calendars and routing revisions;
4. validate standards after engineering or tooling changes;
5. monitor planned-versus-actual time by product family and resource;
6. model scarce skills, tools, and suppliers explicitly;
7. use ranges or scenarios where variability is decision-relevant;
8. review standards with operators rather than imposing unexplained targets;
9. audit master data periodically;
10. treat quality and safety requirements as capacity conditions, not optional losses.

Ethically, standards must not be used to demand unsafe pace or to conceal inadequate staffing. A capacity model is a decision-support instrument, and its outputs should be interpreted with attention to worker welfare, product quality, legal working-time limits, and responsible supplier relationships.

Chapter Summary

Reliable CRP begins with a reliable resource model. Operation time must be decomposed into setup, run, queue, move, wait, inspection, and maintenance elements, with each element assigned to the resource that actually consumes it. Labor and machine standards require clear units, methods, assumptions, and revision control. Setup is generally batch-related, while run time is generally quantity-related; confusing the two distorts both lot sizing and capacity.

Calendars convert nominal resources into usable capacity by representing shifts, breaks, maintenance, holidays, and exceptional events. Skills, tooling, subcontractors, and alternate routings are genuine capacity constraints and should not be hidden in generic work-centre data. Effective capacity calculations are useful only when availability, performance, and quality definitions are consistent and losses are not counted twice.

Standard times are valuable planning abstractions, but they are method-dependent and subject to variation, learning, data bias, and behavioral consequences. Professional practice therefore combines standards with observation, historical evidence, statistical analysis, governance, and shop-floor validation. The strongest capacity model is not the most detailed model; it is the model whose assumptions are visible, technically valid, maintained, and suitable for the decision being made.

Key Terms

Resource model: Structured representation of resources, routings, calendars, skills, tools, and time requirements.

Capacity requirement planning: Process of comparing time-phased resource workload with available capacity.

Setup time: Time required to prepare a resource for a job or batch.

Run time: Active processing time associated with producing a unit or batch.

Queue time: Time a job waits before processing.

Move time: Time spent transporting material between operations.

Machine standard: Documented expected equipment time under specified conditions.

Labor standard: Documented expected person-time under a specified method and performance condition.

Person-hour: One hour of work by one person.

Resource calendar: Schedule defining when a resource is available or unavailable.

Gross capacity: Nominal scheduled capacity before deductions.

Effective capacity: Capacity after defined availability, performance, and quality losses.

Utilization: Workload divided by available capacity.

Skill matrix: Record of personnel qualifications and authorized operations.

Tooling constraint: Limitation caused by insufficient fixtures, tools, programs, or gauges.

Alternative routing: A technically valid alternative path for completing an operation.

Subcontracting: Use of an external supplier to perform a production activity.

Standard time: Approved expected time for a defined method and condition.

Data lineage: Traceable record of a data value’s source, ownership, and revisions.

Review Questions

Short Answer Questions

1. What is the difference between machine time and person-time?
2. Why is setup time usually associated with a batch rather than each unit?
3. Identify three items that should appear in a resource calendar.
4. What is an alternative routing?
5. Why can a standard time become obsolete?

Long Answer Questions

1. Explain how setup, run, queue, move, wait, inspection, and maintenance times should be represented in CRP.
2. Discuss the differences between gross, scheduled, effective, and demonstrated capacity.
3. Explain how skills and tooling can create bottlenecks even when machine hours are available.
4. Evaluate the strengths and weaknesses of using standard time for managerial capacity decisions.
5. Design a governance process for maintaining routing, calendar, and resource master data.

Analytical / Application Questions

1. A batch of 250 units requires 3 hours of setup and 0.06 machine-hours per unit. Calculate the total machine load and average machine time per unit. State the assumptions used.
2. A work centre has two machines, each scheduled for 9 hours per day over four days. Each machine has 0.5 hours of planned maintenance per day. Calculate weekly scheduled capacity and explain which additional information is needed to estimate effective capacity.
3. A product can be processed on Resource X or Y. X has 30 available hours and requires 8 hours per batch; Y has 20 available hours and requires 10 hours per batch. Determine whether five batches can be assigned using both resources, ignoring setup interactions, and propose a feasible allocation if one exists.

Multiple Choice Questions with Answers

Question 1

A batch of 80 units has a setup time of 2 hours and a run time of 0.10 machine-hours per unit. What is the total machine load?

A. 8 hours

B. 10 hours

C. 10.5 hours

D. 82 hours

Correct Answer: B. 10 hours

Question 2

Which time element is normally incurred once per batch rather than once per unit?

A. Run time

B. Setup time

C. Unit inspection time

D. Processing cycle time

Correct Answer: B. Setup time

Question 3

Three operators each work for 30 minutes on the same operation. What is the labor content?

A. 0.5 person-hour

B. 1.0 person-hour

C. 1.5 person-hours

D. 3.0 person-hours

Correct Answer: C. 1.5 person-hours

Question 4

A resource has 50 hours of workload and 60 hours of available capacity. Its utilization is:

A. 0.83

B. 1.20

C. 10 hours

D. 120% capacity shortage

Correct Answer: A. 0.83

Question 5

Which item most directly represents a skill constraint?

A. A machine’s electrical rating

B. An operator’s authorization to perform certified inspection

C. A batch’s material weight

D. The distance between two buildings

Correct Answer: B. An operator’s authorization to perform certified inspection

Question 6

If production rate is 20 units per hour, the corresponding run time per unit is:

A. 0.02 hours

B. 0.05 hours

C. 0.20 hours

D. 20 hours

Correct Answer: B. 0.05 hours

Question 7

Which practice prevents double counting in effective-capacity calculations?

A. Ignoring planned maintenance

B. Applying the same downtime deduction in both the calendar and availability factor

C. Ensuring that losses already removed from scheduled hours are not removed again

D. Treating every resource as infinitely available

Correct Answer: C. Ensuring that losses already removed from scheduled hours are not removed again

Question 8

What is the principal purpose of an alternative routing in CRP?

A. To remove all quality requirements

B. To provide a technically valid alternate resource or process path

C. To convert labor hours into material quantity

D. To eliminate the need for calendars

Correct Answer: B. To provide a technically valid alternate resource or process path

Question 9

Which statement about standard time is most accurate?

A. It is valid regardless of method or equipment

B. It is a guaranteed prediction of every future cycle

C. It is an approved expectation under defined conditions and assumptions

D. It should always be used as an individual punitive target

Correct Answer: C. It is an approved expectation under defined conditions and assumptions

Question 10

A standard-time study observes 4 minutes, uses a performance factor of 1.10, and an allowance of 15% under the additive convention. What standard time results?

A. 4.40 minutes

B. 4.60 minutes

C. 5.06 minutes

D. 5.50 minutes

Correct Answer: C. 5.06 minutes