Schedule and Resource Management

Introduction

Schedule and resource management converts project scope into a feasible pattern of work over time. It answers four connected questions: What must be done? In what order? How long will it take? With which people, equipment, materials, and budget? A schedule is therefore more than a list of dates. It is a model of dependencies, capacity, uncertainty, decision points, and delivery commitments.

In operations and project environments, schedule quality strongly influences cost, customer service, safety, quality, cash flow, and organizational credibility. An unrealistic schedule may appear efficient during planning but generate overtime, rework, expediting, burnout, and contractual disputes during execution. Conversely, excessive contingency can protect a date while reducing competitiveness. Effective schedule management balances technical logic, resource constraints, uncertainty, and stakeholder commitments.

This chapter develops an integrated approach to activity definition, network analysis, critical-path and critical-chain methods, estimation, resource allocation, leveling, optimization, and schedule control.

Learning Objectives

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

1. Define activities and construct logically valid network diagrams.
2. Calculate early and late dates, float, and the critical path.
3. Compare critical-path and critical-chain approaches.
4. Select and combine estimation techniques under uncertainty.
5. Allocate, level, and smooth resources without violating project logic.
6. Evaluate schedule compression and optimization decisions.
7. Establish a realistic baseline and control schedule performance.
8. Recognize ethical, technological, and governance issues in scheduling.

1. From Scope to Schedule Model

The schedule-management process normally proceeds through progressive elaboration:

1. Scope decomposition: translate deliverables into work packages.
2. Activity definition: identify executable work, inspections, approvals, procurement, and handoffs.
3. Sequencing: establish logical relationships and constraints.
4. Duration estimation: estimate work under stated assumptions and resource conditions.
5. Resource planning: identify labor, equipment, materials, facilities, and specialist capacity.
6. Schedule development: calculate dates, resolve conflicts, and establish a baseline.
7. Schedule control: measure actual progress, forecast completion, and approve changes.

A useful activity record includes an identifier, description, predecessor and successor relationships, responsible owner, resource requirements, duration estimate, assumptions, constraints, acceptance criteria, and status. Activities should be sufficiently small to estimate and control, but not so small that administration overwhelms management. A practical test is whether the activity has one clear owner, a recognizable completion condition, and a duration that can be monitored meaningfully.

Terminology

• Activity: a defined unit of project work.
• Milestone: a zero-duration event used to mark a decision, approval, or deliverable.
• Dependency: a logical relationship between activities.
• Predecessor: an activity that must occur before another activity.
• Successor: an activity that follows another activity.
• Calendar duration: elapsed time, including non-working periods where applicable.
• Effort: labor time required; for example, 40 person-hours.
• Float: scheduling flexibility without violating a specified date.
• Baseline: an approved time-phased plan against which performance is measured.

Effort and duration must not be confused. If one analyst requires 40 hours, the effort is 40 person-hours. With one available analyst, duration may be five eight-hour days. With two suitably skilled analysts, duration may decrease, but not necessarily to 2.5 days because coordination, learning, and task structure impose limits.

2. Activity Relationships and Network Diagrams

The most common dependency representation is the precedence diagramming method, in which activities are nodes and arrows represent relationships. Four relationship types are important:

A relationship may include a lead or lag. A two-day lag after concrete placement before finishing begins is a technological waiting period. A lead allows controlled overlap, such as beginning documentation before testing is entirely complete. Leads should not be used to conceal missing logic.

Dependencies may be:

• Mandatory: imposed by technology, physics, law, or contract.
• Discretionary: preferred sequencing based on good practice.
• External: involving parties outside the project team.
• Internal: controlled within the project.

Network logic must be acyclic for ordinary critical-path calculations: work should not eventually depend on itself. Every activity should have a justified predecessor and successor, except the project start and finish. Excessive use of date constraints such as “must start on” can hide the true network logic and make forecasts unreliable.

3. Critical-Path Method

The critical path method (CPM) calculates the longest logical path through a network, where “longest” means greatest total duration, not greatest number of activities. Activities on the critical path determine the earliest possible project completion under the modeled assumptions.

For a simple finish-to-start network with no lags, the forward pass is:

where ES_i is the early start of activity i, EF_i is its early finish, and d_i is its duration, all measured in consistent time units. For an activity with predecessors:

where j ranges over immediate predecessors. The maximum is required because the activity cannot begin until all predecessors have finished.

The backward pass begins with the project completion time T:

Here LF_i and LS_i are late finish and late start, and k ranges over immediate successors. Total float is:

The formulas assume finish-to-start relationships, deterministic durations, a consistent time calendar, and no unmodeled resource restrictions. For other relationship types, software or manual calculations must apply the specific relationship and lag correctly.

Worked Example: CPM

Consider the following activities:

Forward-pass results:

• A: ES 0, EF 3
• B: ES 3, EF 7
• C: ES 3, EF 8
• D: ES max(7, 8) = 8, EF 14
• E: ES 14, EF 17
• F: ES 7, EF 9
• G: ES max(17, 9) = 17, EF 18

The project’s earliest completion is 18 days. The controlling path is A–C–D–E–G, with duration 3 + 5 + 6 + 3 + 1 = 18 days. Activity F is not initially critical. Its late dates, calculated backward from day 18, give LF = 17 and LS = 15 for F, so its total float is 15 − 7 = 8 days. Float belongs to the network model, not automatically to the person performing the activity; consuming it may affect downstream flexibility.

A project can have multiple critical paths. Near-critical paths deserve active monitoring because a small delay can make them critical. Criticality may also change when actual durations, calendars, resources, or approved scope change.

4. Estimation Under Uncertainty

Duration estimation should state its basis. Common methods include:

• Analogous estimation: uses a comparable completed project; fast but dependent on genuine similarity.
• Parametric estimation: uses a measurable rate, such as installation hours per unit.
• Bottom-up estimation: estimates detailed work and aggregates it; often more accurate but more expensive.
• Three-point estimation: uses optimistic, most-likely, and pessimistic cases.
• Expert judgment: valuable when supported by evidence and calibrated feedback.
• Simulation: models many possible outcomes when uncertainty and correlations matter.

For three-point estimates, the triangular expected duration is:

The beta-weighted estimate traditionally associated with PERT is:

where a is optimistic duration, m is most-likely duration, and b is pessimistic duration. These formulas assume the three points are meaningful and that a smooth distribution between them is a reasonable approximation. They do not remove estimation bias or justify false precision.

For the PERT approximation, an often-used activity standard deviation is:

This is an approximation, not a universal law. Project-level uncertainty cannot generally be found by simply adding standard deviations unless assumptions about dependence are satisfied. If activity durations are treated as independent, project variance along a fixed path is approximately:

Correlation, alternative paths, resource contention, and changing critical paths require more advanced analysis, such as Monte Carlo simulation.

For example, if an activity has a = 4 days, m = 7 days, and b = 16 days, the PERT estimate is (4 + 28 + 16) ÷ 6 = 8 days, while the approximate standard deviation is (16 − 4) ÷ 6 = 2 days. The estimate should be documented as “8 days under stated assumptions,” not presented as a guaranteed completion time.

5. Resources and the Critical Chain

Resources include people, equipment, materials, facilities, information, and working capital. A schedule is feasible only when it respects both precedence and resource capacity. Resource requirements should be time-phased so that managers can see when demand exceeds supply.

The critical chain method (CCM) extends precedence-based scheduling by considering resource dependencies. It commonly uses aggressive but achievable activity estimates, removes much of the safety embedded in individual tasks, and places aggregated protection in a project buffer. Feeding buffers protect the critical chain from delays on non-critical paths. Buffer consumption is monitored as a control signal.

CPM and CCM answer different questions:

CCM is not simply “CPM with shorter durations.” It requires credible resource calendars, a policy against unnecessary multitasking, and management commitment to buffer-based control. Its assumptions may be unsuitable where work is highly exploratory, priorities change hourly, or task durations cannot be meaningfully compressed.

6. Resource Allocation, Leveling, and Smoothing

Resource allocation assigns available capacity to activities. Decisions should consider skill, authorization, safety, location, equipment compatibility, and workload—not merely numerical availability.

Resource leveling changes activity dates, and sometimes the completion date, to resolve over-allocation. It may delay non-critical work into its float or extend the project if capacity is insufficient. Resource smoothing adjusts activities within available float to reduce peaks while preserving the required completion date where possible.

A useful resource profile identifies demand by period. If a specialist is required for 8 hours on Monday and 8 hours on Tuesday, but only 8 hours are available each day, the plan is feasible. If three activities each require that specialist for 8 hours on Monday, the schedule has 24 hours of demand against 8 hours of capacity. Leveling must delay, split, resequence, outsource, or add capacity; simply labeling the activities “parallel” does not solve the conflict.

Resource decisions can be framed as constrained optimization. Let x_it indicate whether activity i is scheduled in period t, and let r_ik be the requirement of resource type k. A capacity constraint has the general form:

where C_kt is available capacity of resource k in period t. This representation is simplified: real models must incorporate activity duration, calendars, precedence, setup times, alternative skills, and nonpreemptive work.

Good leveling priorities are explicit. Possible rules include minimum total project duration, protection of contractual milestones, maximum use of critical-path float, continuity of specialist work, safety, or minimum overtime. A schedule is not “optimized” until the objective and constraints are stated.

7. Schedule Compression and Optimization

When a deadline is fixed, managers may use:

• Crashing: adding resources, overtime, specialist support, or premium procurement to reduce duration. It increases direct cost and may create coordination or quality risk.
• Fast tracking: performing activities in parallel that were previously sequential. It increases rework and interface risk and should be used only where overlap is technically defensible.
• Scope or specification adjustment: removing or deferring lower-value work through formal change control.
• Method or technology change: redesigning work to reduce effort or waiting time.
• Calendar change: adding shifts or working days, subject to law, fatigue, safety, and stakeholder constraints.

Crashing should begin with activities on the current critical path that offer credible time reduction. Because criticality can shift, the network must be recalculated after each material intervention. A simple crash comparison uses incremental cost per day saved:

The formula applies when normal and crash points are reasonably represented by a linear trade-off. It should not be used when cost or productivity changes nonlinearly. For example, normal cost of $50,000 at 10 days and crash cost of $62,000 at 8 days gives a cost slope of $6,000 per day saved. This is only one decision input; safety, quality, probability of success, and downstream effects also matter.

Schedule optimization is multi-objective. A useful managerial objective may minimize total cost subject to a promised completion date, quality requirements, safety limits, resource capacities, and regulatory conditions. The cheapest schedule is not necessarily the best schedule if it creates unacceptable risk.

8. Schedule Baselines and Control

A baseline is approved only after scope, assumptions, resources, calendars, and constraints have been reviewed. It should include time-phased work, milestones, dependencies, and responsibility. Baseline changes require documented authorization; otherwise, teams may silently re-plan and destroy performance transparency.

Control involves:

1. Recording actual starts, finishes, remaining duration, and completed quantities.
2. Updating logic when approved changes or discovered dependencies arise.
3. Comparing forecast dates with the baseline.
4. Investigating variance causes rather than merely reporting lateness.
5. Recalculating critical and near-critical paths.
6. Escalating decisions requiring scope, budget, or contract authority.

Earned value can connect schedule and cost, although it should not replace network analysis. Schedule variance is:

and the schedule performance index is:

where EV is earned value and PV is planned value, both in monetary or normalized value units. An SPI below 1 indicates less planned value has been earned than scheduled at the measurement date. However, earned-value schedule indicators can approach 1 near project completion even when activities remain late; therefore, milestone forecasts and critical-path analysis remain necessary.

9. Case Study: Regional Distribution-Center Automation

Illustrative case. A retailer is installing automated sorting equipment in an operating distribution center. The contractual launch date is 20 weeks away. The project includes design, equipment procurement, electrical installation, software configuration, integration testing, operator training, and cutover.

Problem

The initial plan schedules electrical installation and software configuration in parallel. Resource review shows that the same controls engineer is assigned to both activities during weeks 9–11. The procurement activity also has a two-week supplier uncertainty range. Management is concerned that the nominal 20-week plan is unrealistic.

Analysis

The team decomposes the work into deliverable-based activities and identifies mandatory dependencies. Equipment installation cannot begin until delivery and site preparation are complete. Integration testing requires both electrical completion and software configuration. Training can begin after a stable software demonstration but must finish before cutover.

The network calculation produces a 19-week precedence path, but resource loading reveals an over-allocation of the controls engineer. Leveling the configuration activity by two weeks extends the resource-constrained completion to 21 weeks. The project therefore has a resource-constrained critical chain longer than the original CPM path.

A three-point estimate for procurement is 8, 10, and 14 weeks. The PERT estimate is (8 + 40 + 14) ÷ 6 = 10.33 weeks. The team does not promise 10.33 weeks as a certainty; it uses the uncertainty range to create procurement protection and identifies an alternate supplier.

Approach and solution

The project manager adopts the following actions:

• Reassigns a qualified automation engineer to part of configuration work.
• Uses a two-week feeding buffer for procurement uncertainty.
• Begins training with a controlled demonstration rather than waiting for full integration.
• Adds a formal interface review before testing.
• Prohibits uncontrolled multitasking by the controls engineer.
• Establishes weekly forecast reviews of the resource-constrained critical chain.
• Defines a change threshold requiring steering-committee approval if forecast completion exceeds 20 weeks.

Fast tracking is rejected for electrical work and live operations because safety risk is unacceptable. Limited crashing is approved for configuration because the additional engineer has verified competence and the cost is lower than the contractual delay exposure.

Outcome

The revised plan forecasts a 20-week launch with protected procurement uncertainty, no unresolved specialist over-allocation, and a documented contingency route. The plan is not risk-free; it is more credible because its assumptions and resource limits are explicit.

Learning points

1. A precedence-only schedule may be infeasible under resource constraints.
2. Resource leveling can reveal a new critical chain.
3. Buffers are useful only when their purpose and consumption rules are defined.
4. Safety and operational continuity can legitimately outweigh schedule compression.
5. A realistic forecast is more valuable than an optimistic commitment.

10. Governance, Ethics, and Technology

Schedule decisions have ethical consequences. Managers should not manipulate status dates, hide float consumption, pressure staff into unsafe overtime, or report a milestone as complete when acceptance criteria are unmet. Suppliers and customers deserve forecasts that distinguish committed dates, estimates, and confidence levels.

Governance should define who can approve baseline changes, who owns assumptions, how schedule risk is escalated, and how contractual milestones are evidenced. Labor law, working-time regulation, procurement rules, environmental restrictions, and safety requirements may constrain calendars and resource choices.

Modern scheduling platforms support dependency validation, resource histograms, scenario comparison, critical-path recalculation, and integration with enterprise systems. Artificial intelligence can assist with anomaly detection, duration forecasting, pattern recognition, and alternative schedule generation. It should not be treated as an autonomous authority. Historical data may encode past bias, omitted risk, or conditions that no longer apply. AI-generated schedules require explainable assumptions, human validation, access controls, version history, and protection of commercially sensitive data.

Contemporary practice increasingly emphasizes probabilistic schedules, integrated cost and schedule risk, rolling-wave planning, collaborative planning, and digital twins. These approaches improve visibility but do not eliminate uncertainty. Better software cannot compensate for poor scope definition, weak logic, missing resource data, or ineffective governance.

Best-Practice Checklist

• Define completion criteria for every activity.
• Use logical relationships rather than arbitrary imposed dates.
• Separate effort, duration, and elapsed time.
• Document calendars, assumptions, estimates, and uncertainty.
• Check resource capacity before promising a completion date.
• Monitor near-critical paths as well as the current critical path.
• Recalculate after approved changes or major actual deviations.
• Use crashing and fast tracking selectively and transparently.
• Protect safety, quality, and legal compliance during compression.
• Maintain an approved baseline and an auditable change history.
• Present forecasts as evidence-based estimates, not guarantees.

Chapter Summary

Schedule and resource management integrates scope decomposition, activity definition, dependency logic, duration estimation, resource capacity, and control. Network diagrams reveal the sequence of work; CPM identifies the longest logical path and float; resource leveling tests whether that plan is feasible in practice. Critical-chain methods incorporate resource constraints and aggregate protection into buffers. Three-point estimation expresses uncertainty, while simulation may be necessary when risks are correlated or critical paths can change.

A credible schedule is not simply the shortest schedule. It is a transparent, resourced, risk-aware, and governable plan. Compression techniques should be evaluated through cost, quality, safety, probability of success, and changing criticality. During execution, actual progress and forecast dates must be compared with an approved baseline, while ethical reporting and formal change control preserve trust.

Key Terms

Activity: A defined unit of project work.

Baseline: An approved time-phased plan used for comparison.

Critical chain: The longest resource-constrained sequence of activities.

Critical path: The longest logical path determining earliest project completion under modeled assumptions.

Crashing: Adding cost or capacity to shorten activity duration.

Dependency: A logical relationship between activities.

Duration: Elapsed time required for an activity.

Earned value: Budgeted value of work actually performed.

Fast tracking: Overlapping activities that were previously sequential.

Float: Time flexibility available without violating a specified constraint.

Lag: Deliberate waiting time between related activities.

Lead: Permitted overlap between related activities.

Milestone: A significant zero-duration event or decision point.

Network diagram: A graphical representation of activities and logical relationships.

Resource leveling: Adjusting dates or completion to resolve resource over-allocation.

Resource smoothing: Adjusting activities within float to reduce resource variation.

Schedule variance: Difference between earned value and planned value.

Three-point estimate: Estimate based on optimistic, most-likely, and pessimistic cases.

Total float: Difference between late and early dates for an activity.

Review Questions

Short Answer Questions

1. Distinguish effort from duration and explain why the distinction matters.
2. What is the difference between mandatory and discretionary dependency?
3. Define total float and explain how it is calculated in a simple finish-to-start network.
4. Distinguish resource leveling from resource smoothing.
5. Why should near-critical paths be monitored?

Long Answer Questions

1. Explain the forward-pass and backward-pass procedures in CPM, including their assumptions.
2. Compare CPM and critical-chain methods in terms of constraints, protection, and control.
3. Discuss the strengths and limitations of analogous, bottom-up, parametric, and three-point estimation.
4. Explain how crashing and fast tracking affect cost, risk, quality, and schedule logic.
5. Design a governance process for approving, baselining, and controlling a major project schedule.

Analytical / Application Questions

1. Construct a network for a project in which procurement, design, installation, testing, and training have overlapping and sequential relationships. Calculate the critical path and float using stated durations.
2. A project’s critical path is 24 weeks, but a shared specialist creates a three-week resource conflict. Analyze at least three responses and recommend one using cost, risk, and stakeholder criteria.
3. An activity has optimistic, most-likely, and pessimistic durations of 6, 9, and 18 days. Calculate its PERT estimate and approximate standard deviation, then explain why the result should not be treated as a guaranteed duration.

Multiple Choice Questions with Answers

Question 1

In a simple finish-to-start network, the early start of an activity with multiple predecessors is determined by:

A. The minimum predecessor early finish

B. The average predecessor duration

C. The maximum predecessor early finish

D. The latest project milestone

Correct Answer: C. The maximum predecessor early finish

Question 2

Using the PERT estimate E = (a + 4m + b) ÷ 6, what is the estimate when a = 4, m = 7, and b = 16 days?

A. 7 days

B. 8 days

C. 9 days

D. 10 days

Correct Answer: B. 8 days

Question 3

Which statement best describes the critical path?

A. The path with the greatest number of activities

B. The path with the most expensive activities

C. The longest logical path determining earliest completion

D. The path containing all resource-intensive activities

Correct Answer: C. The longest logical path determining earliest completion

Question 4

Resource leveling may:

A. Resolve over-allocation while extending the project completion date

B. Remove all project uncertainty

C. Guarantee that the critical path remains unchanged

D. Reduce every activity duration

Correct Answer: A. Resolve over-allocation while extending the project completion date

Question 5

Which technique overlaps activities that were previously planned sequentially?

A. Crashing

B. Fast tracking

C. Resource smoothing

D. Analogous estimation

Correct Answer: B. Fast tracking

Question 6

If an activity has early start 5 days and late start 9 days, its total float is:

A. 4 days

B. 5 days

C. 9 days

D. 14 days

Correct Answer: A. 4 days

Question 7

In critical-chain management, a project buffer primarily:

A. Replaces all activity dependencies

B. Protects the promised completion date from chain delays

C. Represents additional project scope

D. Measures earned value directly

Correct Answer: B. Protects the promised completion date from chain delays

Question 8

A crash cost is $62,000 at 8 days and normal cost is $50,000 at 10 days. The linear cost slope is:

A. $2,000 per day

B. $4,000 per day

C. $6,000 per day

D. $12,000 per day

Correct Answer: C. $6,000 per day

Question 9

Which is an ethical schedule-control practice?

A. Changing status dates to conceal lateness

B. Reporting incomplete work as accepted

C. Documenting assumptions and approved baseline changes

D. Requiring unsafe overtime without disclosure

Correct Answer: C. Documenting assumptions and approved baseline changes

Question 10

An SPI below 1.0 generally indicates that:

A. Earned value is less than planned value at the measurement date

B. Actual cost is less than budgeted cost

C. The project has no critical path

D. All activities are ahead of schedule

Correct Answer: A. Earned value is less than planned value at the measurement date