Inventory Records and MRP Data Foundations

Introduction

Material Requirements Planning (MRP) is often described as a calculation system that converts a master production schedule into planned orders. In practice, however, MRP is only as reliable as the data from which it calculates. A mathematically correct explosion of an inaccurate bill of material, an obsolete lead time, or an incorrect inventory balance produces operationally harmful recommendations. Thus, inventory records are not merely administrative files; they are the factual foundation of production decisions.

This chapter examines the principal data structures and control practices required for dependable MRP. It explains item masters, inventory status, scheduled receipts, allocations, lead times, safety parameters, scrap factors, and planning calendars. It also considers data ownership, accuracy measurement, exception management, system integration, and the managerial consequences of poor master data.

Learning Objectives

After studying this chapter, the reader should be able to:

1. Explain the role and structure of item master and inventory-record data.
2. Distinguish on-hand, available, allocated, quarantined, rejected, and scheduled-receipt quantities.
3. Apply lead-time, lot-sizing, safety-stock, scrap, and planning-calendar logic in MRP.
4. Evaluate data accuracy, integrity, timeliness, and governance.
5. Diagnose the causes and consequences of unreliable MRP data.
6. Design practical controls for maintaining trustworthy planning records.

1. MRP Data as an Integrated Information System

An MRP record links a planning item to its physical, commercial, and operational characteristics. Typical inputs include:

• item identification and description;
• unit of measure and conversion factors;
• make-or-buy status and approved suppliers;
• bill-of-material relationships;
• current inventory and inventory status;
• open purchase, production, and transfer orders;
• demand, allocations, and reservations;
• lead times and lot-sizing rules;
• safety stock, scrap, yield, and service parameters;
• calendars, workdays, and planning horizons.

These data are interdependent. A wrong unit of measure can distort both stock balances and order quantities. An incorrect parent-child relationship can create or omit requirements. A lead time stated in calendar days rather than working days can shift an order by several production periods. Data quality must therefore be viewed as a system property rather than as a collection of isolated fields.

Three dimensions are especially important:

A fourth dimension, completeness, is also essential. A record may contain accurate values but still be unusable if its lead time, unit of measure, or planning policy is missing.

2. The Item Master

The item master is the authoritative record for each material, component, subassembly, finished product, spare, packaging item, or service-related planning object. Each item should have a unique identifier. Descriptions may assist users, but descriptions should not be used as the primary key because spelling, abbreviations, and language variations create ambiguity.

2.1 Core item-master fields

Important fields normally include:

• unique item number;
• standard description and technical specification;
• item category and product family;
• base unit of measure;
• alternate units and conversion factors;
• procurement type: manufactured, purchased, transferred, or phantom;
• planner or planning group;
• approved sources and supplier information;
• revision or engineering-change status;
• lot-sizing rule;
• order, minimum, and maximum quantities;
• lead-time components;
• safety-stock or service-level policy;
• scrap or yield assumptions;
• storage, quality, hazardous-material, and traceability attributes;
• effective and obsolete dates.

The item master should distinguish technical identity from planning policy. For example, two items may have identical dimensions but different approved suppliers, regulatory status, or inspection requirements. Conversely, a single physical item should not receive multiple item numbers merely because it is stored in different locations unless location-specific control is genuinely required.

2.2 Units of measure and conversions

MRP quantities must be dimensionally consistent. If purchasing is in kilograms, production consumption is in metres, and inventory is recorded in rolls, the system must contain controlled conversion rules. Conversions should not be assumed to be constant when density, cut length, moisture, or packaging variation affects the relationship.

For a fixed conversion, if one purchasing unit contains c base units, then:

where Q_base is the quantity in the base unit and Q_purchase is the purchasing quantity. This relationship is valid only when c is stable and formally approved. Rounding rules must also be defined because fractional units may be impossible to buy or issue.

2.3 Item status and engineering change

Items require lifecycle states such as pending, active, restricted, phase-out, and obsolete. An obsolete item should not be automatically deleted: historical transactions, warranties, traceability, and audit requirements may require retention. Engineering changes should specify effective dates, supersession rules, remaining stock treatment, and whether old and new revisions may be used interchangeably.

3. Inventory Status and Nettable Supply

MRP must distinguish physical possession from usable supply. The quantity in a warehouse is not automatically available for planning. Materials may be reserved for a customer, held for inspection, damaged, blocked by quality, or located in a non-nettable store.

Common statuses include:

• Unrestricted or available stock: usable for approved requirements.
• Allocated or reserved stock: committed to a particular order or demand.
• Quality-hold stock: physically present but awaiting inspection or disposition.
• Rejected or scrapped stock: not legally or technically usable.
• Quarantine stock: segregated pending investigation.
• In-transit stock: shipped but not yet received and accepted.
• Consignment stock: physically present but owned by a supplier; its planning treatment depends on the agreement.
• Safety stock: a protected planning buffer, not necessarily physically separate.

A simplified available-balance relationship is:

Here, all quantities must be expressed in the same unit and time period. The equation is a planning representation, not an accounting definition. Whether safety stock is subtracted explicitly depends on the MRP system and policy. In systems that protect safety stock, net requirements are triggered when projected available inventory falls below the safety-stock target.

For period-by-period planning, projected available balance can be represented as:

where PAB_t is projected available balance in period t. The equation assumes that all receipts and requirements are correctly dated and that there is no unmodelled loss, substitution, or capacity restriction.

Allocations and reservations

An allocation is a commitment of supply to a specific demand, order, customer, or operation. Allocations prevent one demand from consuming material already promised to another. However, excessive or obsolete allocations make inventory appear unavailable and can generate unnecessary purchases. Allocation records therefore require release rules, expiry dates, and review of cancelled orders.

Inventory accuracy

Inventory accuracy should be assessed through cycle counting and transaction discipline rather than relying only on annual physical counts. A basic record-accuracy percentage is:

The tolerance must be stated, such as exact quantity, ±1 unit, or a value-based tolerance. A value-weighted measure may be more appropriate where high-value items carry greater risk, but it can conceal many errors in low-value critical components. Consequently, organisations should use both quantity-based and risk-based measures.

4. Scheduled Receipts and Supply Visibility

A scheduled receipt is an open supply commitment with a quantity and due date. It may be a purchase order, production order, transfer order, subcontract order, or approved inbound shipment. MRP normally treats scheduled receipts differently from planned orders because they represent existing commitments.

Each scheduled receipt should contain:

• order number and line number;
• item and revision;
• open quantity;
• due date and expected arrival date;
• supplier, source, or producing work centre;
• inspection and acceptance status;
• cancellation or rescheduling authority;
• remaining quantity after partial receipts.

An open order is not necessarily usable supply. A receipt that is late, rejected, incomplete, or still awaiting quality approval should be represented according to its realistic availability. Treating every open purchase order as certain supply produces systematic under-ordering and false confidence.

The distinction between order date, promised date, receipt date, inspection-completion date, and usable-date is particularly important. MRP should plan against the date at which the material can actually support production, not merely the date printed on a supplier document.

5. Demand, Allocations, and the Logic of Net Requirements

MRP distinguishes independent demand, such as customer orders and forecasts, from dependent demand, which is calculated from the requirements of a parent item. Gross requirements are derived after considering the master production schedule, bills of material, usage quantities, scrap, and timing offsets.

A simplified net-requirement calculation is:

Available supply may include projected opening inventory and valid scheduled receipts, but its definition must be consistent with the organisation’s treatment of allocations and safety stock. The maximum function prevents a negative order recommendation. A negative balance is instead carried as excess projected supply or used to offset later requirements, depending on the planning design.

MRP records should also distinguish firm demand from forecast demand. If both are included without a forecast-consumption rule, the same requirement may be counted twice. Forecast consumption may reduce forecast quantities when customer orders are received, but the policy must specify the time fence, matching logic, and treatment of cancellations.

6. Lead Times and Time Phasing

Lead time is the elapsed time between authorising an order and obtaining usable material. It may include queue, setup, processing, move, inspection, release, supplier, and transport time. For a purchased item, the usable lead time may be expressed as:

For a manufactured item, it may include:

These are additive models and assume the activities are sequential. If some activities overlap, the total is not simply their sum; the planning system must model the actual precedence and overlap.

Lead time is not the same as processing time. A component that takes two hours to machine may require ten working days from release to usable receipt because of queueing, batching, transport, and inspection. Using only touch time systematically produces late orders.

Fixed, variable, and measured lead times

• Fixed lead time is a policy value used when variation is not modelled explicitly.
• Variable lead time changes with quantity, batch size, supplier, or capacity load.
• Measured lead time is estimated from actual transaction history.

Historical averages can be misleading if demand, supplier, or routing conditions have changed. Median lead time may be more robust to exceptional delays, while a high-percentile lead time may be suitable for service-critical items. The selected parameter should reflect the cost of lateness and the cost of early inventory.

Lead-time offsets must use the correct calendar. If the system works in working days but a supplier promise is in calendar days, weekends and holidays must be translated explicitly. A planning calendar defines valid workdays, receiving days, production shifts, holidays, shutdowns, and period boundaries.

7. Safety Stock and Other Safety Parameters

Safety stock protects against uncertainty in demand, supply, or internal process performance. It is not a substitute for accurate records or disciplined scheduling. Excessive safety stock hides poor data and increases carrying cost; insufficient safety stock creates shortages.

For independent demand with approximately normal demand variation and a known replenishment lead time, a common special-case expression is:

where SS is safety stock, z is the service-factor corresponding to the desired cycle-service probability, and σ_LT is the standard deviation of demand during lead time. This formula assumes that lead time is fixed, demand variation is represented adequately by a standard deviation, and the service target is defined consistently.

If demand per period has standard deviation σ_d and lead time is a fixed L periods, then:

so that SS = z × σ_d × √L. This special case should not be applied when lead time itself has material variability. If demand and lead time are both random, a more general variance model is required, and assumptions about independence must be tested rather than presumed.

Other safety policies include fixed quantity, days of supply, minimum presentation stock, service-level optimisation, and risk-class rules. Criticality, substitution difficulty, supplier reliability, demand intermittency, and item value should influence policy selection.

8. Lot Sizing, Scrap, Yield, and Rounding

MRP does not necessarily recommend exactly the calculated net requirement. Lot-sizing rules convert requirements into feasible orders. Common rules include lot-for-lot, fixed quantity, minimum order quantity, fixed period quantity, economic order quantity, and order-up-to policies.

For a simple fixed-lot rule with lot size Q and net requirement N:

where ceil means rounding upward to the next whole lot. This formula assumes that excess quantity is permitted and that no maximum quantity or shelf-life constraint overrides the rule.

Scrap and yield must be represented carefully. If a process has a scrap fraction s, where 0 ≤ s < 1, and the required good output is G, the input requirement is:

For example, if 950 good units are required and expected scrap is 5%, the input requirement is 950 ÷ 0.95 = 1000 units. The 5% is not obtained by simply adding 5% to 950; that would produce 997.5 and would understate the input needed to yield 950 good units. The formula assumes a stable average scrap fraction and applies to a process where scrap is proportional to input. Known fixed losses require a different model.

Yield may be expressed as y = 1 − s. Then InputRequirement = G ÷ y. Scrap parameters should be differentiated by item, process, equipment, material grade, and operation where evidence supports such differentiation. A single corporate scrap percentage can be convenient but inaccurate.

9. Planning Calendars and Time Fences

A planning calendar determines when demand, supply, capacity, and order actions may occur. It should define:

• working and non-working days;
• shifts and production hours;
• supplier and receiving calendars;
• holidays and planned shutdowns;
• period definitions;
• shop-floor reporting cut-off times;
• planning, demand, and release time fences.

A time fence restricts automatic changes within a protected horizon. For example, near-term orders may require planner approval because material has been purchased, labour assigned, or customers promised delivery. Outside the fence, the system may freely reschedule according to new requirements.

Calendar misalignment is a frequent source of apparent MRP instability. A supplier may promise seven calendar days, while the system interprets seven production days. Similarly, a month-end cut-off may make a receipt appear to move by a full period even though the physical difference is one day.

10. Data Governance and Integrity Controls

Reliable MRP requires clearly assigned ownership. Engineering typically controls specifications and bills of material; purchasing controls supplier and commercial lead-time data; production controls routings and process yields; inventory control controls physical balances and statuses; finance controls valuation-related attributes; and planning controls policy parameters. Shared governance is necessary because no single function sees the whole data chain.

Useful controls include:

1. Validation rules: mandatory fields, permitted values, unit checks, date checks, and duplicate detection.
2. Referential integrity: every component, supplier, work centre, warehouse, and calendar referenced by a record must exist and be active.
3. Effective dating: revisions and planning changes require start and end dates.
4. Workflow approval: changes to critical parameters require documented authorisation.
5. Audit trails: the system records who changed what, when, and why.
6. Exception reports: focus attention on negative balances, overdue receipts, inactive sources, unusual scrap rates, and missing parameters.
7. Cycle counting: frequency is based on value, criticality, movement, and historical error.
8. Reconciliation: inventory, purchasing, production, sales, and finance records are compared routinely.
9. Change management: users understand the operational consequences of altered data.

A master-data key performance indicator should be interpreted carefully. A high percentage of completed fields does not prove that the values are correct. Quality measurement should combine completeness, accuracy, timeliness, consistency, and business impact.

11. Technology, Integration, and AI

Modern ERP, MRP, warehouse-management, manufacturing-execution, supplier-collaboration, and product-lifecycle systems increasingly exchange data automatically. Barcodes, radio-frequency identification, electronic data interchange, and shop-floor reporting can improve timeliness, but automation also propagates errors rapidly when item identities or transaction rules are wrong.

Analytics can identify unusual lead times, negative inventory, repeated quantity adjustments, or supplier performance deterioration. Machine-learning methods may predict lead-time distributions or scrap rates, but predictions should not replace governance. Models require clean historical data, stable definitions, monitoring for drift, and human accountability. A planner should be able to understand why a parameter changed and challenge an implausible recommendation.

Important controls for AI-supported planning include versioned models, explainable features, approval thresholds, segregation of duties, protection of supplier and customer data, and testing for systematic bias. An algorithm that prioritises high-volume items may neglect low-volume but safety-critical parts.

12. Illustrative Case: The Missing Valve Problem

Background

A fictional industrial-pump manufacturer experienced repeated shortages of a specialised valve despite apparently adequate inventory. The MRP system showed 240 valves on hand and two open purchase orders totalling 120 valves.

Problem

Production orders for several pump models were delayed. Investigation found that 80 valves were allocated to a cancelled customer order, 60 were in quality quarantine, and 40 belonged to an obsolete revision. The remaining 60 unrestricted valves were recorded in pieces, while the bill of material consumed them in kits containing two valves. In addition, the main supplier’s usable lead time had increased from 15 to 28 working days, but the item master still showed 15 days.

Analysis

The apparent supply of 360 valves was not equivalent to usable supply. The company had failed to distinguish on-hand from nettable stock and had not closed obsolete allocations. The unit-of-measure conversion also distorted requirements, while the outdated lead time caused late releases.

The investigation produced the following simplified reconciliation:

Thus, immediate and date-valid supply was 130 valves, not 360. If a production plan required 90 kits, the requirement was 180 valves. The shortage was therefore 50 valves before considering safety stock.

Approach and solution

The firm corrected inventory statuses, cancelled the obsolete allocation, separated revisions, and established a two-valve-per-kit conversion. Purchasing confirmed revised delivery dates, and planning changed the usable lead time to 28 working days. A supplier-performance review was initiated, and valves were classified as a critical purchased component. Cycle-count frequency was increased, and a workflow was introduced for closing allocations when customer orders were cancelled.

Outcome and learning points

Shortages declined because the system’s representation of usable supply became realistic. The central lesson is that MRP accuracy depends on transaction discipline and semantic clarity. “On hand,” “ordered,” and “available” are different concepts. Correcting the arithmetic without correcting the definitions would not have solved the problem.

13. Challenges and Limitations

MRP data parameters are estimates, not permanent truths. Demand changes, supplier capacity fluctuates, processes improve or deteriorate, and engineering revisions alter consumption. Excessive parameter maintenance can create instability, while infrequent review creates obsolescence.

Several limitations deserve attention:

• deterministic MRP may underrepresent uncertainty;
• safety stock calculations can be unreliable with intermittent demand;
• averages conceal variability and tail risk;
• planned orders may ignore finite capacity unless linked to capacity planning;
• data ownership disputes can delay corrections;
• automated rescheduling can create nervousness and unnecessary expediting;
• inventory accuracy may be high in quantity but poor in condition or revision;
• supplier promises may be optimistic or commercially biased.

Managers should therefore combine MRP outputs with supplier collaboration, capacity analysis, quality information, inventory observation, and professional judgement. MRP is a decision-support system, not an autonomous substitute for operational responsibility.

Best-Practice Checklist

Before relying on an MRP run, an organisation should confirm that:

• each item has one controlled identity and valid unit of measure;
• bills of material and revisions are effective-dated;
• on-hand quantities are reconciled and statuses are meaningful;
• allocations and reservations are current;
• scheduled receipts reflect realistic usable dates;
• lead times are measured, documented, and expressed in the correct calendar;
• safety parameters match item risk and demand behaviour;
• lot sizes, minimums, rounding, scrap, and yield are explicit;
• planning calendars and time fences are maintained;
• exceptions are reviewed by accountable planners;
• changes are approved, traceable, and periodically audited.

Chapter Summary

Inventory records provide the factual foundation for MRP. The item master establishes identity, units, sourcing, lifecycle, and planning policy. Inventory status distinguishes physically present material from material that is unrestricted, allocated, quarantined, rejected, obsolete, or otherwise unavailable. Scheduled receipts must represent realistic, date-valid supply rather than merely open paperwork.

Lead times must include the elapsed time to obtain usable material and must be aligned with working calendars. Safety stock protects against uncertainty but should be based on explicit assumptions and service objectives. Lot-sizing, rounding, scrap, and yield rules convert theoretical requirements into feasible orders. Planning calendars and time fences determine when requirements and actions can occur.

Effective MRP data governance combines validation, ownership, auditability, cycle counting, exception management, and cross-functional review. Technology and AI can improve visibility and prediction, but they cannot compensate for ambiguous definitions or weak transaction discipline. The managerial objective is not simply complete data; it is trustworthy data that produces feasible, explainable, and timely supply decisions.

Key Terms

Item master: Authoritative record describing a planning item and its attributes.

Inventory status: Classification indicating whether stock is usable, allocated, quarantined, rejected, or otherwise restricted.

Nettable inventory: Stock that MRP is permitted to treat as available supply.

Scheduled receipt: Open inbound or production supply with a quantity and due date.

Allocation: Commitment of supply to a specified demand or order.

Projected available balance: Expected inventory after adding receipts and subtracting requirements.

Independent demand: Demand originating from customers, forecasts, or external requirements.

Dependent demand: Demand calculated from the requirements of a parent item.

Lead time: Elapsed time from order release to usable receipt.

Planning calendar: Definition of valid working, receiving, production, and non-working periods.

Safety stock: Inventory or planning protection against uncertainty.

Service factor: Statistical factor associated with a selected service objective.

Lot sizing: Rule for converting net requirements into order quantities.

Scrap factor: Proportion of input expected to be lost or unusable.

Yield: Proportion of input converted into acceptable output.

Time fence: Horizon within which planning changes are restricted or controlled.

Referential integrity: Consistency between linked records and valid master references.

Cycle counting: Repeated physical verification of selected inventory records.

Master-data governance: Policies, responsibilities, controls, and workflows for maintaining data.

MRP nervousness: Frequent and undesirable changes in planned orders caused by data or planning variation.

Review Questions

Short Answer Questions

1. Why is an item number preferable to a text description as the primary identifier?
2. Distinguish on-hand inventory from nettable inventory.
3. What information should a scheduled receipt contain?
4. Why is processing time different from total lead time?
5. What is the purpose of a planning time fence?

Long Answer Questions

1. Explain how item-master fields, inventory status, scheduled receipts, and allocations interact in an MRP calculation.
2. Discuss the assumptions and limitations of safety-stock calculations based on demand standard deviation.
3. Describe a comprehensive governance system for maintaining accurate MRP data.
4. Explain how units of measure, lot sizing, scrap, yield, and rounding affect planned order quantities.
5. Evaluate the benefits and risks of using AI and analytics to maintain lead-time and scrap parameters.

Analytical / Application Questions

1. A component has a gross requirement of 950 good units and an expected scrap rate of 5%. Calculate the required input quantity and explain why simply adding 5% is inaccurate.
2. An item has 500 units physically present: 120 are allocated, 80 are quarantined, and 50 are obsolete. It also has a scheduled receipt of 200 units, but only 150 will be usable within the planning period. Determine the date-valid usable supply and identify the data controls that should be investigated.
3. A supplier’s average lead time is 20 working days, but its performance varies from 12 to 35 days. Recommend a parameter-review approach for a critical component and explain the trade-off between inventory and shortage risk.

Multiple Choice Questions with Answers

Question 1

Which quantity is most appropriate for immediate MRP netting?

A. All physically present inventory

B. Only unrestricted, usable inventory

C. All inventory plus cancelled orders

D. Only rejected inventory

Correct Answer: B. Only unrestricted, usable inventory

Question 2

A process has a 5% scrap fraction and requires 950 good units. What input quantity is required?

A. 900 units

B. 950 units

C. 997.5 units

D. 1,000 units

Correct Answer: D. 1,000 units

Question 3

Which field primarily prevents the same physical material from being represented ambiguously?

A. Unique item number

B. Planner name

C. Supplier payment term

D. Warehouse temperature

Correct Answer: A. Unique item number

Question 4

A scheduled receipt should be treated cautiously when it is open but

A. linked to an active item

B. dated in the future

C. late, rejected, or not yet usable

D. shown on a purchase order

Correct Answer: C. late, rejected, or not yet usable

Question 5

If demand standard deviation per period is σ_d and fixed lead time is L periods, the standard deviation during lead time is

A. σ_d ÷ L

B. σ_d × L

C. σ_d × √L

D. L ÷ σ_d

Correct Answer: C. σ_d × √L

Question 6

What is the principal purpose of safety stock?

A. To eliminate all cycle counting

B. To protect against demand or supply uncertainty

C. To replace an item master

D. To increase scrap deliberately

Correct Answer: B. To protect against demand or supply uncertainty

Question 7

Which control most directly verifies that a component referenced by a bill of material exists and is active?

A. Referential-integrity validation

B. Supplier discount analysis

C. Annual sales forecasting

D. Packaging inspection

Correct Answer: A. Referential-integrity validation

Question 8

A fixed lot size is 100 units and the net requirement is 230 units. Ignoring other constraints, the planned order quantity is

A. 200 units

B. 230 units

C. 300 units

D. 330 units

Correct Answer: C. 300 units

Question 9

Why must lead time be aligned with a planning calendar?

A. To convert all materials into kilograms

B. To account for working days, holidays, and shutdowns

C. To remove all supplier variation

D. To determine item descriptions

Correct Answer: B. To account for working days, holidays, and shutdowns

Question 10

Which statement best describes MRP nervousness?

A. A stable inventory balance over time

B. Frequent undesirable changes to planned orders

C. A formal quality inspection method

D. A fixed supplier contract

Correct Answer: B. Frequent undesirable changes to planned orders