flowchart LR A[Start] --> B[Activity 1<br/>Days 0-3] A --> C[Activity 2<br/>Days 0-5] B --> D[Activity 3<br/>Days 3-8] C --> E[Activity 4<br/>Days 5-9] D --> F[End] E --> F style A fill:#E3F2FD,stroke:#1565C0 style F fill:#E8F5E9,stroke:#2E7D32
77 Scheduling, Loading and Sequencing
77.1 What are Scheduling, Loading and Sequencing?
Once capacity and materials are planned, operations needs to decide when each job gets done, which machine handles it, and in what order. The three closely-related functions:
| Function | Question it answers |
|---|---|
| Loading | Which work-centre / machine handles each job? |
| Sequencing | In what order should the jobs be done? |
| Scheduling | When (start, finish) should each job be done? |
Heizer and Render frame scheduling as “the assignment of jobs to specific times in the operations plan” (heizerrender2020?).
77.2 Forward vs Backward Scheduling
| Type | What it does | Use |
|---|---|---|
| Forward scheduling | Start work as soon as the order is received; finish as early as possible | Jobs without firm due dates |
| Backward scheduling | Start with the due date and work backward to when each task must begin | Jobs with firm due dates |
77.3 Loading Methods
| Approach | Description |
|---|---|
| Infinite loading | Assume infinite capacity; check feasibility later |
| Finite loading | Schedule no job that exceeds capacity in any period |
The classical visual tools:
| Tool | What it shows |
|---|---|
| Gantt load chart | Current load on each work-centre |
| Gantt schedule chart | Time-phased plan for each job |
| Assignment method (Hungarian) | Optimal one-to-one assignment of jobs to machines / workers |
The Hungarian assignment method (Harold Kuhn, 1955) is the textbook algorithm for the special case where each job must be assigned to one machine and each machine to one job, minimising total cost.
77.4 Sequencing Rules
When several jobs await processing on the same machine, which job goes first? Common priority rules:
| Rule | What it does |
|---|---|
| FCFS (First-Come-First-Served) | Process in order of arrival |
| SPT (Shortest Processing Time) | Shortest jobs first |
| LPT (Longest Processing Time) | Longest jobs first |
| EDD (Earliest Due Date) | Jobs nearest due date first |
| CR (Critical Ratio) | (Time remaining ÷ Work remaining); lowest first |
| Slack per Operation | Slack ÷ remaining operations |
| Rule | Best for |
|---|---|
| SPT | Average flow time, average lateness, average WIP |
| EDD | Maximum tardiness, minimising lateness |
| FCFS | Fairness; simplicity |
SPT minimises average flow time is one of the most-tested rules.
77.5 Johnson’s Rule for Two-Machine Sequencing
S.M. Johnson’s algorithm (1954) finds the optimal sequence for n jobs on two machines in series, minimising total processing time (makespan) (johnson1954?):
| Step | Action |
|---|---|
| 1 | List the jobs and their times on Machine 1 and Machine 2 |
| 2 | Find the smallest time across all jobs and machines |
| 3 | If on Machine 1 → schedule that job first (earliest available slot) |
| 4 | If on Machine 2 → schedule that job last (latest available slot) |
| 5 | Cross out the scheduled job and repeat |
The result is the optimal sequence — minimum makespan and idle time on Machine 2.
77.6 Network Scheduling — PERT and CPM
For projects with many activities and dependencies, use network methods. Two classical:
| Method | Origin | What it captures |
|---|---|---|
| PERT (Program Evaluation and Review Technique) | US Navy Polaris (1958) | Probabilistic activity times — uses three time estimates (optimistic, most likely, pessimistic) |
| CPM (Critical Path Method) | DuPont (1957) | Deterministic activity times — focuses on the critical path |
| Concept | Definition |
|---|---|
| Activity | Task that consumes time and resources |
| Event | Start or end of an activity (a node) |
| Critical path | Longest path through the network — determines minimum project duration |
| Slack / Float | Time an activity can be delayed without delaying the project |
| Forward pass | Compute Earliest Start (ES) and Earliest Finish (EF) |
| Backward pass | Compute Latest Start (LS) and Latest Finish (LF) |
| Crashing | Reduce activity time at additional cost |
PERT’s expected time (Beta distribution):
\[T_e = \frac{a + 4m + b}{6}, \quad \sigma^2 = \left(\frac{b - a}{6}\right)^2\]
where \(a\) = optimistic, \(m\) = most likely, \(b\) = pessimistic.
77.7 Theory of Constraints (TOC) — Drum-Buffer-Rope
Eliyahu Goldratt’s Theory of Constraints (goldratt1984?) focuses on the bottleneck — the constraint that limits throughput. The drum-buffer-rope scheduling approach:
- Drum — the bottleneck sets the pace.
- Buffer — protective inventory before the bottleneck.
- Rope — pull signal that releases work into the system.
77.8 Practice Questions
Choosing the order in which jobs are processed is:
View solution
The Shortest Processing Time (SPT) rule minimises:
View solution
Johnson's rule applies to sequencing:
View solution
In PERT, the expected activity time is computed as:
View solution
The "critical path" of a project is:
View solution
PERT differs from CPM in that:
View solution
The Hungarian method is used for:
View solution
Goldratt's "Drum-Buffer-Rope" scheduling approach within Theory of Constraints uses the bottleneck as the:
View solution
- Three functions: Loading (which machine) · Sequencing (in what order) · Scheduling (when).
- Forward vs Backward scheduling. Infinite vs finite loading.
- Visual tools: Gantt charts; Hungarian assignment (Kuhn 1955).
- Sequencing rules: FCFS, SPT (best for avg flow time), LPT, EDD (best for max tardiness), CR.
- Johnson’s rule (1954) — optimal sequence for n jobs on two machines.
- PERT (probabilistic, 1958, Polaris) vs CPM (deterministic, 1957, DuPont). T_e = (a + 4m + b) / 6.
- Critical path = longest path = min project duration; activities on it have zero slack.
- TOC (Goldratt) — drum-buffer-rope, bottleneck-driven scheduling.