Aggregate Planning, Scheduling and Resource Control

Introduction and context

Operations management converts market requirements into coordinated decisions about people, materials, equipment, time and information. Aggregate planning, master scheduling and short-term scheduling form a hierarchy: aggregate planning establishes an economically feasible medium-term direction; the master production schedule (MPS) translates that direction into end-item commitments; and detailed scheduling assigns work to specific resources and time slots. Resource-control systems then monitor execution and revise plans when demand, capacity or material availability changes.

These decisions are central to supply-chain performance. A plan that promises more than the factory can produce creates lateness and expediting. A plan that protects capacity excessively creates inventory and idle resources. Effective planning therefore balances customer service, cost, flexibility, labour stability, working capital and operational risk.

Learning objectives

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

1. Explain the hierarchy and purpose of aggregate planning, MPS and short-term scheduling.
2. Construct a basic aggregate plan using workforce, inventory, overtime and subcontracting alternatives.
3. Develop a master schedule and calculate projected available balance and available-to-promise inventory.
4. Apply loading, sequencing, Gantt charts and finite-capacity concepts.
5. Explain the theory of constraints (TOC), MRP, MRP II and ERP-based resource planning.
6. Evaluate planning systems critically, including their assumptions, data requirements and governance risks.

1. The planning hierarchy

Planning horizons are linked but not identical. Strategic capacity decisions concern facilities, technology and long-term sourcing. Aggregate planning commonly covers approximately 3–18 months and uses product families, labour categories and monthly or weekly periods. The MPS usually covers a shorter, rolling horizon and specifies individual end items or service outputs. Detailed scheduling may cover days, shifts or hours.

The hierarchy is not a one-way cascade. Capacity feedback can force revision of the MPS, while shop-floor disruptions can require changes to the aggregate plan. A sound system uses a rolling process: completed periods are closed, new periods are added, actual results are compared with the plan, and assumptions are updated.

Figure: The planning hierarchy is integrated by capacity and execution feedback.

Aggregate planning

Aggregate planning determines how total expected demand will be met over a medium-term horizon. Demand is expressed in aggregate units, such as tonnes, labour hours or product-family units, rather than every stock-keeping unit (SKU). The planner chooses a strategy and evaluates its consequences.

A level strategy keeps production or workforce relatively stable and absorbs demand variation through inventory, backlog or capacity buffers. A chase strategy varies production to follow demand, often through hiring, layoffs, variable hours or external capacity. A mixed strategy combines these methods and is common in practice.

Typical decision variables include regular-time production, overtime, inventory, backorders, subcontracting, workforce size and temporary labour. Relevant costs include hiring and training, layoffs, regular and overtime labour, inventory carrying, shortage or backorder, subcontracting and changeover costs. Fixed facility costs may be excluded when they do not vary among alternatives, but this assumption should be stated.

A simplified aggregate-planning balance is:

Here, I_t is ending inventory in period t, P_t is production during the period and D_t is demand. The equation assumes inventory is measured in consistent units and that production and demand are available in the same period. If shortages are permitted, I_t may be negative and represents net inventory or backlog; if shortages are prohibited, the constraint I_t ≥ 0 applies.

A basic cost objective is:

where Z is total relevant cost; R_t and O_t are regular and overtime production; I_t^+ is positive ending inventory; B_t is backlog; S_t is subcontracted quantity; H_t and L_t are hires and layoffs; and the c terms are their unit costs. This is a planning model, not a universal cost law. It assumes costs are known or estimated, units are comparable, and the chosen variables adequately represent feasible operations.

Worked example: a mixed aggregate plan

An assembly operation begins with 100 units in inventory. Forecast demand for three months is 500, 700 and 400 units. Regular production is fixed at 600 units per month, and overtime production is used only when needed. Assume no safety stock requirement and no subcontracting.

The inventory balance gives:

For month 1, 100 + 600 − 500 = 200. Month 2 gives 200 + 600 − 700 = 100, and month 3 gives 100 + 600 − 400 = 300. Total demand is 1,600 units; opening inventory plus regular production is 1,900 units, so the plan ends with 300 units. If the desired ending inventory were zero, regular production could be reduced by 300 units, subject to workforce and minimum-run constraints. The example illustrates why a feasible plan is not automatically an economical plan.

2. Master production scheduling

The MPS disaggregates the aggregate plan into end-item quantities and dates. It is a contractual-looking promise inside the organisation: sales uses it to promise delivery, procurement uses it to anticipate components, and production uses it to reserve capacity. The MPS must therefore be stable enough to execute but responsive enough to reflect real demand.

Important concepts include:

• Time fence: a boundary beyond which schedule changes are restricted or require approval.
• Planning horizon: the number of future periods visible to the planning system.
• Projected available balance (PAB): expected inventory after scheduled supply and demand.
• Available-to-promise (ATP): uncommitted supply that can be promised to customers.
• Lot size: the quantity released or produced at one time.
• Safety stock: inventory protection against uncertainty, not a substitute for reliable data.

For a periodic MPS, a useful balance is:

where PAB_t is projected available balance, MPS_t is the scheduled receipt in period t, and D_t is demand used by the MPS. If a safety stock SS is required, feasibility normally requires PAB_t ≥ SS.

ATP is calculated differently depending on the system convention. For the first period, a common expression is:

For later MPS receipt periods, ATP equals that receipt minus customer orders from the current receipt period up to, but not including, the next MPS receipt. Forecast demand is generally used for PAB, whereas confirmed customer orders are used for ATP; organisations must document their convention.

MPS example

Suppose opening inventory is 80 units. Weekly MPS receipts are 100 units in weeks 2 and 4. Demand is 60, 70, 90 and 80 units in weeks 1–4.

The negative PAB in week 3 shows that the MPS is infeasible unless a receipt is moved forward, demand is rescheduled, safety stock is relaxed, or an alternative source is used. This is a capacity-and-service issue, not merely a spreadsheet error.

3. Loading, sequencing and Gantt charts

Loading assigns work to resources. In infinite loading, all required work is placed on a resource without initially checking whether the resource can perform it in the available time. Infinite loading reveals workload and overloads, but it does not produce an executable schedule. In finite loading, work is assigned only when the resource has available capacity, so operation dates may move or orders may require alternative resources.

If a job requires processing time p_j and setup time s_j on a resource, its basic load is:

For a resource in period t, utilisation is:

Both numerator and denominator must use the same unit, such as machine-hours. Utilisation above 1, or 100%, indicates overload. Practical capacity is often below theoretical capacity because breaks, maintenance, absenteeism, quality losses and changeovers consume time.

A Gantt chart displays jobs against time, showing planned start, processing, setup, waiting and completion. It supports visual control and communication, but it does not by itself optimise the schedule or prove that all constraints are satisfied.

Sequencing rules include:

• First-come, first-served: perceived as fair and simple, but often poor for flow.
• Shortest processing time: tends to reduce average flow time, but may delay long jobs.
• Earliest due date: prioritises delivery performance.
• Critical ratio: CR = (time ext{ remaining until due date})/(work ext{ remaining}). Values below 1 indicate a job is behind its required rate. The rule is dynamic and must be recalculated.
• Slack per operation: remaining slack divided by remaining operations, useful when routing complexity differs.

No rule dominates in every environment. Due-date reliability, setup families, rush orders, material availability and fairness must be considered together.

4. Finite capacity and the theory of constraints

Finite-capacity planning recognises that a machine, labour group, supplier or warehouse has a limited amount of usable capacity. It must account for calendars, skills, maintenance, tooling, material availability, precedence relationships and sometimes shared resources. A schedule can be mathematically feasible yet operationally fragile if it leaves no allowance for variability.

The theory of constraints views every system as having at least one constraint that limits throughput. Its five focusing steps are:

1. Identify the constraint.
2. Exploit it—use its available time on value-adding work and prevent avoidable stoppage.
3. Subordinate other activities—do not release work that overwhelms the constraint.
4. Elevate the constraint—add capacity through overtime, tooling, redesign, outsourcing or investment.
5. Repeat when the constraint moves.

TOC commonly uses drum-buffer-rope. The drum is the constraint’s pace; the buffer protects it from disruption; the rope controls release of work so that upstream WIP does not grow without purpose. The approach challenges local efficiency: keeping every machine busy can increase queues and lead time while failing to increase throughput.

A simplified throughput relationship is:

In operational analysis, throughput should be distinguished from production output: producing items that are not sold may increase inventory but not economic throughput. TOC decisions should also consider due dates, quality and strategic bottlenecks, not only physical machine speed.

5. Material and resource planning: MRP, MRP II and ERP

Material requirements planning (MRP) is a time-phased system for determining dependent-component requirements. It uses the MPS, bill of materials (BOM), inventory records, open orders and lead times. The basic logic is: explode the BOM, net available inventory and scheduled receipts against gross requirements, apply lot-sizing rules, and offset planned order receipts by lead time to obtain planned releases.

For a component, net requirements can be represented as:

A negative result normally becomes zero, because no new requirement is needed in that period. However, safety stock, allocated inventory, scrap, lot sizes and order policies alter the practical calculation. Planned release in period t becomes a planned receipt after the stated lead time; this assumes lead time is sufficiently stable and the BOM is accurate.

MRP is especially effective for dependent demand, where component demand is derived from planned parent production. It is less suitable as a stand-alone forecast system for independent customer demand.

MRP II (manufacturing resource planning) extends MRP by integrating capacity planning, shop-floor control, purchasing, finance and feedback. It asks not only “what materials are required?” but also “can labour, machines and money support the plan?”

Enterprise resource planning (ERP) integrates broader organisational processes, including finance, procurement, inventory, production, sales, maintenance, quality, human resources and distribution. Modern ERP platforms may use cloud architecture, application programming interfaces, advanced planning modules, Internet-of-Things data and analytics. Integration creates a shared record, but it also means that a poor master datum—such as an incorrect lead time, unit of measure or BOM—can propagate across the enterprise.

Implementation priorities include data ownership, controlled master-data changes, role-based access, audit trails, segregation of duties, cybersecurity, backup and recovery, user training, testing and disciplined process design. ERP should not be treated as a substitute for sound policies; automating a flawed process makes errors faster and less visible.

6. Integrated applied case: Northstar Pumps

The following is an illustrative case. Northstar Pumps assembles industrial circulation pumps. Demand is seasonal, the final-test station is shared by three product families, and sales has promised short lead times. The company experiences overtime, late orders and excess work-in-process.

Background and problem

The aggregate forecast for the next quarter is 1,200 pump-equivalent units. The plant has regular capacity of 400 units per month, but demand is 300, 500 and 400 units. Opening finished-goods inventory is 100 units. The final-test station has 360 available hours per month and requires 1.0, 1.2 and 0.8 hours for product families A, B and C. The MPS has been created from sales orders without checking the station’s finite capacity.

Analysis

At the aggregate level, regular output plus opening inventory equals 1,300 units, so the quarterly quantity is feasible with 100 units ending inventory. However, the month-two product mix creates a bottleneck. If month-two requirements are 150 A, 250 B and 100 C, final-test load is:

(150×1.0) + (250×1.2) + (100×0.8) = 530 hours.

The load is 170 hours above the 360-hour capacity, or approximately 147.2% utilisation. A volume-only aggregate plan therefore conceals a resource overload.

Approach and solution

Northstar first freezes customer commitments inside a short time fence and validates the BOM, routings and test-time standards. It then applies TOC: the final-test station becomes the drum. Preventive maintenance is scheduled outside peak demand, test fixtures are prepared in advance, and quality defects are prevented from consuming scarce test time. The rope is implemented through a controlled release of subassemblies; upstream departments no longer release work merely to achieve local utilisation.

The MPS is revised by moving 80 units of family B to month one and 90 units to month three, subject to customer agreement and component availability. A finite schedule sequences similar test setups together, while a small time buffer protects the station from disruption. Procurement uses the revised MPS to run MRP, and the ERP system records revised due dates, planned orders and capacity exceptions.

Outcome and learning points

The revised plan removes the month-two overload, reduces WIP release and makes customer promises visible. Its success depends on accurate standards and cross-functional agreement; a software reschedule alone would not create capacity. The case demonstrates four principles: aggregate volume does not guarantee mix feasibility; the MPS must be capacity checked; the constraint should govern release; and ERP is most valuable when supported by disciplined data and decision rights.

7. Challenges, ethics and contemporary practice

Planning systems face forecast error, unreliable suppliers, absenteeism, breakdowns, engineering changes, inaccurate inventory, nervousness caused by frequent rescheduling and conflicting departmental incentives. Planners should use scenario analysis, time fences, exception reports, safety capacity and post-period variance review rather than assuming the forecast is exact.

Ethical and governance issues include truthful customer commitments, fair treatment of workers during overtime or layoffs, safe staffing, responsible subcontracting, protection of personal and commercial data, and avoidance of algorithmic decisions that conceal discrimination or shift risk to vulnerable suppliers. Sustainability adds energy, emissions, waste and circularity considerations to cost and service objectives. A plan that minimises labour cost while increasing unsafe hours is not operational excellence.

Artificial intelligence can assist demand sensing, anomaly detection, predictive maintenance, dynamic sequencing and scenario generation. Its outputs require human review because training data may encode past bias, unusual events may not resemble historical patterns, and optimisation may exploit an incomplete objective. Governance should specify data lineage, model validation, override authority, monitoring and explainability appropriate to the decision’s risk.

Best practice is to integrate, but not confuse, planning levels. Aggregate plans should be financially and operationally feasible. MPS changes should be governed by time fences. MRP should use accurate BOMs, lead times and inventory records. Finite-capacity checks should focus on bottlenecks and critical resources. Schedulers should measure delivery reliability, flow time, WIP, inventory, utilisation, schedule adherence, changeover performance and quality—not utilisation alone.

Chapter Summary

Aggregate planning balances medium-term demand and capacity through workforce, inventory, overtime, subcontracting and backlog decisions. Its inventory balance links opening stock, production and demand, while cost models compare feasible strategies. The MPS converts family-level plans into end-item receipts and customer commitments; projected available balance and ATP expose service and inventory consequences. Loading identifies resource requirements, finite loading respects capacity, sequencing rules establish job priority, and Gantt charts communicate timing. The theory of constraints focuses improvement and release control on the system constraint. MRP plans dependent materials through BOM explosion and time-phased netting; MRP II adds manufacturing resources and feedback; ERP integrates enterprise processes but depends on accurate data, governance and disciplined implementation. Effective resource control combines quantitative models with professional judgement, ethical practice and continuous feedback.

Key Terms

Aggregate plan: Medium-term plan for product families, capacity and resources.

Available-to-promise (ATP): Uncommitted supply available for customer commitments.

Bill of materials (BOM): Structured record of components required for a parent item.

Capacity: Maximum feasible output or resource time under stated conditions.

Critical ratio: Ratio of time remaining to work remaining for a job.

Drum-buffer-rope: TOC method linking constraint pace, protection and release control.

Enterprise resource planning (ERP): Integrated information system for organisational processes.

Finite loading: Assigning work only within available resource capacity.

Gantt chart: Time-based visual display of jobs and activities.

Infinite loading: Assigning workload without initially enforcing capacity limits.

Master production schedule (MPS): Time-phased schedule for end items or final outputs.

Material requirements planning (MRP): Time-phased planning of dependent materials.

MRP II: Extension of MRP incorporating manufacturing resources and feedback.

Projected available balance (PAB): Expected inventory after scheduled supply and demand.

Safety stock: Inventory held to protect against uncertainty.

Sequencing: Determining the order in which jobs are processed.

Time fence: Boundary controlling permitted schedule changes.

Theory of constraints (TOC): Improvement approach centred on the system constraint.

Utilisation: Workload divided by available capacity, expressed as a proportion or percentage.

Review Questions

Short Answer Questions

1. What is the purpose of aggregate planning?
2. Distinguish the MPS from an aggregate plan.
3. Define finite loading.
4. What does a negative PAB indicate?
5. State the five focusing steps of the theory of constraints.

Long Answer Questions

1. Explain how aggregate planning, MPS and detailed scheduling interact in a rolling planning system.
2. Compare level, chase and mixed aggregate-planning strategies, including their principal risks.
3. Explain how MRP differs from MRP II and ERP.
4. Discuss the strengths and limitations of common sequencing rules.
5. Explain why ERP implementation requires master-data governance and process discipline.

Analytical / Application Questions

1. A plant begins with 50 units and plans production of 300, 250 and 350 units over three periods. Demand is 280, 320 and 250 units. Calculate ending inventory for each period and interpret the result.
2. A bottleneck has 400 available hours. Three jobs require 120, 180 and 160 hours. Calculate total load and utilisation, then recommend two TOC-based actions.
3. Design a short MPS and capacity-check procedure for a company whose sales department frequently changes promised dates. Identify time-fence, data, escalation and performance-measure requirements.

Multiple Choice Questions with Answers

Question 1

Which planning level normally specifies individual end items and their planned completion periods?

A. Strategic facility planning

B. Aggregate planning

C. Master production scheduling

D. Annual financial budgeting

Correct Answer: C. Master production scheduling

Question 2

Opening inventory is 100 units, production is 600 units and demand is 450 units. What is ending inventory?

A. 150 units

B. 250 units

C. 350 units

D. 1,150 units

Correct Answer: B. 250 units

Question 3

A resource has 360 available hours and a scheduled load of 450 hours. What is its utilisation?

A. 80%

B. 90%

C. 125%

D. 810%

Correct Answer: C. 125%

Question 4

Which statement best describes finite loading?

A. It ignores resource limits until execution.

B. It assigns work only within available capacity.

C. It schedules only purchased materials.

D. It always prioritises the shortest job.

Correct Answer: B. It assigns work only within available capacity.

Question 5

In the theory of constraints, what is the purpose of the rope in drum-buffer-rope?

A. To increase finished-goods inventory everywhere

B. To control the release of work to the system

C. To replace the bottleneck with a supplier

D. To calculate customer profitability

Correct Answer: B. To control the release of work to the system

Question 6

MRP primarily plans which type of demand?

A. Dependent demand for components

B. Independent demand forecasts only

C. Employee demand for training

D. Demand for financial capital

Correct Answer: A. Dependent demand for components

Question 7

A negative projected available balance in an MPS most directly indicates that:

A. Supply exceeds demand

B. Demand has been eliminated

C. The schedule may fail to meet requirements

D. The BOM is necessarily correct

Correct Answer: C. The schedule may fail to meet requirements

Question 8

Which sequencing rule generally prioritises jobs with the nearest due dates?

A. Earliest due date

B. Longest processing time

C. Random selection

D. Largest batch first

Correct Answer: A. Earliest due date

Question 9

What is a key difference between MRP II and basic MRP?

A. MRP II excludes capacity considerations

B. MRP II incorporates broader manufacturing resources and feedback

C. MRP II is used only for retail pricing

D. MRP II does not use a bill of materials

Correct Answer: B. MRP II incorporates broader manufacturing resources and feedback

Question 10

Why can an ERP system amplify operational problems?

A. It prevents all data from being shared

B. It automatically removes all planning constraints

C. Incorrect master data can propagate across integrated processes

D. It eliminates the need for process ownership

Correct Answer: C. Incorrect master data can propagate across integrated processes