Aggregate Planning, Production Planning and Scheduling

Introduction and Context

Manufacturing and service organizations must decide not only what to produce, but also when to produce it, in what quantities, with which resources, and in what sequence. These decisions connect market demand with employees, equipment, materials, suppliers, inventories, and delivery commitments. Aggregate planning, production planning, and scheduling form a hierarchy of decisions that translates strategic intentions into executable operating plans.

Aggregate planning determines broad production, workforce, inventory, subcontracting, and back-order levels over an intermediate horizon, commonly 3–18 months. The master production schedule (MPS) converts this aggregate plan into specific end items and time-phased quantities. Material requirements planning (MRP) determines when components and raw materials are needed. Capacity requirements planning (CRP) checks whether available work-centre capacity can support the proposed plan. Detailed scheduling then assigns jobs to machines, employees, and time intervals.

The central challenge is coordination. A schedule that ignores material availability is infeasible; a material plan that ignores capacity creates overloads; and a production plan that ignores workforce constraints may be costly or unsafe. Effective operations planning therefore combines quantitative models with managerial judgment, reliable data, and continuous feedback.

Learning Objectives

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

1. Explain the hierarchy from aggregate planning to detailed scheduling.
2. Develop aggregate plans using workforce, inventory, overtime, subcontracting, and back-order alternatives.
3. Prepare an MPS and interpret time-phased production requirements.
4. Apply bill-of-materials logic and calculate MRP gross and net requirements.
5. Explain CRP and identify capacity overloads.
6. Compare job, batch, flow, and service scheduling environments.
7. Apply sequencing rules, Gantt charts, finite scheduling, and workforce-planning principles.
8. Evaluate the practical, technological, ethical, and organizational limitations of planning systems.

1. The Planning Hierarchy

Operations planning is usually organized into four levels:

The hierarchy reduces complexity. Strategic planning does not specify which component is released on Tuesday, while shop-floor scheduling should not independently decide the organization’s annual capacity strategy. Plans must, however, be linked through feedback. If the MPS repeatedly exceeds capacity, the aggregate plan or resource strategy must be revised.

Demand may be independent or dependent. Independent demand arises directly from customers, such as demand for finished bicycles. Dependent demand is calculated from the need for another item, such as tyres required for bicycles. Forecasting is central for independent demand, whereas bills of materials and planned-order logic are central for dependent demand.

2. Aggregate Planning

Aggregate planning develops a feasible and economical response to forecast demand for product families rather than individual stock-keeping units. The word “aggregate” means that products with similar processing and resource requirements are grouped together.

The plan typically balances:

• forecast demand;
• regular-time production;
• overtime production;
• workforce hiring and layoffs;
• inventory investment;
• subcontracting;
• planned shortages or back orders; and
• service-level and policy constraints.

A plan may use a level strategy, a chase strategy, or a mixed strategy. A level strategy maintains a relatively stable workforce and production rate, using inventory or back orders to absorb demand variation. A chase strategy adjusts production or workforce to follow demand closely. Level plans improve workforce stability but may increase inventory; chase plans reduce inventory but can create hiring, training, morale, and quality costs.

Aggregate planning costs

Common costs include regular production, overtime, hiring, layoffs, inventory holding, subcontracting, and back-order penalties. A simplified total-cost expression is:

where TC is total plan cost; Q_R, Q_O, and Q_S are regular-time, overtime, and subcontracted quantities; I is average or ending inventory measured according to the selected convention; B is back-order quantity; H is hires; L is layoffs; and the C terms are corresponding unit or event costs. Units must be consistent: production costs may be currency per unit, inventory cost may be currency per unit-period, and hiring cost may be currency per employee.

The inventory balance for period t is:

where I_t is ending inventory, P_t is total production, and D_t is demand in period t. If shortages are permitted, net inventory may be negative; alternatively, inventory and back orders can be represented separately:

Here I_t and B_t are nonnegative ending inventory and back orders. The equation assumes all quantities use the same unit and that yield losses, scrap, and timing within a period are ignored.

A basic workforce relationship is:

where W_t is workforce size, H_t is hires, and L_t is layoffs. If each worker produces r units per period, regular production is:

This expression applies only when labour productivity, attendance, product mix, and available hours are treated as constant. In practice, learning, absenteeism, maintenance, changeovers, and skill restrictions may make capacity lower than the theoretical value.

Worked example: a level plan

An assembly department begins with 20 workers. Each worker produces 40 units per month. Demand for three months is 700, 1,000, and 900 units. Initial inventory is 100 units, and the organization wishes to end the horizon with 100 units. A level workforce of 20 produces 800 units per month.

The inventory balance is:

• Month 1: 100 + 800 − 700 = 200 units.
• Month 2: 200 + 800 − 1,000 = 0 units.
• Month 3: 0 + 800 − 900 = −100 units.

The negative value indicates a 100-unit shortage at the end of month 3. To achieve the desired ending inventory of 100 units, total production must satisfy:

Thus, total production = 2,600 + 100 − 100 = 2,600 units. The level workforce produces only 2,400 units, so an additional 200 units are required through overtime, subcontracting, a larger workforce, or a revised service policy. This illustrates why an apparently stable plan must be checked against cumulative demand, not only average demand.

Aggregate planning methods

The trial-and-error method evaluates feasible alternatives in a spreadsheet. It is transparent and useful for teaching and managerial discussion, but it may not find the least-cost plan.

Linear programming can minimize cost subject to production, workforce, inventory, and capacity constraints. Its assumptions often include linear costs, divisible quantities, and known demand. Integer or mixed-integer programming is needed when decisions such as hiring one whole employee, opening a shift, or activating a supplier are indivisible.

Simulation evaluates policies under uncertain demand, processing times, absenteeism, or supply disruptions. It does not automatically identify an optimum; it estimates performance under specified rules.

The chosen plan should be evaluated using cost, service, stability, utilization, flexibility, risk, and environmental impact. The lowest direct cost is not necessarily the best plan if it increases safety risks, supplier dependence, emissions, or customer failures.

3. Master Production Scheduling

The MPS is a time-phased statement of when specific finished products will be completed. It is more detailed than the aggregate plan but usually less detailed than component-level MRP. An MPS must respect the aggregate plan, available-to-promise commitments, lead times, lot-sizing rules, and rough-cut capacity.

Important MPS terms include:

• Forecast: estimated demand independent of confirmed orders.
• Customer orders: committed demand received from customers.
• Projected available balance (PAB): expected inventory after considering demand and scheduled production.
• Available-to-promise (ATP): uncommitted inventory or production that can be promised to customers.
• Time fence: a boundary defining how freely the schedule may be changed.
• Firm planned order: an order protected from automatic rescheduling.

For a simple periodic system, projected available balance can be calculated as:

where MPS_t is the quantity scheduled for completion in period t and D_t is the demand used by the organization’s policy. Within a frozen time fence, customer orders may be accepted only if sufficient ATP exists. Outside the fence, the MPS can usually be changed more easily.

The MPS should be expressed in meaningful units. A family-level aggregate plan may specify 2,000 refrigerators per month, whereas the MPS may specify 800 white 300-litre units, 700 silver 300-litre units, and 500 premium units in particular weeks.

4. Bills of Materials and Material Requirements Planning

A bill of materials (BOM) is a structured description of the components, subassemblies, quantities, and relationships required to make one unit of a parent item. A single-level BOM lists immediate components; a multilevel BOM shows the complete product structure.

MRP uses the MPS, BOM, inventory records, scheduled receipts, lead times, and lot-sizing rules to calculate planned component orders. The fundamental logic is:

1. Start with the MPS or independent demand for a parent item.
2. Determine gross component requirements using usage quantities.
3. Net the requirements against projected inventory and scheduled receipts.
4. Apply lot-sizing rules.
5. Offset planned receipts by lead time to obtain planned order releases.
6. Explode the releases to lower BOM levels.

For a component used in quantity q per parent, gross requirements are:

where GR is gross requirement and PORel is the parent planned-order release. The timing is important: components are needed when the parent order begins production, not necessarily when the parent is shipped.

A simplified net-requirement calculation is:

where NR_t is net requirement, GR_t is gross requirement, SS is desired safety stock, OH_t is available on-hand inventory, and SR_t is usable scheduled receipt in the relevant period. The exact calculation should also account for earlier planned receipts, allocations, rejected material, and lot-sizing rules.

If the lot-for-lot rule is used, planned receipt equals net requirement. Under a fixed-order-quantity rule, the planned receipt is the smallest permitted multiple that covers the net requirement. Under economic order quantity, the order quantity may be estimated as:

where D is annual demand in units per year, S is ordering or setup cost per order, and H is annual holding cost per unit per year. This classical expression assumes relatively stable demand, instantaneous replenishment, no quantity discounts, and no shortages. It should not be used uncritically where setup times, capacity, perishability, or supplier minimums dominate.

Worked MRP example

Suppose a weekly MPS requires 50 units of a finished product in week 3. Each finished unit uses two units of component C. Therefore, the gross requirement for C in week 3 is:

There are 30 units of C on hand, no scheduled receipt, and safety stock is zero. Net requirement is 100 − 30 = 70 units. If the supplier uses a fixed order quantity of 100 units and has a one-week lead time, the planned receipt is 100 units in week 3 and the planned release is 100 units in week 2. The excess 30 units remain as projected inventory after the requirement.

MRP is a planning logic, not a guarantee of physical availability. Its quality depends on accurate BOMs, inventory records, lead times, transaction discipline, and appropriate nervousness controls.

5. Capacity Requirements Planning

CRP translates planned orders into hours at specific work centres and compares those requirements with available capacity. It is more detailed than rough-cut capacity planning, which checks major resource groups during MPS development.

For operation i, required load may be calculated as:

where Q_j is the number of orders of product or job j, t_s is setup time per order, q_j is quantity processed, and t_r is run time per unit. The summation includes all jobs assigned to work centre i. If setup time is already included per batch, it must not be multiplied by every unit.

Available capacity may be estimated as:

where N is the number of resources, H is scheduled hours per resource, A is availability, E is performance efficiency, and S is the fraction reserved for planned preventive maintenance or other exclusions. All factors must be dimensionless except H, so the result is in resource-hours.

A utilization indicator is:

A value above 1 indicates overload under the stated assumptions. Utilization below 1 does not automatically prove feasibility because variability, sequence-dependent setups, due dates, and material shortages may still create delays.

Capacity responses include overtime, alternate routing, subcontracting, preventive-maintenance changes, cross-training, lot splitting, additional equipment, and demand rescheduling. A useful practice is to identify the bottleneck resource and protect it from unnecessary interruptions. However, maximizing utilization everywhere can increase queues and lead times; flow performance must be considered across the system.

6. Enterprise Resource Planning

ERP integrates planning and transaction data across functions such as sales, purchasing, inventory, production, quality, finance, maintenance, and human resources. MRP is generally a central production-planning logic within an ERP environment.

ERP benefits include a common database, transaction visibility, coordinated master data, standardized workflows, and improved traceability. Risks include expensive implementation, poor data migration, excessive customization, cybersecurity exposure, and the mistaken belief that software can compensate for weak processes.

Critical master data include item identifiers, units of measure, BOMs, routings, work-centre calendars, lead times, safety stocks, lot sizes, supplier records, and inventory balances. Governance should define who may change each data element, how changes are approved, and how errors are detected. Barcode, RFID, industrial Internet of Things sensors, and machine integration can improve data timeliness, but automated data are not automatically accurate.

Artificial intelligence can support demand forecasting, anomaly detection, predictive maintenance, dispatching, and scenario analysis. Human oversight remains necessary because models may reproduce historical bias, fail during structural changes, or optimize local measures at the expense of safety, quality, or contractual commitments. Planning systems should provide explanations, audit trails, access controls, and override procedures.

7. Scheduling Environments

Scheduling assigns work to resources and time periods while considering processing times, due dates, release dates, setups, priorities, labour skills, and machine availability.

Job shops produce varied products in low volumes. Jobs follow different routings through functional departments. The main challenge is coordinating many possible routes and variable workloads.

Batch operations produce groups of similar items. Setup time and sequence-dependent changeovers are significant. Batches reduce setup frequency but increase waiting and work-in-process if they are too large.

Flow shops process products through a common sequence of work centres. Repetitive manufacturing and assembly lines are examples. Line balancing and takt-time analysis are important.

Service operations schedule people, rooms, vehicles, or appointments rather than only physical products. Demand often arrives in real time, customer waiting is visible, and service duration may be uncertain.

Scheduling objectives and measures

Common objectives include minimizing completion time, lateness, tardiness, work-in-process, overtime, or total cost. For job j, completion time is C_j and due date is d_j. Lateness is:

Tardiness cannot be negative:

Average flow time is the mean time jobs spend in the system, often measured from release to completion. Makespan is the completion time of the final job. No single sequence rule optimizes every objective.

Sequencing rules

• First-come, first-served (FCFS): processes jobs in arrival order; perceived as fair but may create long waits.
• Shortest processing time (SPT): processes the shortest job first; often reduces average flow time, but long jobs may wait.
• Earliest due date (EDD): processes the nearest due date first; useful for reducing maximum lateness under suitable single-machine assumptions.
• Critical ratio:

where t is the current time and p_j is remaining processing time. Smaller ratios generally receive higher priority; a ratio below 1 indicates that the job is behind its nominal schedule if no recovery occurs.

• Slack per operation: prioritizes low remaining slack divided by remaining operations, useful when jobs have multiple stages.

Rules are dispatching heuristics, not universal laws. They should be evaluated against due-date performance, setup costs, priority customers, fairness, quality, and safety.

Gantt charts and finite scheduling

A Gantt chart displays jobs or activities against time. It shows planned start and finish times, machine occupancy, idle intervals, maintenance, and often progress. A schedule is infinite-capacity when operations are placed according to material and precedence logic without checking whether a resource is assigned conflicting work. A finite-capacity schedule does not assign overlapping work to the same constrained resource.

Finite scheduling may be forward or backward. Forward scheduling starts as soon as possible and estimates completion. Backward scheduling starts from a due date and places operations as late as feasible, reducing early inventory but leaving less protection against disruption. Practical systems often combine both approaches with frozen zones and rescheduling rules.

A Gantt chart is a communication and control tool, not by itself an optimization method. It must be updated using actual completions, downtime, quality holds, and material status.

8. Workforce Planning and Line Balancing

Workforce planning aligns labour quantity and skills with expected workload. Decisions include staffing levels, shifts, overtime, temporary labour, cross-training, job rotation, and subcontracting. Managers should account for legal working-hour limits, fatigue, breaks, equal-opportunity requirements, and health and safety.

For a paced assembly line, takt time is the available production time divided by customer demand:

If a shift provides 27,000 seconds of productive time and demand is 450 units, takt time is 60 seconds per unit. Takt time is a demand-pacing concept, not necessarily the actual processing time of every task.

If total task time is ∑t_i and takt time is T, the theoretical minimum number of stations is:

The ceiling function rounds up to the next whole station. Actual station count may be higher because tasks cannot be divided, precedence constraints exist, or worker skills and equipment differ. Line efficiency can be expressed as:

where N is the actual number of stations. This measure does not capture ergonomic quality, reliability, or uneven demand.

9. Detailed Illustrative Case: Apex Pumps

The following is a fictional illustrative case.

Background

Apex Pumps assembles industrial coolant pumps in three product variants. Monthly demand is seasonal, rising from 800 units in April to 1,300 units in June. The company has 30 assemblers, each capable of 40 units per month under normal conditions. The final assembly line is the bottleneck, while machining has spare capacity. Two pump variants share most components, but the premium variant requires a special impeller supplied with a two-week lead time.

Problem

Sales has accepted June orders without checking the MPS. Inventory records for the special impeller are inaccurate, and the existing schedule releases all variants in large batches. The result is overtime, late deliveries, and excessive work-in-process.

Analysis

Aggregate analysis shows normal monthly capacity of 30 × 40 = 1,200 units. June demand of 1,300 therefore creates a 100-unit gap before considering beginning inventory. A mixed aggregate plan is selected: maintain the core workforce, use limited overtime in June, and subcontract 50 standard units if supplier quality is verified.

The MPS then assigns weekly quantities by variant and protects a time fence for the next two weeks. The BOM identifies one special impeller per premium pump. MRP nets the MPS against the corrected inventory record and schedules the impeller release two weeks before assembly. CRP shows that final assembly is overloaded during the second June week, so 30 units are moved to the first week and the remaining 50 standard units are subcontracted.

A finite schedule sequences products to reduce changeovers: standard variants are grouped, but the plan avoids a batch so large that premium orders become late. The shop uses a Gantt chart showing assembly, inspection, maintenance, and supplier-receipt milestones. Cross-trained machining employees support kitting and inspection during the peak.

Outcome

The revised plan meets the committed premium orders, reduces peak assembly load, and limits overtime. Inventory accuracy improves after cycle counting and barcode confirmation at receipt and issue. The company also establishes a rule that sales orders inside the frozen time fence require operations approval.

Learning points

1. Aggregate capacity decisions must precede detailed promises.
2. MPS quantities must be exploded through an accurate BOM.
3. MRP cannot correct inaccurate inventory or lead-time data.
4. CRP should identify bottlenecks before schedules are released.
5. Finite scheduling and sensible sequencing reduce avoidable congestion.
6. Governance between sales, operations, purchasing, and finance is essential.

10. Challenges, Limitations, and Best Practices

Planning models simplify reality. Forecasts are uncertain, processing times vary, suppliers fail, workers are absent, machines break down, and customers change orders. Excessive schedule changes create nervousness, while excessive stability can preserve an obsolete plan. Safety stock and capacity buffers provide protection but increase cost.

Best practices include:

• use a documented sales and operations planning cycle;
• maintain one agreed demand and inventory picture;
• separate frozen, slushy, and liquid planning zones;
• audit BOMs, routings, calendars, and inventory transactions;
• validate MRP plans through capacity checks;
• identify bottlenecks and protect them from avoidable variation;
• use small, controlled batches where setup and quality permit;
• measure delivery reliability, flow time, inventory, overtime, and schedule adherence together;
• involve employees in realistic standard times and ergonomic decisions;
• maintain contingency plans for critical materials and resources; and
• govern algorithmic decisions with human review, security, and traceability.

Environmental performance should also be considered. Large batches may reduce setups but increase obsolete inventory and energy use. Overtime and expedited freight may improve delivery while increasing emissions and fatigue. Sustainable planning evaluates total system consequences rather than a single cost measure.

Chapter Summary

Aggregate planning connects medium-term demand with workforce, inventory, overtime, subcontracting, and capacity decisions. Level, chase, and mixed strategies offer different trade-offs between stability, inventory, flexibility, and cost. The MPS converts family-level plans into specific end-item quantities and dates. BOMs provide product-structure logic, while MRP calculates time-phased component requirements using gross requirements, netting, lot sizing, and lead-time offsets. CRP checks whether work-centre capacity can support planned orders.

ERP integrates these planning activities with purchasing, sales, inventory, finance, quality, and human resources, but its effectiveness depends on accurate master data and disciplined processes. Scheduling assigns work to resources using Gantt charts, finite-capacity logic, and sequencing rules such as FCFS, SPT, EDD, and critical ratio. Workforce planning and line balancing align skills and labour capacity with demand. Effective practice combines quantitative analysis, cross-functional governance, employee involvement, technology, and ethical attention to safety, fairness, privacy, and sustainability.

Key Terms

Aggregate planning: Intermediate-term planning of production, workforce, inventory, and related resources for product families.

Back order: Customer demand intentionally delayed until a later period.

Bill of materials (BOM): Structured list of components and quantities required to make a parent item.

Capacity requirements planning (CRP): Process of comparing planned work-centre loads with available capacity.

Chase strategy: Aggregate plan that varies production or workforce to follow demand.

Critical ratio: Priority measure comparing remaining time until due date with remaining processing time.

Dependent demand: Demand calculated from the requirements of another item.

Enterprise resource planning (ERP): Integrated information system linking organizational planning and transactions.

Finite scheduling: Scheduling that prevents simultaneous assignment of conflicting work to a constrained resource.

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

Level strategy: Aggregate plan using relatively stable production and workforce levels.

Master production schedule (MPS): Time-phased plan for specific finished products.

Material requirements planning (MRP): Logic for calculating component requirements and planned order releases.

Net requirement: Remaining material requirement after considering inventory and scheduled receipts.

Projected available balance: Expected inventory position after planned receipts and demand.

Rough-cut capacity planning: Preliminary capacity check of major resources during MPS development.

Sequencing rule: Procedure for determining the order in which jobs are processed.

Takt time: Available production time divided by required customer output.

Time fence: Planning boundary that limits schedule changes during a defined interval.

Utilization: Required resource load divided by available resource capacity.

Review Questions

Short Answer Questions

1. Distinguish aggregate planning from master production scheduling.
2. What is the difference between independent and dependent demand?
3. Explain the purpose of a bill of materials in MRP.
4. What information does a Gantt chart communicate?
5. Why can utilization below 100% still be associated with late deliveries?

Long Answer Questions

1. Compare level, chase, and mixed aggregate-planning strategies, including their cost and workforce implications.
2. Explain the complete MRP process from the MPS through BOM explosion and planned-order release.
3. Describe how CRP supports MPS validation and discuss possible responses to a capacity overload.
4. Compare job-shop, batch, flow-shop, and service scheduling environments.
5. Explain how ERP systems improve planning integration and identify their principal implementation risks.

Analytical / Application Questions

1. A department has 18 workers, each producing 35 units per week. Beginning inventory is 60 units, and demand over three weeks is 500, 700, and 650 units. Calculate the inventory balance each week if the workforce remains constant, and identify the required response if shortages are not permitted.
2. A parent order releases 120 units in week 4. Each parent uses three components. There are 100 components on hand, no safety stock, a fixed order quantity of 300, and a two-week lead time. Determine the gross requirement, planned receipt, and planned release week.
3. Three jobs are available at one machine. Their processing times are 6, 2, and 4 hours, and their due dates from the current time are 10, 5, and 8 hours. Compare the FCFS sequence 1–2–3 with the SPT sequence and calculate completion times and total tardiness for each, assuming all jobs are available now and no setup time exists.

Multiple Choice Questions with Answers

Question 1

Which planning level specifies quantities and timing for specific finished products?

A. Strategic capacity planning

B. Aggregate planning

C. Master production scheduling

D. Supplier qualification

Correct Answer: C. Master production scheduling

Question 2

A product requires four units of component X per finished unit. An MPS release is 75 finished units. What is the gross requirement for X?

A. 19 units

B. 79 units

C. 300 units

D. 375 units

Correct Answer: C. 300 units

Question 3

Which strategy generally maintains a relatively stable workforce and uses inventory to absorb demand variation?

A. Chase strategy

B. Level strategy

C. Random dispatching

D. Backward scheduling only

Correct Answer: B. Level strategy

Question 4

If available capacity is 800 hours and required load is 960 hours, utilization is:

A. 0.83 or 83%

B. 1.00 or 100%

C. 1.20 or 120%

D. 1.76 or 176%

Correct Answer: C. 1.20 or 120%

Question 5

Which document describes the components and quantities needed to make a parent item?

A. Bill of materials

B. Gantt chart

C. Cash budget

D. Control chart

Correct Answer: A. Bill of materials

Question 6

Under the shortest-processing-time rule, which job is normally given priority?

A. The job with the earliest arrival date only

B. The job with the longest processing time

C. The job with the shortest processing time

D. The job with the largest batch size

Correct Answer: C. The job with the shortest processing time

Question 7

A component has a gross requirement of 250 units, 80 usable units on hand, and no safety stock or scheduled receipt. Its net requirement is:

A. 80 units

B. 170 units

C. 250 units

D. 330 units

Correct Answer: B. 170 units

Question 8

What is the principal purpose of finite-capacity scheduling?

A. To ignore machine conflicts

B. To prevent conflicting assignments on constrained resources

C. To eliminate all setup times

D. To replace the bill of materials

Correct Answer: B. To prevent conflicting assignments on constrained resources

Question 9

A line has 24,000 seconds of available production time and demand is 400 units. Its takt time is:

A. 40 seconds per unit

B. 60 seconds per unit

C. 240 seconds per unit

D. 9,600 seconds per unit

Correct Answer: B. 60 seconds per unit

Question 10

Which statement about ERP is most accurate?

A. ERP removes the need for accurate master data

B. ERP integrates planning and transaction information across functions

C. ERP is limited to machine maintenance

D. ERP guarantees an optimal schedule without human judgment

Correct Answer: B. ERP integrates planning and transaction information across functions