MRP – Material Requirements Planning: Turning Demand into Time-Phased, Lot-Traceable Supply in Regulated Manufacturing

This topic is part of the SG Systems Global regulatory & operations glossary.

Updated October 2025 • Planning & Scheduling • ERP, MES, LIMS, QMS, WMS

Material Requirements Planning (MRP) is the time-phased computation that converts independent and dependent demand into dated purchase, production, and packaging orders matched to actual lead times, yields, and shelf life constraints. In regulated industries, the stakes are higher than “keep the line running.” MRP must respect BOM and versioned MMR/MBR rules, enforce component release and Hold/Release states, drive FEFO allocation, and generate orders that the plant can actually execute within validated capacities and finite constraints. Done right, MRP is not a spreadsheet—it is a governed algorithm embedded in the digital thread from Goods Receipt to Finished Goods Release, explaining every date, quantity, and lot choice in a way that auditors and supply chain leaders can both trust.

“MRP is the difference between guessing what to buy and proving what must arrive, in which quantity, to hit a validated schedule without compromising genealogy.”

1) What It Is and Where It Applies

MRP consumes demand from multiple sources—customer orders, safety stock policies, Kanban signals, and internal projects—and explodes that demand through effective BOM versions and process routings to determine what to buy, what to make, and when. In discrete and batch environments the explosion walks the MMR/MBR to identify required intermediates, packaging materials, and test steps that affect availability. The planning horizon must be long enough to cover cumulative lead time—supplier manufacturing + transport + quarantine + QA release—and short enough to maintain relevance as forecasts shift and MOC updates hit. In pharma, nutraceutical, and device operations, MRP is tightly coupled to shelf life and storage class constraints: it cannot simply “suggest” orders; it must ensure that the lots it will cause to be consumed remain within expiry through the entire routing, including hold time stages and transit. That makes FEFO and dynamic allocation first-class planning inputs rather than warehouse afterthoughts.

2) Regulatory Anchors & Data Integrity

While the MRP math itself is not licensed by a regulator, the effects of MRP decisions are. 21 CFR 211 expects control over components, labeling, and in-process materials; 21 CFR 820 expects control over purchasing and production and process controls; FSMA and HACCP drive hazard-based planning and preventive controls; and ICH Q10 stresses lifecycle management and change control. If your planning tool releases or modifies electronic instructions, reservations, or labels, Part 11 / Annex 11 controls apply: unique users, e-signatures with meaning, time sync, audit trails, and validated interfaces under CSV/GAMP 5. Data integrity matters because bad planning data begets nonconformances: expired lots issued, out-of-status materials consumed, or labels without current artwork. The minimum bar is this: the MRP record must reconstruct why an order was suggested (demand source, effective BOM, lead times, policies), what rules constrained it (shelf life, QA status, asset status availability), and who approved any overrides.

3) The Planning Data Model—What Drives the Math

Good MRP lives or dies on master data. The BOM must be reconciled to the MMR and MBR, including alternates and by-products. Routings define operation times, queue and wait steps, line-clearance and cleanup, cleaning validation hold, and test lags. Item masters carry MOQ, EOQ, lead time, lot size, shelf life, storage class, and special compliance flags. Yield and scrap are versioned and conservative. Calendars encode shutdowns and FAT/SAT windows. Incoming quality steps and component release durations are explicit, not “part of lead time.” And every packaging component—artwork, inserts, UDI/GTIN, GS1 codes—exists as a separately planned item because labels derail schedules as often as APIs. If any of this sits in a spreadsheet, you are planning on hope rather than facts.

4) Demand Management—Forecasts, MPS, and Reality

MRP takes its marching orders from the Master Production Schedule (MPS), which arbitrates between forecast and actual orders at the planning fence. In regulated environments, the MPS must also observe validation capacity limits and campaign rules (e.g., allergen or potency segregation). Forecasts should be traceable to commercial assumptions and seasonality; safety stock should be argued with service-level math, not folklore. When complaints or stability studies trigger tighter expiry rules, demand must be re-netted against FEFO, potentially pulling orders forward. When MOC deprecates a component or artwork, demand must shift to the superseding item with Document Control effectivity, and MRP must both drain old stock compliantly and prevent re-buying obsoletes. The goal is not 100% forecast accuracy; the goal is transparent logic so that planners can defend why the system scheduled what it did, and quality can verify that constraints were honored.

5) Netting, Lot Sizing, and Shelf Life

Classic netting subtracts on-hand and scheduled receipts from gross requirements to derive net requirements by date bucket. In GxP, that math must be lot-aware. On-hand that expires before use is not available; on-hand on Hold is not available; on-hand tied to other quality statuses (MRB) is not available. Lot sizing must balance setup economics with expiry risk and storage constraints. Large lot sizes minimize changeovers but inflate scrap at expiry and increase cross-contamination risk unless campaigns are well ruled. For perishable intermediates with hold-time limits, MRP should prefer smaller, more frequent orders even if unit costs rise. The right answer is policy-driven and visible: if the business chooses larger lots, it should also accept higher expiry write-offs measured explicitly in the MRP KPI set.

6) Capacity, Constraints, and Finite Reality

Pure MRP is materials-finite and capacity-infinite. Plants are not. The planning stack must check feasibility against finite capacity and shared resource constraints—operators, suites, utilities, QC slots—and then either adjust dates or raise exceptions. QC laboratories are chronic bottlenecks; if LIMS indicates a 5-day queue for HPLC, the MRP promise date that assumed 2 days is fiction. Packaging lines constrained by label verification and camera checks dictate a different cadence than blending. And utilities (clean steam, WFI) and cleaning validation cooldowns create line balancing challenges that must be encoded as calendar capacity or minimum gaps in routing. If your “feasible plan” depends on ignoring these, expect deviations and emergency MRB decisions later.

7) Inventory, Warehouse Integration, and Picking Logic

MRP that ignores the warehouse is a liar. The plan must respect Bin / Location Management, quarantine areas, cold-chain lanes, and lot/serial attributes. Suggested orders should create reservations that WMS converts into Directed Picking lists using FEFO and suitability rules (potent vs general, allergen segregation). Barcode validation at pick and dispense protects the plan from “right date, wrong lot” mistakes. Cycle counting should target items with high MRP sensitivity—APIs, printed components, and long-lead materials—so that availability is real, not theoretical. Finally, the MRP engine must subscribe to EPCIS events or equivalent so that moves, holds, and releases re-net immediately rather than next week when the spreadsheet jockey appears.

8) Quality, Testing, and Release Coupling

Every MRP suggestion is conditional on quality outcomes. Raw materials are not “available” until component release. Intermediates that require IPC testing or stability holds cannot be scheduled back-to-back as if data were instant. When LIMS results turn marginal or fail, MRP must see the MRB hold and replan downstream orders. For packaging, labeling control and artwork effectivity dates are as real a constraint as API lead times; if artwork is not approved, the kit is not plannable. Release sequencing matters too: if finished goods release depends on long QC tests, MRP must ensure earlier campaign completion so orders ship on time without paying air freight for lateness created on the planner’s desk.

9) Change Control and Obsolescence

Management of Change (MOC) is the fulcrum that keeps MRP honest. When specifications, suppliers, or routings change, the system must know the effectivity date and lot impact. MRP should stop buying the obsolete component, plan consumption of remaining compliant stock, and schedule cut-over batches with the superseding revision. If MRP cannot “see” MOC, planners will work around the system with emails and private trackers, creating the classic audit finding: why did the plant buy a deprecated excipient or print the wrong insert? The antidote is simple: require that MRP runs only against approved master data versions with effective dates, and block purchase order release when effectivity conflicts exist. Tie this to Document Control so artwork, IFU, and eIFU are synchronized with component changes.

10) Common Failure Modes & How to Avoid Them

• Planning on theoretical lead time. Supplier says 4 weeks; real world is 8 with QA. Fix: model quarantine and test lags explicitly; measure actuals and feed back.
• Ignoring expiry in netting. Lots “available” on paper, unusable in reality. Fix: FEFO-aware availability; show “use-by in routing” dates.
• Unversioned BOMs vs MMR/MBR. Explosion mismatches the shop instruction. Fix: one source of truth under Document Control with effectivity.
• Spreadsheet overrides. Private allocations beat WMS. Fix: reservations and Directed Picking driven from the system; audit trail on changes.
• Lab bottlenecks invisible to plan. LIMS waits not modeled. Fix: integrate test capacity and turnaround assumptions; trigger exception when queues exceed plan.
• Artwork and labeling afterthought. Everything ready except the insert. Fix: plan printed components as critical items with their own lead times and approvals.
• MRP suggests against Held/MRB lots. Status not checked. Fix: net availability by status; enforce Hold at planning and execution.
• Capacity denial. Infinite schedule on finite lines. Fix: finite checks; if infeasible, move dates or escalate for capacity action.
• Supplier variability unmodeled. All vendors treated equal. Fix: vendor calendars, reliability factors, and safety lead time per item/supplier.
• No closed-loop learning. Plan errors repeat. Fix: measure plan vs actual for lead time, yield, and expiry scrap; feed back monthly.

11) Metrics That Prove Control

Core KPIs include plan adherence (OTD vs MPS), material availability at start (no-short starts), expiry scrap (value and % by family), quarantine dwell vs target, MRB incidence due to planning, label/insert availability at packaging start, forecast accuracy at the planning fence, supplier reliability (on-time and quality acceptance), and inventory turns by risk class. For governance, track override rate (manual changes to MRP suggestions), effectivity violation attempts, and blocked issues at pick caused by FEFO or status. If these are not measured and reviewed in APR/PQR, you are not running a controlled planning process—you are improvising.

12) How This Fits with V5

V5 by SG Systems Global makes MRP executable by binding planning to inventory truth, quality status, and execution gates. V5 ingests approved BOM, MMR/MBR versions from Document Control, explodes demand with shelf-life awareness, and publishes dated orders to V5 MES and WMS. Reservations convert directly into Directed Picking with Dynamic Lot Allocation and FEFO. Status is enforced end-to-end: lots on Hold or in MRB are invisible to netting and cannot be picked; LIMS results flow into availability; label verification prevents packaging starts without current artwork. MOC effectivity dates are respected automatically: purchase releases and production orders block if they would violate the approved cut-over. All planning actions are captured with audit trails and e-signatures aligned to Part 11/Annex 11. The result: a plan you can execute without spreadsheets, deviations, or excuses.

13) FAQ

Q1. Should MRP be FEFO-aware or is that a WMS concern?
FEFO belongs in both. MRP must recognize expiry windows during the routing; WMS enforces the actual lot pull. If FEFO is only a warehouse rule, planning will overpromise with lots that cannot survive the process timeline.

Q2. Can we treat QC time as part of supplier lead time?
You can, but you shouldn’t. QC is variable and driven by lab capacity and result quality. Model it explicitly, integrate with LIMS, and measure plan vs actual so buffers are evidence-based.

Q3. How does MRP interact with MRB?
When material enters MRB, its availability goes to zero. MRP should immediately re-net demand and either advance alternate supplies, propose reschedule, or flag an exception. When MRB resolves, effectivity and regrade rules determine whether stock returns to the plan.

Q4. We have a great ERP; why pull MRP into MES/WMS?
Keep MRP where it fits—but integrate tightly. MES/WMS own the truth about status, locations, and actual starts. If ERP MRP cannot “see” those truths in near-real time, it will continue to suggest fictional orders and force planners back to spreadsheets.

Q5. What proves MRP data integrity to inspectors?
Reconstructability. You must show the demand source, effective BOM/MMR/MBR, lead time/yield assumptions, shelf-life checks, quality status at suggestion time, and approvals for overrides—all under audit trail and Part 11 controls when electronic.

Related Reading
• Planning & Execution: MES | Finite Capacity Scheduling | Job Queue & Dispatch | Kanban
• Inventory & Traceability: Bin / Location Management | FEFO | Lot Traceability | EPCIS
• Quality & Release: LIMS | MRB | Component Release | Finished Goods Release | Label Verification / UDI
• Governance & Integrity: MOC | Document Control | Audit Trail (GxP) | Annex 11 | 21 CFR Part 11