MRP cheat sheet

Hierarchy levels top-down: strategic business plan (annual, units and dollars) → sales and operations planning (monthly, product families) → master production schedule (weekly, end items) → material requirements planning (daily/weekly, components) → production activity control (daily, orders).

MRP answers three questions: what components are needed, how many, and when; it is a time-phased dependent-demand calculation, not a scheduling or execution system.

Independent demand is forecast-driven for end items; dependent demand is calculated from parent item requirements via the BOM, so it does not need separate forecasting.

MRP inputs: master production schedule, bill of materials, inventory records (on-hand, scheduled receipts, lead times, lot-sizing data). MRP outputs: planned orders, released order recommendations, action/exception messages.

Time buckets are normally weekly in MRP but daily in repetitive and JIT environments; bucketless (time-phased) systems place orders on exact due dates rather than bucket boundaries.

Planning horizon must exceed the longest cumulative lead time of any end item, otherwise the plan cannot react to changes before the first order must be released.

MRP II extends MRP with capacity planning, costing, and financial integration; ERP extends MRP II across the enterprise and supply chain.

Closed-loop MRP feeds capacity results and execution feedback back into the plan, so the MPS is revised when capacity is infeasible.

MPS states, per end item and per time bucket, the quantity planned to be produced; it drives MRP and must be stated in actual buildable configurations, not product-family aggregates.

Available-to-promise (ATP) = scheduled MPS receipts minus already-committed customer orders; the first-period ATP subtracts beginning inventory as well.

Projected available balance = prior balance + MPS receipts − max(forecast, actual demand) when forecast and orders are combined by the larger-value rule; otherwise subtract their sum.

MPS should be frozen for the cumulative lead time (firm zone), then slushy (trading zone) where limited changes are allowed, then liquid (free zone) where changes are routine.

Forecast consumption prevents double counting: customer orders consume forecast within a period or backward over a consumption window before the forecast is treated as demand.

Time fences define the boundaries between frozen, slushy, and liquid zones and are the mechanism for controlling nervousness in the MPS.

Planning bill (super bill) or percentage-blend BOMs translate family-level forecasts into option-level MPS quantities for configure-to-order products.

Rough-cut capacity planning should validate the MPS before it is released to MRP; an unvalidated MPS generates infeasible component plans.

BOM is a hierarchical listing of components, quantities per parent, and effectivity; each parent-component pair is one BOM line with its own scrap and effectivity dates.

Low-level coding places each item at its deepest level in the product structure so MRP explodes it only after all parent requirements are known.

Single-level explosion: component requirement = parent planned order release × quantity per × (1 + scrap %).

Multi-level explosion applies the single-level calculation level by level, top to bottom, using low-level codes; cumulative lead time is the sum of lead times along the longest path.

Phantom (blow-through) items are assemblies that are exploded through rather than stocked; they have no inventory and no planned orders, only their children are planned.

Where-used (implosion) reports list every parent that consumes an item; they are essential for engineering change impact analysis and recall/quality containment.

BOM effectivity dates and engineering change control ensure that a component is only planned for orders whose dates fall inside its effective range.

Modular (planning) BOMs and option classes support configure-to-order: the 100% BOM is assembled from selected options at order entry, then exploded by MRP.

Item master fields that drive MRP: item number, description, unit of measure, item type (make/buy/phantom), planner, low-level code, lead time, safety stock, lot-sizing rule, and scrap percentage.

Lead time components: queue, setup, run, move, wait, and inspection; cumulative lead time runs from purchase of the longest-lead component to final assembly.

Planned order release date = planned order due date − lead time; if lead time is longer than the remaining horizon, the order is released immediately and flagged.

Safety stock covers demand and supply variability; safety lead time covers timing variability and shifts the due date rather than the quantity.

Inventory record accuracy is measured by cycle counting; a common tolerance is ±0.2% of inventory value for A items and ±5% for C items, with 95%+ record accuracy as an MRP prerequisite.

Shrinkage, scrap, and yield must be carried in the item master; planning scrap = order quantity × scrap % and is added to the component requirement, not subtracted from supply.

Item status codes (active, inactive, obsolete, pending) prevent MRP from planning superseded parts while preserving history for where-used analysis.

Fixed, variable, and lot-for-lot lead times: fixed lead time applies per order, variable lead time scales with quantity, and lot-for-lot ignores quantity effects.

Net requirement = gross requirement + safety stock − projected available balance − scheduled receipts; if negative, it is zero.

Lot-for-lot (L4L) orders exactly the net requirement each period; it minimizes inventory but maximizes order count and setup cost.

Fixed order quantity orders a constant lot size and may generate planned orders larger than the net requirement, creating residual inventory.

Economic order quantity EOQ = √(2DS/H), where D = annual demand, S = ordering/setup cost per order, H = annual holding cost per unit.

Period order quantity (POQ) sets the order cycle to EOQ / average periodic demand, ordering to cover an integer number of periods.

Least total cost compares setup cost against holding cost across candidate coverage periods and selects the coverage where the two are closest to equal.

Least unit cost divides total setup plus holding cost by the units ordered and selects the coverage period with the lowest cost per unit.

Net change MRP recalculates only items affected by a change; regenerative MRP recalculates the entire plan, typically on a fixed cycle.

Pegging traces a component requirement back to the end item and customer order that generated it, which is essential for order promising and change impact analysis.

RCCP validates the MPS using capacity bills, resource profiles, or rough-cut capacity planning by work center; it uses product-family load factors rather than detailed routings.

Capacity requirements planning (CRP) converts planned and released orders into work-center load using routings, standard times, and order dates; it is the capacity check for MRP.

Load profile per work center per period = Σ (order quantity × run time per unit + setup time) for all orders assigned to that period.

Available capacity = work center hours per shift × shifts × days in period × utilization × efficiency, less planned downtime and maintenance.

Load percentage = required hours / available hours; above 100% indicates a capacity constraint, below a threshold (commonly 70–80%) indicates underutilization.

Infinite loading assumes unlimited capacity and shows pure demand; finite loading schedules within capacity and produces a feasible sequence.

Capacity is adjusted by overtime, extra shifts, subcontracting, alternate routings, or moving load to another period; load is adjusted by lot splitting, order release timing, or rescheduling.

Input/output control monitors actual work-center throughput against planned input and output; a growing queue indicates capacity shortfall or upstream release problems.

MRP outputs: planned orders (release, reschedule in, reschedule out, cancel), scheduled receipts, action messages, exception reports, and pegging reports.

Release message means the planned order's start date has arrived and the order should be converted to a production or purchase order.

Reschedule-in moves an existing scheduled receipt earlier; reschedule-out moves it later; cancel recommends deleting a receipt no longer needed.

Exception messages are prioritized by planner action date and severity; a common rule is to work the highest-severity messages first and never let the exception queue age past a defined threshold.

Nervousness is excessive plan change caused by small demand changes; it is dampened by time fences, firm planned orders, and lot-sizing stability.

Firm planned orders are not changed by MRP regeneration; they are used to protect committed supply and to stabilize the near-term plan.

Exception categories typically include late orders, past-due receipts, negative inventory, orders beyond the planning horizon, and orders with no routing or BOM.

Planner should review action messages by exception, not by scanning every order; the message is a recommendation and requires planner judgment before execution.

ERP integrates MRP with finance, purchasing, sales, and shop-floor execution; a single item master and BOM is the shared source of truth across modules.

Implementation phases: readiness assessment, data cleansing and conversion, conference room pilot, cutover, go-live, and post-go-live stabilization.

Critical success factors: executive sponsorship, accurate item/BOM/inventory data, user training, process ownership, and a defined cutover and rollback plan.

Common failure causes: poor data accuracy, scope creep, unrealistic timelines, insufficient testing, and treating the project as IT-only rather than a business change.

Delivery performance is measured by on-time delivery, order fill rate, and perfect order rate; plan stability by the percentage of orders rescheduled per period.

Inventory metrics: inventory turns = cost of goods sold / average inventory value; days of supply = average inventory / average daily usage.

MRP effectiveness metrics include forecast accuracy (MAPE), schedule adherence, capacity utilization, exception message volume, and record accuracy.

Integration testing must verify that transactions flow correctly across modules: a purchase order receipt must update inventory, MRP, and the general ledger consistently.

The planning hierarchy flows from the strategic business plan down to production activity control, with each level covering a shorter horizon and finer granularity than the one above it.

MRP is a time-phased dependent-demand system: it calculates component requirements from the MPS and BOM rather than forecasting them independently.

The four core MRP inputs are the master production schedule, bill of materials, inventory records, and item master data such as lead times and lot-sizing rules.

MRP outputs include planned orders, action messages, exception reports, and pegging information used to trace requirements back to their source.

The planning horizon must cover the longest cumulative lead time so the system can release orders early enough to meet the end-item schedule.

MRP II adds capacity planning, costing, and financial integration to basic MRP; ERP extends the same logic across the whole enterprise.

Closed-loop MRP feeds capacity and execution results back into the plan, so the MPS is revised when the plan proves infeasible.

Time buckets are typically weekly in MRP and daily in repetitive or JIT environments; bucketless systems schedule orders on exact dates.

The MPS states, per end item and per time bucket, the quantity planned to be produced; it drives MRP and should reflect buildable configurations.

Available-to-promise (ATP) = scheduled MPS receipts minus committed customer orders; the first period also subtracts beginning inventory.

Projected available balance = prior balance + MPS receipts − demand, where demand is the forecast or actual orders per the combining rule in use.

Time fences divide the horizon into frozen, slushy, and liquid zones, controlling how much the MPS may change in each period.

Forecast consumption prevents double counting by netting customer orders against forecast before the forecast is treated as demand.

Planning bills translate product-family forecasts into option-level MPS quantities for configure-to-order product lines.

Rough-cut capacity planning should validate the MPS before it is released to MRP, so component plans are not built on infeasible schedules.

A frozen zone typically extends for the cumulative lead time, protecting committed supply from plan changes.

A BOM lists each parent-component pair with quantity per, scrap percentage, and effectivity dates; each pair is one BOM line.

Low-level coding places each item at its deepest level in the product structure so MRP explodes it only after all parent requirements are known.

Single-level explosion: component requirement = parent planned order release × quantity per × (1 + scrap %).

Multi-level explosion applies the single-level calculation level by level, top to bottom, using low-level codes.

Phantom items are exploded through rather than stocked; they have no inventory and no planned orders, only their children are planned.

Where-used (implosion) reports list every parent consuming an item and are essential for engineering change impact analysis.

Effectivity dates ensure a component is only planned for orders whose dates fall inside its effective range.

Modular and planning BOMs support configure-to-order by assembling a 100% BOM from selected options at order entry.

Item master fields that drive MRP include item number, unit of measure, item type, planner, low-level code, lead time, safety stock, lot-sizing rule, and scrap percentage.

Lead time components are queue, setup, run, move, wait, and inspection; cumulative lead time runs along the longest path to final assembly.

Planned order release date = planned order due date − lead time; orders whose release date has passed are flagged for immediate action.

Safety stock covers quantity variability; safety lead time covers timing variability by shifting due dates rather than quantities.

Cycle counting measures inventory record accuracy, with tighter tolerances for A items than for C items; high accuracy is an MRP prerequisite.

Planning scrap adds to component requirements as order quantity × scrap %, increasing the quantity planned rather than reducing supply.

Item status codes (active, inactive, obsolete, pending) prevent MRP from planning superseded parts while preserving history.

Lead times may be fixed per order, variable with quantity, or lot-for-lot, depending on the item's setup and run characteristics.

Net requirement = gross requirement + safety stock − projected available balance − scheduled receipts; a negative result is set to zero.

Lot-for-lot orders exactly the net requirement each period, minimizing inventory but maximizing order count and setup cost.

Fixed order quantity orders a constant lot size and may create residual inventory when the lot exceeds the net requirement.

Economic order quantity EOQ = √(2DS/H), where D is annual demand, S is ordering or setup cost per order, and H is annual holding cost per unit.

Period order quantity sets the order cycle to EOQ divided by average periodic demand, ordering to cover an integer number of periods.

Least total cost compares setup cost against holding cost across candidate coverage periods and selects where the two are closest to equal.

Least unit cost divides total setup plus holding cost by the units ordered and selects the coverage with the lowest cost per unit.

Pegging traces a component requirement back to the end item and customer order that generated it, supporting order promising and change impact analysis.

RCCP validates the MPS using capacity bills, resource profiles, or work-center load factors from product families rather than detailed routings.

Capacity requirements planning converts planned and released orders into work-center load using routings, standard times, and order dates.

Load per work center per period = Σ (order quantity × run time per unit + setup time) for all orders assigned to that period.

Available capacity = hours per shift × shifts × working days × utilization × efficiency, less planned downtime and maintenance.

Load percentage = required hours / available hours; above 100% signals a capacity constraint, well below 100% signals underutilization.

Infinite loading assumes unlimited capacity and shows pure demand; finite loading schedules within capacity and produces a feasible sequence.

Capacity is adjusted by overtime, extra shifts, subcontracting, alternate routings, or moving load to another period.

Input/output control compares actual work-center throughput against planned input and output; a growing queue signals capacity shortfall.

MRP outputs include planned orders, scheduled receipts, action messages, exception reports, and pegging reports.

Release message means the planned order's start date has arrived and the order should be converted to a production or purchase order.

Reschedule-in moves an existing scheduled receipt earlier; reschedule-out moves it later; cancel recommends deleting a receipt no longer needed.

Exception messages are prioritized by planner action date and severity, and the highest-severity messages are worked first.

Nervousness is excessive plan change caused by small demand changes; time fences, firm planned orders, and lot-sizing stability dampen it.

Firm planned orders are not changed by MRP regeneration and are used to protect committed supply and stabilize the near-term plan.

Exception categories include late orders, past-due receipts, negative inventory, orders beyond the horizon, and orders lacking a routing or BOM.

Action messages are recommendations requiring planner judgment; they are reviewed by exception rather than by scanning every order.

ERP integrates MRP with finance, purchasing, sales, and shop-floor execution; a single item master and BOM is the shared source of truth.

Implementation phases include readiness assessment, data cleansing and conversion, conference room pilot, cutover, go-live, and post-go-live stabilization.

Critical success factors include executive sponsorship, accurate item/BOM/inventory data, user training, process ownership, and a cutover and rollback plan.

Common failure causes include poor data accuracy, scope creep, unrealistic timelines, insufficient testing, and treating the project as IT-only.

Delivery performance is measured by on-time delivery, order fill rate, and perfect order rate; plan stability by the percentage of orders rescheduled per period.

Inventory turns = cost of goods sold / average inventory value; days of supply = average inventory / average daily usage.

MRP effectiveness metrics include forecast accuracy (MAPE), schedule adherence, capacity utilization, exception message volume, and record accuracy.

Integration testing must verify transactions flow consistently across modules, such as a purchase order receipt updating inventory, MRP, and the general ledger.