Activity Definition, Sequencing and Data Preparation

Introduction

A project schedule does not begin with dates, software, or a critical path calculation. It begins with a disciplined translation of project scope into a logically connected set of manageable activities. The Work Breakdown Structure (WBS) describes what the project must deliver; schedule planning determines the work required to produce those deliverables, the order in which that work can occur, and the information needed to estimate and control it.

This chapter examines the preparation stage that precedes network diagramming and the Critical Path Method (CPM). It focuses on activity definition, activity coding, sequencing, dependencies, milestones, duration bases, assumptions, schedule decomposition, rolling-wave planning, and the quality of schedule data. These subjects are foundational because errors introduced here propagate into the network, float calculations, completion forecasts, resource decisions, and managerial commitments.

A high-quality schedule is not merely a list of tasks. It is a time-based model of project work and logic. Its usefulness depends on whether activities are measurable, logically related, sufficiently detailed, consistently coded, and supported by explicit assumptions.

Learning Objectives

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

1. Convert scope statements and WBS work packages into measurable schedule activities.
2. Distinguish activities, work packages, milestones, summary tasks, and planning packages.
3. Establish an activity coding structure and a usable activity register.
4. Identify and classify logical dependencies, including mandatory, discretionary, external, and internal relationships.
5. Apply the four standard precedence relationships and recognize inappropriate use of leads and lags.
6. Define duration bases, calendars, assumptions, constraints, and milestones.
7. Apply schedule decomposition and rolling-wave planning appropriately.
8. Prepare reliable data for network diagrams and CPM analysis.
9. Critically evaluate schedule quality, governance, and the use of scheduling technology.

1. From Scope and WBS to Schedule Activities

The WBS is a hierarchical decomposition of the total project scope. At its lower levels it identifies work packages, which are components of scope that can be planned, estimated, assigned, and controlled. A work package is not automatically equivalent to a single schedule activity. It may contain several activities with different owners, durations, predecessors, successors, and completion criteria.

For example, a work package titled “Install electrical system” may be decomposed into:

• approve electrical installation drawings;
• procure cable and switchgear;
• install cable trays;
• pull and terminate cables;
• install switchgear;
• inspect and test the system; and
• obtain approval for energization.

The decomposition should be sufficient to make progress observable and responsibility clear, but not so detailed that the schedule becomes administratively burdensome. The appropriate level depends on project scale, risk, uncertainty, reporting requirements, and the planning horizon.

1.1 Characteristics of a well-defined activity

A schedule activity should normally have:

• a unique identifier;
• a clear verb–noun description, such as “Approve foundation design”;
• a defined start and finish condition;
• an accountable owner or responsible organizational unit;
• an estimated duration and duration unit;
• known or expected predecessors and successors;
• a relationship to a WBS element or deliverable;
• an applicable calendar;
• assumptions and constraints recorded separately; and
• an objective method of measuring completion.

Descriptions such as “design,” “construction,” or “testing” are often too broad. A stronger description specifies an action and an object: “Complete detailed design for pump station,” “Pour east foundation slab,” or “Execute integrated user acceptance test.”

1.2 Activity versus milestone

An activity consumes time and may consume resources. A milestone is a significant point in time with zero planned duration. Examples include contract award, design approval, system ready for testing, regulatory submission, and project handover.

Milestones are useful for governance and communication, but they should not replace the activities that produce them. “Design approved” is a milestone; “Review and approve detailed design” is an activity that leads to it. A milestone should have an objectively verifiable achievement condition and an identified owner.

1.3 Summary tasks and planning packages

A summary task groups lower-level activities for reporting. It is not usually an independently executable piece of work and should not be used as a substitute for detailed logic. A planning package is a WBS component for which near-term work is not yet decomposed into detailed activities because information is incomplete or the work lies in a later planning horizon.

Planning packages are legitimate when used deliberately. They become problematic when they conceal poorly understood scope, omit interfaces, or remain unchanged despite the arrival of better information.

2. Activity Definition Techniques

Activity definition combines decomposition, expert judgment, historical information, templates, and planning workshops. A practical process is as follows:

1. Review the scope baseline, WBS dictionary, requirements, contracts, and acceptance criteria.
2. Select a work package or planning package.
3. Identify the actions needed to produce its deliverable.
4. Separate procurement, design, approval, construction, integration, testing, and handover work where their logic differs.
5. Define start and finish conditions for each activity.
6. Identify the responsible party and completion evidence.
7. Estimate the information currently known and record uncertainty.
8. Validate the decomposition with technical specialists and delivery managers.
9. Link each activity to its WBS element and deliverable.

A useful test is the “if–then” test: if an activity is complete, what observable result exists? If the answer is unclear, the activity may be poorly defined. Another test is ownership: if no person or organizational unit can accept responsibility for completion, the activity may be too broad or may represent a deliverable rather than work.

2.1 Activity attributes

Activity attributes are structured information about an activity. A basic activity register may contain the following fields:

The activity register becomes a principal input to network analysis. It should be controlled as a project management record rather than treated as an informal spreadsheet.

3. Coding Structures and Data Governance

Coding structures allow schedule information to be sorted, filtered, summarized, and integrated with cost, risk, procurement, document control, and organizational systems. A code should communicate useful information without becoming unnecessarily complex.

Possible coding dimensions include:

• WBS or deliverable;
• geographic area;
• discipline or technical package;
• responsible organization;
• contract or supplier;
• work type;
• phase;
• cost account;
• control account; and
• schedule level or reporting tier.

An illustrative identifier might be CIV-02-045, where CIV denotes civil works, 02 denotes a location or subsystem, and 045 is a sequential activity number. The exact format is less important than uniqueness, stability, documented meaning, and governance.

A good coding system has four properties:

1. Uniqueness: no two active activities share the same identifier.
2. Stability: an identifier is not casually changed because a description or date changes.
3. Traceability: the code can be related to the WBS, responsible party, and deliverable.
4. Extensibility: new activities can be added without redesigning the entire structure.

Codes should not carry information that changes frequently. For example, embedding a forecast month in an activity ID creates administrative problems when the forecast changes. The code should identify the activity; separate fields should hold status, dates, and forecast information.

4. Dependency Logic and Activity Sequencing

Sequencing establishes the order in which activities are planned to occur. It does not necessarily mean that every activity must be performed strictly one after another. Some work can proceed in parallel, while other work must wait for information, materials, approvals, access, or technical completion.

4.1 Types of dependencies

Dependencies can be classified in several ways.

Mandatory dependencies arise from the nature of the work, physical conditions, safety requirements, law, or contract. For example, concrete must be placed before it can cure, and a system normally must be installed before it can be tested.

Discretionary dependencies reflect a preferred approach or organizational convention. A team may choose to complete all architectural design before beginning structural coordination, even though partial overlap may be technically possible. Such logic should be documented because it may be reconsidered when the schedule is compressed.

External dependencies involve parties or events outside the project team, such as regulatory approval, delivery by a supplier, utility connection, or a client decision.

Internal dependencies are controlled within the project, such as the relationship between two design activities conducted by the same organization.

The classification matters because different risks and management responses apply. Mandatory logic is generally difficult to remove; discretionary logic may be challenged; external logic requires stakeholder management; and internal logic may be changed through resource or process decisions.

4.2 Precedence relationships

The Precedence Diagramming Method (PDM) represents activities as nodes and connects them with logical relationships. The four standard relationships are:

FS is common, but using FS for every relationship can create an artificially sequential schedule. SS and FF can represent legitimate overlap when supported by the method of work. SF is rare and should be used only when its meaning is clearly understood.

A relationship should express a real condition, not merely force a desired date. If a manager wants an activity to occur later for convenience, that may be a constraint or a planning preference rather than a technical dependency.

4.3 Leads and lags

A lag is a planned delay between related activities. For example, a curing period may be represented as a finish-to-start relationship with a positive lag, although modelling curing as an explicit activity is often clearer when it has risk, ownership, or monitoring significance.

A lead is an intentional overlap that allows a successor to begin before a predecessor is fully complete. Leads can be difficult to interpret and can conceal risk. Explicitly decomposing work into sections or using SS/FF logic is often more transparent.

Lags and leads should have a documented basis, unit, and owner. A vague “wait 10 days” is weak schedule data unless the reason, calendar, and measurement basis are known.

5. Duration Bases, Calendars and Assumptions

Duration is the amount of time planned between an activity’s start and finish under a specified calendar. It is not automatically equal to effort. Ten person-days of effort may require two calendar days with five available people, or ten calendar days with one person, subject to productivity, coordination, and resource constraints.

A duration estimate should state its basis, such as:

• expert assessment;
• analogous performance on a previous project;
• productivity rate;
• supplier quotation;
• quantity and production-rate calculation;
• three-point estimate; or
• statistically derived historical data.

For a quantity-based estimate, a simple duration relationship is:

where D is duration, Q is the quantity of work, and R is the production rate in quantity units per unit of time. The equation is applicable only when the quantity, production rate, work conditions, resource availability, and calendar are compatible. For example, 1,200 metres of cable installed at 300 metres per working day gives D = 1200 ÷ 300 = 4 working days, before considering access, inspection, weather, rework, and mobilization.

A three-point estimate may use optimistic, most likely, and pessimistic values. In the Program Evaluation and Review Technique (PERT) convention, the expected duration is:

where a is the optimistic duration, m is the most likely duration, and b is the pessimistic duration. The result is an expected-value estimate under the PERT assumptions and should not be presented as a guaranteed duration. The time units must be identical, and the three values should describe the same activity scope and calendar basis.

For example, if a = 4 days, m = 7 days, and b = 16 days:

The estimate is eight days. It remains an estimate and should be accompanied by the assumptions that explain the range.

5.1 Working time and elapsed time

Working days depend on the activity calendar. A five-day workweek, a seven-day construction calendar, a two-shift manufacturing calendar, and a calendar containing public holidays will produce different finish dates for the same nominal duration. Schedule records should therefore distinguish duration units from calendar rules.

Assumptions may include uninterrupted access, timely approval, availability of a specialist, normal weather, or delivery by an agreed date. Assumptions are not facts. They should be recorded, assigned an owner, and connected to risks where failure would affect the schedule.

Constraints include contractual completion dates, regulatory windows, imposed start dates, seasonal restrictions, resource limitations, and required handover events. Hard constraints can reduce the network’s ability to reveal actual flexibility and should be used only when justified.

6. Milestones and Schedule Boundaries

Milestones provide control points across the project life cycle. They can be classified as contractual, regulatory, technical, decision, procurement, or internal management milestones. A milestone should answer three questions:

1. What event has occurred?
2. What evidence confirms it?
3. Who has authority to declare it achieved?

A milestone such as “ready for commissioning” should be supported by completion of defined prerequisites, approved records, test results, and safety authorization. Without such criteria, milestone reporting becomes subjective and can create false confidence.

Milestones also establish schedule boundaries. A design approval milestone may separate concept development from detailed design; a notice to proceed may separate planning from execution; and operational handover may mark the end of project delivery and the beginning of operations.

7. Schedule Decomposition and Rolling-Wave Planning

Schedule decomposition is the progressive division of work into activities that can be estimated, sequenced, assigned, and controlled. Excessive decomposition produces numerous low-value activities, frequent status updates, and false precision. Insufficient decomposition hides interfaces, delays, and accountability.

A practical activity should be small enough that progress can be assessed reliably, but large enough to represent meaningful work. There is no universal duration threshold. A highly repetitive project may use short installation activities, while a research project may use longer knowledge-development activities with milestone-based evidence.

Rolling-wave planning recognizes that planning detail should increase as work approaches and information improves. Near-term work is decomposed in detail; distant work remains at a higher level until uncertainty is reduced. The approach is not permission to postpone all planning. The project team should establish future planning triggers, such as design maturity, procurement award, regulatory decision, or a defined time horizon.

A rolling-wave cycle includes:

1. identify a future work package or planning package;
2. record its current assumptions and uncertainty;
3. define a provisional milestone and boundary;
4. monitor information and decision triggers;
5. decompose the package when sufficient information becomes available;
6. sequence and estimate the newly defined activities;
7. validate interfaces with adjacent work; and
8. control changes to the schedule baseline.

Rolling-wave planning is particularly valuable in research, product development, complex engineering, and projects with progressive design. It must be governed carefully so that later elaboration does not silently change approved scope or conceal schedule growth.

8. Preparing the Activity and Dependency Register

Before a network diagram is constructed, the project team should assemble a dependency register. A compact register might contain the following fields:

The register separates logic from dates. This is important because dates change as forecasts change, while the underlying relationship may remain valid. It also makes review easier: subject-matter experts can challenge the rationale without navigating a complex software file.

8.1 Network quality checks

Before CPM analysis, the following checks should be performed:

• every non-start activity has an appropriate predecessor unless intentionally authorized;
• every non-finish activity has an appropriate successor;
• no unintended loops or circular relationships exist;
• relationship types reflect actual work methods;
• leads, lags, and constraints have documented reasons;
• durations use consistent units and calendars;
• milestones have zero duration and clear acceptance criteria;
• activities are linked to WBS elements;
• external dependencies have named owners;
• assumptions are recorded and linked to risks; and
• the network has a clear start and finish basis.

A schedule with numerous open ends, unexplained constraints, excessive lags, or large numbers of date overrides may produce attractive dates but weak analytical results.

9. Worked Application: Community Health Facility

The following is an illustrative case.

Background

A public agency is delivering a small community health facility. The scope includes design, permitting, procurement, construction, medical-equipment installation, commissioning, and handover. The sponsor requires a transparent schedule that can be reviewed by the contractor, regulator, and operations team.

Problem

The initial schedule contains broad lines such as “design,” “build facility,” and “commission.” It shows a completion date but does not explain how approval, procurement, inspection, or handover affect that date. The project manager therefore cannot identify the schedule drivers or assess the effect of a permit delay.

Analysis and approach

The team decomposes the relevant work package into the following activities:

The team identifies a discretionary choice: procurement could begin after design approval, before the permit is received. That choice is represented by the relationship between B and D, not by forcing procurement to wait for C. The team also records the external dependency between permit submission and the authority’s response. “Obtain construction permit” is an activity because it consumes elapsed time and requires monitoring; “permit received” is a milestone that may follow it.

The initial network contains two branches after B. The construction branch is B → C → E, with duration 5 + 15 + 30 = 50 working days from the start of B through completion of E. The procurement branch is B → D, with duration 5 + 20 = 25 working days through completion of D. Installation cannot begin until both E and D are complete. Therefore, from the start of B, installation becomes possible after the longer branch, 50 working days, and commissioning adds 8 + 5 = 13 working days. The controlling sequence is therefore B → C → E → F → G, assuming the stated durations and relationships.

If the permit activity increases from 15 to 25 working days, the controlling branch increases by 10 working days and the projected handover moves by 10 working days, unless an approved mitigation changes the logic or duration. If procurement is delayed by 10 working days while construction remains the longer branch, the handover date may not change because the procurement branch retains 15 working days of relative allowance before it controls installation. This illustrates why a network is more informative than a task list: it shows both dependency structure and potential schedule flexibility.

Outcome

The revised schedule provides:

• measurable activities linked to the WBS;
• an explicit permit dependency and owner;
• a separate procurement branch for long-lead equipment;
• a clear handover milestone;
• documented assumptions regarding approval and delivery; and
• a basis for later CPM and float analysis.

Learning points

The case demonstrates that activity definition and sequencing precede meaningful critical-path analysis. A calculated critical path is only as credible as the activities, durations, calendars, and relationships on which it is based.

10. Managerial, Governance and Ethical Considerations

Schedule data influences contracts, funding decisions, staffing, public commitments, and claims. Deliberately optimistic durations, hidden constraints, or omitted dependencies can create misleading confidence and transfer risk unfairly to other parties. Schedule integrity is therefore a governance concern.

Managers should distinguish between a target date, a forecast date, and a contractual date. A target expresses intent; a forecast reflects current information; and a contractual date may carry legal consequences. Treating these as interchangeable encourages manipulation.

Schedule changes should be traceable. A controlled process should record the change, reason, impact, approving authority, and relationship to scope, risk, cost, and quality. In regulated or publicly funded projects, auditability may be mandatory.

Ethical practice includes reporting uncertainty honestly, preserving unfavorable information, avoiding artificial activity splitting to hide delay, and ensuring that schedule compression does not compromise safety or quality. A schedule should support responsible decisions rather than merely satisfy reporting expectations.

11. Technology, Automation and Contemporary Practice

Modern scheduling platforms can generate activity IDs, identify open ends, detect circular logic, apply calendars, produce network views, and connect schedule data to cost and risk systems. These capabilities improve consistency but do not replace professional judgment. Software can calculate dates from incorrect logic with great speed.

Artificial intelligence and machine learning may assist with activity extraction from documents, duration benchmarking, anomaly detection, and forecasting from historical progress data. However, automated suggestions require human validation. Risks include biased historical data, unexplained recommendations, confidentiality breaches, and the mistaken assumption that statistical similarity guarantees technical equivalence.

Digital twins, model-based planning, and Building Information Modelling can improve sequencing by linking activities to physical elements, access requirements, and installation zones. Their value depends on disciplined information standards and agreement about ownership of model and schedule data.

Best-Practice Checklist

Before approving activity data for network development, confirm that:

1. every activity represents identifiable work or a justified milestone;
2. each activity has a clear completion condition;
3. the activity is linked to scope and a responsible owner;
4. durations have a stated basis and calendar;
5. assumptions, constraints, and external dependencies are visible;
6. relationship types reflect the actual method of work;
7. leads and lags are justified and measurable;
8. planning packages have future decomposition triggers;
9. milestones have objective acceptance evidence; and
10. the register has been reviewed by the people who will perform or accept the work.

Chapter Summary

Activity definition transforms WBS scope into measurable schedule work. Activities should have clear action-oriented descriptions, completion criteria, owners, durations, calendars, and links to deliverables. Milestones mark significant zero-duration events, while summary tasks and planning packages serve reporting and progressive-planning purposes.

Sequencing establishes the logical relationships among activities. Mandatory, discretionary, internal, and external dependencies should be distinguished because they have different management implications. The four precedence relationships are finish-to-start, start-to-start, finish-to-finish, and start-to-finish. Leads, lags, assumptions, and constraints require explicit justification.

Duration data must identify its estimation basis and time calendar. Quantity-rate calculations and three-point estimates are useful only when their assumptions and units are consistent. Coding structures and dependency registers improve traceability, review, reporting, and integration with other project systems.

Rolling-wave planning enables detailed near-term planning while preserving appropriate flexibility for uncertain future work. It must be governed through decomposition triggers and change control. Ultimately, CPM results are only as reliable as the activity definitions, durations, calendars, and logic prepared before network analysis begins.

Key Terms

Activity: A defined unit of project work that consumes time and may consume resources.

Activity attribute: Structured information describing an activity, such as owner, duration, calendar, or WBS code.

Activity register: A controlled record containing schedule activities and their attributes.

Assumption: A condition treated as true for planning purposes but subject to uncertainty.

Calendar: The working-time pattern used to calculate activity dates and duration.

Constraint: A restriction that limits when an activity or milestone may occur.

Dependency: A logical relationship between two activities or events.

Finish-to-start: A relationship in which the successor cannot start until the predecessor finishes.

Lag: A planned delay between related activities.

Lead: An intentional overlap that allows a successor to begin before the predecessor is complete.

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

Planning package: A future work component not yet decomposed into detailed activities.

Precedence Diagramming Method: A network method that represents activities as nodes connected by logical relationships.

Rolling-wave planning: Progressive elaboration in which near-term work is planned in greater detail than distant work.

Schedule decomposition: The division of work into manageable, estimable, and controllable activities.

Summary task: A reporting-level grouping of subordinate activities.

Work Breakdown Structure: A hierarchical decomposition of the total project scope.

Work package: A WBS component that can be planned, estimated, assigned, and controlled.

Review Questions

Short Answer Questions

1. Why is a work package not necessarily equivalent to one schedule activity?
2. What distinguishes a milestone from an activity?
3. Define mandatory, discretionary, internal, and external dependencies.
4. What information should be recorded in an activity register?
5. Why should activity duration always be associated with a calendar?

Long Answer Questions

1. Explain how a project team should transform a WBS work package into a validated set of schedule activities.
2. Compare the four precedence relationships and discuss circumstances in which each may be appropriate.
3. Explain rolling-wave planning, including its benefits, risks, and governance requirements.
4. Discuss how assumptions, constraints, leads, and lags can affect the credibility of a network schedule.
5. Evaluate the managerial and ethical consequences of poor-quality or deliberately manipulated schedule data.

Analytical / Application Questions

1. A project includes design approval, regulatory permitting, procurement, construction, testing, and handover. Develop a preliminary activity register with at least eight activities, including owners, durations, milestones, and logical relationships. Identify two external dependencies and two assumptions.
2. An activity requires 2,400 units of work at a production rate of 400 units per working day. Calculate its duration and explain at least three conditions that could make the estimate unreliable.
3. A future work package is six months away, its technical solution is not yet selected, and its interfaces are uncertain. Recommend a rolling-wave planning approach. Specify what should be recorded now and what information should trigger detailed decomposition.

Multiple Choice Questions with Answers

Question 1

Which statement best describes a schedule activity?

A. A broad statement of the project’s business case

B. A measurable unit of work with defined start and finish conditions

C. A zero-duration governance event only

D. A cost account without a time dimension

Correct Answer: B. A measurable unit of work with defined start and finish conditions

Question 2

Which relationship means that a successor cannot start until its predecessor finishes?

A. Start-to-start

B. Finish-to-finish

C. Finish-to-start

D. Start-to-finish

Correct Answer: C. Finish-to-start

Question 3

Which dependency is most likely discretionary?

A. A legally required inspection before operation

B. A curing period required by the material specification

C. A supplier delivery date stated in a contract

D. A preferred organizational practice to complete one design discipline before another

Correct Answer: D. A preferred organizational practice to complete one design discipline before another

Question 4

A work package requires 1,200 metres of installation at 300 metres per working day. What is the basic production-rate duration?

A. 3 working days

B. 4 working days

C. 300 working days

D. 1,500 working days

Correct Answer: B. 4 working days

Question 5

Using the PERT convention, what is the expected duration when the optimistic estimate is 4 days, most likely is 7 days, and pessimistic is 16 days?

A. 6 days

B. 7 days

C. 8 days

D. 16 days

Correct Answer: C. 8 days

Question 6

Which item is a milestone rather than an activity?

A. Install medical equipment

B. Review design documents

C. Obtain regulatory approval

D. Design approved

Correct Answer: D. Design approved

Question 7

What is the principal purpose of an activity coding structure?

A. To guarantee that every activity has zero float

B. To replace the need for logical relationships

C. To support identification, sorting, reporting, and traceability

D. To convert all durations into elapsed time

Correct Answer: C. To support identification, sorting, reporting, and traceability

Question 8

What is the central feature of rolling-wave planning?

A. All project work is fully detailed before execution begins

B. Near-term work is planned in greater detail than uncertain distant work

C. Future work is excluded from the schedule until it starts

D. Only contractual milestones are progressively elaborated

Correct Answer: B. Near-term work is planned in greater detail than uncertain distant work

Question 9

Which practice most improves the credibility of a lag in a schedule?

A. Leaving its basis undocumented

B. Using it to force a preferred date

C. Recording its reason, unit, calendar basis, and owner

D. Applying it to every finish-to-start relationship

Correct Answer: C. Recording its reason, unit, calendar basis, and owner

Question 10

Why is a dependency register useful before CPM analysis?

A. It removes the need to estimate activity durations

B. It documents and supports review of relationships and their rationale

C. It automatically identifies the legally responsible contractor

D. It ensures that all activities will be on the critical path

Correct Answer: B. It documents and supports review of relationships and their rationale