Dependent Demand and Product Structure Management

Introduction

Inventory control becomes substantially more effective when managers distinguish between demand that originates in the marketplace and demand that is calculated from a production plan. Independent demand is demand for an item that is determined primarily by external customers or markets. Finished products, service parts, and replacement units commonly have independent demand. Dependent demand is demand for an item that is derived from the requirement for another item. Components, subassemblies, raw materials, packaging, and manufacturing supplies often have dependent demand.

This distinction is central to material requirements planning (MRP). Independent demand is generally forecast, monitored through customer orders, or managed with replenishment policies such as reorder points and safety stock. Dependent demand is calculated by exploding a product structure: the system determines which components are needed, in what quantities, and when they must be available. A weak product structure, inaccurate lead time, or poorly governed engineering change can therefore produce shortages even when the MRP logic is mathematically correct.

This chapter examines the information architecture and managerial controls required for dependable dependent-demand planning. It addresses bills of material, product structures, low-level codes, lead times, engineering changes, scrap and yield, lot traceability, and planning master data.

Learning Objectives

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

1. Distinguish independent demand from dependent demand and select appropriate planning methods.
2. Construct and interpret bills of material and multilevel product structures.
3. Explain low-level coding and its role in MRP explosion.
4. Calculate gross component requirements, adjusted requirements, and timing requirements.
5. Incorporate lead times, scrap, yield, lot sizes, and safety stock into planning.
6. Control engineering changes, effectivity, revisions, and obsolete inventory.
7. Design lot-traceability and master-data governance practices.
8. Evaluate the managerial, operational, and ethical consequences of inaccurate product data.

1. Independent and Dependent Demand

Independent demand cannot normally be derived with certainty from the requirements of another inventory item. It must be estimated from forecasts, customer orders, market intelligence, or service-level policies. Examples include demand for finished bicycles, replacement pumps, and retail packages.

Dependent demand is logically linked to a parent item. If one finished product requires four bolts, the planned requirement for bolts is related to the planned production quantity of the finished product. In a deterministic setting, the relationship is structural rather than probabilistic.

The distinction is not absolute. A component may be dependent in one context and independent in another. A standard motor used in a configured product may be planned from a parent schedule, but the same motor may also be sold as an aftermarket replacement. In that case, the organization may need separate demand streams and a carefully designed planning policy.

A common error is to forecast every component independently. This can create inconsistent requirements: the forecast may imply one quantity of finished products while component forecasts imply another. MRP reduces this inconsistency by deriving component requirements from a coordinated master production schedule (MPS), subject to valid product data and planning parameters.

2. Product Structures and Bills of Material

A bill of material (BOM) is a formally controlled list of all components, ingredients, materials, and sometimes operations required to make, assemble, or package one unit of a parent item. A BOM normally specifies:

• parent item and revision;
• component item identifiers;
• quantity per parent;
• unit of measure;
• scrap or yield information;
• effectivity dates or serial-number ranges;
• phantom, co-product, or bulk-material status;
• engineering and manufacturing references; and
• revision and approval status.

A product structure represents the hierarchical relationship among items. For example, a finished assembly may contain a frame and a drive module. The drive module may contain a motor, controller, and fasteners. The hierarchy is important because requirements at one level become inputs to the next level.

Figure: A multilevel product structure. MRP explodes requirements from the finished product toward lower-level components.

Single-level and multilevel BOMs

A single-level BOM lists only the immediate components of a parent. A multilevel BOM includes all levels beneath the parent. A planning BOM may be designed for forecasting families, while an engineering BOM describes design relationships and a manufacturing BOM describes how the factory actually builds the item. These representations may differ, but their relationships must be reconciled through governance and controlled transformation.

A phantom item is a logical grouping used for planning or engineering convenience but not normally stocked or scheduled as a separate physical item. During MRP explosion, its requirements may pass directly to its components. Phantom use simplifies structures but can obscure responsibility if effectivity and component ownership are poorly maintained.

3. MRP Explosion and Low-Level Codes

Suppose a parent item has planned production quantity Q_p and requires q_c units of component per parent. Ignoring inventory, scheduled receipts, scrap, and lot-size effects, the gross requirement is:

Here, GR_c is the gross requirement for the component, Q_p is the planned quantity of the parent, and q_c is the usage quantity. The formula assumes that every parent unit has the same structure and that the component is required in the same period as the parent requirement is exploded.

For a component used at multiple levels, the extended usage is the product of the usage quantities along every path from the top-level item to that component. If component C is used twice in subassembly B, and B is used three times in product A, then each unit of A requires:

If 100 units of A are planned, the gross requirement for C is 600 units before other adjustments.

A low-level code (LLC) is the lowest level at which an item appears anywhere in a product structure. The top-level item is commonly assigned code 0. An item appearing directly below it may have code 1; an item used below multiple parents receives the deepest applicable code. MRP processes items in low-level sequence so that all parent requirements are consolidated before lower-level requirements are calculated. Without this sequencing, a component shared by multiple assemblies could be planned incompletely or repeatedly.

Net requirements

A simplified net-requirement calculation is:

where NR_t is net requirement in period t, GR_t is gross requirement, SS_t is desired safety stock, PAB_t is projected available balance before the new receipt, and SR_t is usable scheduled receipt. The expression assumes that all quantities are in the same unit and period and that scheduled receipts are available when stated. Actual systems may apply allocated stock, firm planned orders, reservations, negative balances, and lot-sizing rules in a more detailed sequence.

Worked example: multilevel explosion

An assembly plan calls for 80 units of product A. Each A requires 2 units of subassembly B and 1 unit of component D. Each B requires 3 units of component C and 1 unit of D.

• Gross requirement for B: 80 × 2 = 160 units.
• Gross requirement for C: 160 × 3 = 480 units.
• Direct requirement for D: 80 × 1 = 80 units.
• Requirement for D through B: 160 × 1 = 160 units.
• Total requirement for D: 80 + 160 = 240 units.

This example illustrates why the system must consolidate requirements for a shared component. Planning only the direct relationship would understate D by 160 units.

4. Lead Times, Offsetting, and Time-Phased Planning

MRP is not merely a quantity calculation. It is a time-phased plan. A component must be available before the operation or assembly that consumes it. Lead time is the elapsed time between release of an order and availability of the resulting item. It may include queue, setup, processing, inspection, move, receiving, and supplier-transit time.

If a component is required in period t and its effective lead time is L, the planned release period is conceptually:

where PR is the planned release period. The formula assumes a fixed lead time and discrete planning periods. In practice, calendars, weekends, holidays, transportation lanes, capacity constraints, and supplier cutoffs may make date arithmetic more complex.

Lead-time data should distinguish manufacturing, purchasing, inspection, and interoperation elements when the planning system supports them. A single inflated number can protect against some uncertainty but may hide the real cause of delay and increase inventory. Conversely, an unrealistically short lead time creates systematic lateness and expediting.

Lead times should be measured rather than guessed. Useful measures include median and high-percentile actual elapsed time, schedule adherence, queue time, supplier confirmation time, and receiving-to-available time. Governance should define who may change lead times, what evidence is required, and how frequently parameters are reviewed.

5. Scrap, Yield, and Lot Sizing

A BOM quantity describes nominal usage, but actual production may require additional material because of scrap or imperfect yield.

If the expected scrap fraction is s, where 0 ≤ s < 1, and the process consumes one unit of input for each nominal unit of output, the planned input requirement is:

For an expected output of 1,000 units and a scrap rate of 4%:

If whole units are required, the organization must round according to its operational rules, normally upward to 1,042 units. The calculation assumes a stable average scrap rate and independence between units. It does not guarantee that every batch will achieve the average.

Yield is often expressed directly as y, where 0 < y ≤ 1. The equivalent formula is:

A 96% yield produces the same result as a 4% scrap rate. However, “scrap percentage” and “yield” must not be confused when the organization defines them using different denominators.

Scrap can occur at a component level, operation level, or batch level. Double counting must be avoided. If a 2% loss is already embedded in a routing operation and also added to the BOM, requirements will be overstated.

Lot sizing converts net requirements into order quantities. Common policies include lot-for-lot, fixed quantity, minimum order quantity, fixed-period quantity, and economic order quantity. Lot sizing can reduce ordering or setup costs but may increase inventory, obsolescence, and exposure to engineering changes. For dependent demand, lot sizing should be evaluated with the parent schedule and capacity constraints rather than treated as an isolated purchasing decision.

6. Engineering Changes and Effectivity Control

An engineering change order (ECO) formally authorizes a change to design, specification, process, or product structure. Effective change management is essential because inventory may exist in several states: raw material, work in process, finished goods, supplier stock, and customer-installed product.

A controlled change process normally includes:

1. Request and justification: identify the problem, benefit, risk, and affected items.
2. Impact analysis: evaluate cost, quality, safety, regulatory compliance, capacity, suppliers, inventory, service parts, and customer commitments.
3. Disposition decision: determine whether existing stock is used, reworked, scrapped, returned, or segregated.
4. Approval: obtain engineering, operations, quality, supply-chain, finance, and—where relevant—regulatory authorization.
5. Effectivity definition: specify the date, serial number, lot, batch, or order from which the revision applies.
6. System implementation: update BOMs, routings, item revisions, drawings, work instructions, and planning parameters.
7. Verification: confirm that the released structure produces the intended material plan.

Effectivity by date is simple but can be unsafe when production orders cross the change date. Serial- or lot-based effectivity offers stronger traceability where product safety or regulatory compliance is important. A change should never be implemented in the BOM before the organization has decided how to control old and new inventory.

Engineering changes create an inventory transition problem. Continuing to buy the old component can create obsolescence; switching too early can make existing work orders unusable. MRP must therefore be integrated with change-control records, open-order review, and disposition decisions.

7. Lot Traceability and Genealogy

Lot traceability is the ability to identify the origin, movement, transformation, and destination of materials by lot, batch, serial number, or other controlled identity. Forward traceability asks where a lot was used. Backward traceability asks which supplier lots and processes contributed to a finished item.

Traceability is particularly important for food, pharmaceuticals, aerospace, medical devices, chemicals, and safety-critical products. It supports recalls, nonconformance containment, warranty analysis, regulatory reporting, and root-cause investigation.

A robust traceability design defines:

• the identity unit: lot, batch, serial, or container;
• points of capture: receiving, issue, transformation, return, shipment, and disposal;
• parent-child genealogy between consumed and produced lots;
• retention period and data ownership;
• rules for split, merge, dilution, rework, and substitution; and
• controls against unauthorized lot substitution.

Traceability is not achieved merely by printing labels. Transactions must connect physical movement to system records. Barcode, RFID, warehouse-management systems, manufacturing-execution systems, and supplier portals can improve accuracy, but technology cannot compensate for ambiguous item definitions or undisciplined transaction practices.

8. Planning Master Data Governance

Planning master data is the set of values used by MRP and related systems to calculate, prioritize, and release supply. Important fields include item identifiers, descriptions, units of measure, procurement type, make-or-buy status, lead times, order policies, minimum and maximum quantities, safety stock, calendar, supplier, location, planner code, lot-control status, yield, scrap, and BOM relationships.

Data quality has several dimensions:

• Accuracy: the value reflects reality.
• Completeness: required fields are populated.
• Consistency: related records use compatible definitions and units.
• Timeliness: changes are implemented before they affect planning.
• Traceability: the source and approval history are known.
• Uniqueness: duplicate item records are prevented.

Units of measure are a frequent source of error. If a supplier sells kilograms while production consumes grams, the conversion must be explicit, controlled, and dimensionally consistent. Similar risks arise with pack sizes, length-based material, yield-based ingredients, and variable-weight products.

A practical governance model assigns ownership by data domain. Engineering owns design structures; manufacturing engineering owns routings and process losses; purchasing owns supplier and commercial parameters; operations planning owns planning policies; quality owns inspection and traceability rules; and a data-governance council resolves conflicts. Segregation of duties is important: the person requesting a change should not necessarily be the sole person approving and implementing it.

Useful data-quality indicators include BOM accuracy, percentage of items with reviewed lead times, invalid-unit incidents, planned-order exception rates, emergency purchase frequency, and inventory caused by obsolete revisions. These indicators should be connected to operational outcomes rather than treated as purely administrative measures.

9. Integrated Applied Case: Illustrative Electric Pump Assembly

Background

The following is an illustrative case. A manufacturer produces an electric pump. Each finished pump requires one drive module and one housing assembly. Each drive module requires one motor and one controller. Each housing assembly requires one housing and four seals. The plant plans to produce 500 pumps in week 6.

The motor has a two-week purchasing lead time, the controller has a one-week purchasing lead time, and the drive module has a one-week assembly lead time. The housing and seals are available internally with one-week and zero-week lead times respectively. The drive module requires two seals for assembly testing in addition to the seals installed in the finished pump. The planned scrap rate for seals is 5% at the testing operation.

Problem

The planner must determine gross requirements and release timing. A recent controller revision is also scheduled to become effective for pumps beginning in week 7. Existing controllers are usable in week 6 but must not be issued to week-7 production.

Analysis

For 500 pumps:

• Drive modules: 500 × 1 = 500.
• Motors: 500 × 1 = 500.
• Controllers: 500 × 1 = 500.
• Housings: 500 × 1 = 500.
• Seals before scrap adjustment: installed seals 500 × 4 = 2,000, plus testing seals 500 × 2 = 1,000; total = 3,000.
• Seal input after 5% scrap: 3,000 ÷ 0.95 = 3,157.89, rounded upward to 3,158 seals.

The drive modules must be available for final assembly in week 6. With a one-week drive-module assembly lead time, their planned release is week 5. Motors are required for drive-module assembly in week 5. With a two-week purchasing lead time, motor orders should be released in week 3, subject to supplier calendars and existing receipts. Controllers are also required in week 5 and, with a one-week lead time, should be released in week 4.

The revision effectivity requires a separate check. The 500 controllers for week 6 may use the old revision if authorized, but procurement and inventory records must prevent old-revision controllers from being allocated to week 7. If the old revision is physically indistinguishable from the new one, the organization faces a serious traceability and configuration-control risk.

Approach and solution

The planner should:

1. freeze the week-6 parent schedule sufficiently to protect component timing;
2. explode the approved revision of the pump BOM;
3. apply the seal scrap factor once, at the testing operation;
4. offset component requirements using actual calendars;
5. segregate controller inventory by revision and effectivity;
6. review open purchase orders against the ECO; and
7. validate the resulting plan through a shortage and exception report.

Outcome

The plan identifies 500 motors and 500 controllers, released at different times, 500 housings, 500 drive modules, and 3,158 seals. More importantly, it exposes the configuration risk that a mathematically correct MRP plan could still issue the wrong controller revision. The case demonstrates that quantity, timing, revision, and traceability are inseparable planning dimensions.

Learning points

A product structure must be interpreted with its operations, calendars, effectivity rules, and loss factors. MRP output is not automatically a feasible or compliant execution plan. Professional judgment remains necessary for revision transitions, supplier constraints, and physical identification.

10. Technology, AI, and Contemporary Developments

Modern ERP, advanced planning and scheduling, manufacturing-execution, warehouse-management, and product-lifecycle-management systems increasingly exchange product and planning data through integrated platforms. Digital twins can link engineering structures to manufacturing and service information. RFID, machine vision, and IoT devices can capture lot movements and process conditions closer to the point of activity.

AI can assist with anomaly detection, duplicate-item identification, lead-time estimation, supplier-risk assessment, and prediction of scrap. However, AI-generated recommendations should not silently alter safety-critical BOMs or effectivity rules. Training data may reflect historical errors, and a statistically likely substitution may be unacceptable for regulatory or engineering reasons. Human approval, explainability, access control, and audit logs are essential.

Contemporary organizations are also managing product variants, mass customization, circular-economy requirements, remanufacturing, and sustainability information. Product structures may need to represent recycled content, carbon data, repairable modules, disassembly relationships, and reverse logistics. These developments expand the meaning of a BOM from a production list into a lifecycle information model.

11. Challenges, Limitations, and Governance Risks

MRP assumes that the master production schedule, BOM, inventory records, lead times, and planning parameters are sufficiently reliable. It does not independently discover physical reality. Common failure modes include:

• inaccurate on-hand balances caused by unrecorded scrap or consumption;
• duplicate item numbers and inconsistent units of measure;
• obsolete BOM revisions remaining active;
• lead times that ignore queues or supplier variability;
• unapproved substitutions;
• safety stock used to conceal structural data problems;
• phantom assemblies misunderstood by planners;
• lot genealogy broken by manual transactions; and
• excessive system messages that cause planners to ignore genuinely urgent exceptions.

Governance must also address ethical and regulatory issues. A planner should not conceal shortages by changing a lead time or BOM quantity without authorization. Safety, quality, and regulatory requirements take precedence over short-term schedule performance. Traceability records must be protected from unauthorized alteration, and personal or supplier-sensitive data should be governed according to applicable privacy and information-security requirements.

Best-Practice Framework

An effective dependent-demand environment generally follows these principles:

1. Use forecasts and customer orders for independent demand; derive component demand from authorized parent plans.
2. Maintain one controlled definition for each item, unit, revision, and effectivity rule.
3. Separate engineering, manufacturing, and planning views while preserving reconciliation among them.
4. Measure actual lead times, yields, and scrap rather than relying on inherited assumptions.
5. Process low-level codes and consolidate shared-component requirements correctly.
6. Link ECO approval to inventory disposition, open orders, and system implementation.
7. Capture lot or serial genealogy at every significant transformation and movement.
8. Monitor data-quality metrics and exception trends.
9. Use automation and AI to support—not replace—engineering, quality, and planning accountability.
10. Audit the physical process against system records through cycle counts, revision audits, and traceability exercises.

Chapter Summary

Independent demand originates outside the immediate product structure and is commonly forecast or managed through customer orders. Dependent demand is calculated from parent-item requirements and therefore depends on accurate BOMs, product structures, schedules, and planning data. A BOM defines the quantity and identity relationships among parents and components; a multilevel structure allows MRP to explode requirements through several levels.

Low-level codes ensure that shared components are processed after all relevant parent requirements have been consolidated. Net requirements combine gross demand with projected availability, scheduled receipts, and safety-stock policy. Lead-time offsetting converts requirements into planned releases, while scrap and yield adjustments recognize process losses. Lot-sizing policies determine the quantity ordered or produced but may trade setup efficiency against inventory and obsolescence.

Engineering-change control must coordinate revisions, effectivity, open orders, inventory disposition, and regulatory obligations. Lot traceability connects supplier inputs to production and customers, supporting recalls and root-cause analysis. Finally, planning master data requires formal ownership, validation, measurement, and audit. MRP is a decision-support system, not a substitute for disciplined operational governance.

Key Terms

Bill of material: Controlled list of components and quantities required for a parent item.

Dependent demand: Demand derived from the requirement for another item.

Engineering change order: Authorized record governing a design, process, or structure change.

Effectivity: Rule specifying when or where a revision becomes applicable.

Engineering BOM: Product structure describing design relationships.

Manufacturing BOM: Product structure describing how the factory builds an item.

Gross requirement: Total requirement before inventory and scheduled receipts are considered.

Independent demand: Demand determined by customers, markets, or external service needs.

Lead time: Elapsed time from order release to item availability.

Low-level code: The deepest product-structure level at which an item appears.

Lot-for-lot: Lot-sizing policy that orders the net requirement for each period.

Lot traceability: Ability to track material origin, transformation, and destination.

Material requirements planning: Time-phased system for calculating and scheduling dependent requirements.

Net requirement: Requirement remaining after usable supply and policy stock are considered.

Phantom item: Logical product-structure grouping normally not stocked independently.

Product structure: Hierarchical relationship among a product and its components.

Scrap factor: Planned allowance for material or output lost in a process.

Safety stock: Planned inventory held to protect against uncertainty or service risk.

Yield: Proportion of input expected to become acceptable output.

Review Questions

Short Answer Questions

1. What is the principal difference between independent and dependent demand?
2. Why are low-level codes required in multilevel MRP processing?
3. What information should an effective BOM contain besides component quantity?
4. How does an engineering change affect inventory and open purchase orders?
5. Distinguish scrap rate from yield.

Long Answer Questions

1. Explain how MRP converts a master production schedule into time-phased component requirements.
2. Discuss the relationship among BOM accuracy, lead-time accuracy, and inventory performance.
3. Compare engineering BOMs, manufacturing BOMs, and planning BOMs.
4. Describe a controlled engineering-change process, including effectivity and inventory disposition.
5. Evaluate the organizational responsibilities required for planning master-data governance.

Analytical / Application Questions

1. A product requires three units of subassembly X and two units of component Y directly. Each X requires four units of Y. If 250 products are planned, calculate the total gross requirement for Y and explain the consolidation logic.
2. A process must produce 2,400 acceptable units with an expected yield of 92%. Calculate the required input, round appropriately for whole units, and state the assumption behind the calculation.
3. A component is required in week 10, has a three-week lead time, and is subject to a one-week receiving and inspection period included in the stated lead time. Determine the planned release week and identify two practical factors that could invalidate the result.

Multiple Choice Questions with Answers

Question 1

Which item most commonly represents dependent demand?

A. A finished product sold to a retail customer

B. A replacement product forecast from market history

C. A component required by a planned assembly

D. A service contract purchased by a customer

Correct Answer: C. A component required by a planned assembly

Question 2

A product requires 2 units of B, and B requires 3 units of C. For 100 products, what is the gross requirement for C before other adjustments?

A. 100 units

B. 200 units

C. 300 units

D. 600 units

Correct Answer: D. 600 units

Question 3

What is the primary purpose of a low-level code?

A. To calculate supplier prices

B. To sequence MRP processing at the deepest occurrence of an item

C. To identify warehouse aisles

D. To determine product selling prices

Correct Answer: B. To sequence MRP processing at the deepest occurrence of an item

Question 4

If 1,000 acceptable units are required and the scrap rate is 4%, what input is required before rounding?

A. 960.00 units

B. 1,000.00 units

C. 1,040.00 units

D. 1,041.67 units

Correct Answer: D. 1,041.67 units

Question 5

A component is required in period 8 and has a lead time of two periods. Ignoring calendars, when should its planned order be released?

A. Period 5

B. Period 6

C. Period 7

D. Period 10

Correct Answer: B. Period 6

Question 6

Which practice best supports engineering-change control?

A. Changing the BOM immediately without reviewing open orders

B. Defining revision effectivity and deciding how existing inventory will be handled

C. Allowing each operator to choose the preferred revision

D. Deleting the old revision without retaining history

Correct Answer: B. Defining revision effectivity and deciding how existing inventory will be handled

Question 7

What does backward lot traceability identify?

A. Customers who received a finished lot

B. Future demand for a component

C. Supplier lots and inputs used in a finished product

D. The selling price of a finished item

Correct Answer: C. Supplier lots and inputs used in a finished product

Question 8

Which statement about a phantom item is correct?

A. It must always be stocked as a finished physical item

B. It is a logical grouping whose requirements may pass to its components

C. It is an item with no relationship to a BOM

D. It eliminates the need for effectivity control

Correct Answer: B. It is a logical grouping whose requirements may pass to its components

Question 9

If a process yield is 80% and 400 acceptable units are needed, what input is required before rounding?

A. 320 units

B. 400 units

C. 480 units

D. 500 units

Correct Answer: D. 500 units

Question 10

Which is primarily a planning master-data quality dimension?

A. Decorative appearance of the item label

B. Accuracy of lead times and units of measure

C. Number of product advertisements

D. Color of the warehouse floor

Correct Answer: B. Accuracy of lead times and units of measure