Inferensys

Glossary

Net Flow Equation

The DDMRP calculation of on-hand plus on-order inventory minus qualified sales order demand, used to determine current buffer status and replenishment priority.
Operations manager reviewing inventory AI on tablet, stock levels and reorder dashboards visible, warehouse office setup.
DDMRP INVENTORY POSITIONING

What is Net Flow Equation?

The Net Flow Equation is the fundamental DDMRP calculation used to determine the current inventory buffer status and prioritize replenishment orders.

The Net Flow Equation is a Demand Driven Material Requirements Planning (DDMRP) calculation defined as on-hand inventory plus on-order inventory minus qualified sales order demand. This equation determines the current 'flow' position of a buffered item, representing the net quantity of stock available to satisfy future demand after accounting for all incoming supply and committed outgoing orders.

The resulting net flow value is compared against the item's dynamically sized DDMRP buffer zones—green, yellow, and red—to establish the current replenishment priority. A net flow position deep in the red zone signals a critical stockout risk requiring immediate action, while a position in the green zone indicates a healthy inventory posture. This calculation is the core mechanism that decouples supply from demand variability.

DDMRP CORE MECHANIC

Key Characteristics of the Net Flow Equation

The Net Flow Equation is the central calculation in Demand Driven Material Requirements Planning (DDMRP) that determines the current health of an inventory buffer. It provides a single, actionable number used to prioritize replenishment and signal execution urgency.

01

The Core Formula

The equation is defined as On-Hand Inventory + On-Order Inventory - Qualified Sales Order Demand. This calculation represents the net available stock position after accounting for all committed future supply and demand. It is not a forecast; it is a snapshot of reality. On-Hand is physical stock. On-Order is stock that has been released but not yet received. Qualified Demand includes sales orders that are due today, past due, or within a defined demand spike horizon.

02

Buffer Status Determination

The result of the Net Flow Equation directly maps to the DDMRP buffer zones:

  • Green Zone: Net Flow is high. No action required; the system is stable.
  • Yellow Zone: Net Flow has dropped into the rebuild zone. A supply order is typically generated to restore the buffer to green.
  • Red Zone: Net Flow is critically low. The system prioritizes this item for immediate replenishment to prevent a stockout. The red zone is further split into a safety base and a safety zone.
03

Decoupling from Forecast Error

Unlike traditional MRP which relies on forecasted demand, the Net Flow Equation uses only qualified actual demand. This decouples the execution signal from forecast inaccuracy. By only subtracting real, qualified sales orders, the equation prevents the nervousness and bullwhip effect caused by constantly changing forecasts. The buffer absorbs variability, while the net flow calculation provides a stable, trustworthy signal for execution.

04

Priority Calculation

The Net Flow Equation is the input for calculating replenishment priority. The percentage of the buffer consumed is derived by comparing the Net Flow position against the total buffer level. A lower percentage indicates higher urgency. This allows planners to focus on items where Net Flow is deepest in the red zone, providing an objective, mathematically sound method for prioritizing work orders and purchase orders across thousands of SKUs.

05

Qualified Order Spike Handling

A critical component is the Qualified Sales Order Demand term. To prevent a single large order from instantly crashing the Net Flow and triggering a false emergency, orders are qualified against a defined spike threshold. Demand exceeding this threshold is not fully subtracted from the equation immediately. Instead, it is managed separately, ensuring the Net Flow reflects normal operational demand and prevents overreaction to known, planned large events.

06

Real-Time Recalculation

The Net Flow Equation is designed for continuous recalculation. Every inventory transaction, order release, or sales order commitment triggers an update. This provides a real-time, always-accurate view of buffer health. The dynamic nature allows the system to instantly detect emerging problems, such as a supplier delay that prevents on-order stock from arriving, and immediately escalate the priority as the Net Flow position deteriorates.

NET FLOW EQUATION

Frequently Asked Questions

Clarifying the core DDMRP calculation that determines buffer status and drives replenishment priority across the supply chain.

The Net Flow Equation is the foundational calculation in Demand Driven Material Requirements Planning (DDMRP) that determines the current buffer status by computing on-hand inventory plus on-order inventory minus qualified sales order demand. This single equation provides a real-time snapshot of available stock relative to actual demand, enabling planners to prioritize replenishment based on actual buffer penetration rather than arbitrary dates. The formula is expressed as: Net Flow = On-Hand + On-Order - Qualified Sales Order Demand. On-hand represents physically available stock, on-order includes all open purchase and production orders not yet received, and qualified sales order demand includes only actual, due-now demand—not forecasts. This calculation is performed continuously, allowing dynamic buffer adjustments as conditions change.

Prasad Kumkar

About the author

Prasad Kumkar

CEO & MD, Inference Systems

Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.

His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.