Event-Based Supply Chain Replenishment for Granular Forecast Simulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Large-scale supply chain systems face challenges in accurately modeling and simulating inventory flow due to non-deterministic factors like demand and transit times, leading to uncertainty in outcomes and significant computational resource requirements, especially when dealing with complex networks involving hundreds or thousands of locations.
Innovation Solution
An event-based simulation system that receives forecasted demand distribution, domain logic, and policies to generate action events, which are applied to the current run-state of the supply chain to yield observation events, transforming these into predicted metrics for inventory management, allowing for 'what-if' scenarios and reducing uncertainty through repetitive simulations with different discrete values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If simulation is performed with high level of granularity to obtain insights regarding potential supply chain interruptions, then measurement precision is improved, but use of energy by stationary object worsens
Solution Approach 1:
The patent segments the supply chain simulation into multiple independent time epochs, where each epoch represents a discrete time period. By dividing the overall simulation into smaller epoch-based units, the system can process simulations in manageable increments rather than attempting to simulate the entire supply chain network continuously, thereby reducing computational resource requirements while maintaining detailed granularity at each epoch level.
Solution Approach 2:
The patent applies preliminary action by pre-calculating and storing deterministic supply chain parameters (such as inventory levels, order quantities, and transit times) before executing the simulation. This allows the simulation to reuse these pre-computed values across multiple epochs and scenarios, avoiding redundant calculations and significantly reducing the computational energy required for high-granularity simulations.
2Reliability
If repeated execution of simulations is performed to provide insight into possible outcomes, then reliability is improved, but use of energy by stationary object worsens
Solution Approach 1:
The patent creates multiple simulation traces by copying the same epoch structure and applying different discrete values to non-deterministic inputs across each trace. Instead of fully re-executing complex simulations multiple times, the system copies the epoch framework and varies only the necessary input parameters, thereby achieving reliable outcome analysis through repeated execution while minimizing redundant computational energy consumption.
3Adaptability or versatility
If simulation models include non-deterministic inputs such as demand and availability times, then adaptability is improved, but measurement precision worsens
Solution Approach 1:
The patent applies dynamics by allowing the simulation model to adaptively select discrete values from non-deterministic input distributions during epoch execution. The system dynamically adjusts simulation parameters based on the specific epoch context and observed supply chain states, enabling the model to capture real-world variability while maintaining measurement precision through structured probabilistic sampling rather than fixed deterministic values.
Solution Approach 2:
The patent changes parameters by transforming non-deterministic continuous inputs (such as demand rates and availability times) into discrete parameter values suitable for epoch-based simulation. This parameter transformation allows the system to maintain adaptability to supply chain variability while achieving measurement precision by working with discrete, computationally manageable parameter representations rather than continuous non-deterministic variables.
Data Source
AI summary
Event-based replenishment simulation for an enterprise supply chain is described. On a per item, per node, per epoch basis, the simulation may generate a stream of action events based on forecasted demand, supply chain logic, and policy inputs that are applied to a current-run state of the supply chain in order to yield a stream of observation events. Requested metrics may be received, and the observation events may then be transformed to predict values for the metrics as output of the simulation. The simulation may be repeated for a given epoch using discrete demand values from a demand distribution, for a plurality of epochs, and/or across a plurality of items at a plurality of nodes. Resultantly, the simulation output can be used for predicting a future run-state of the supply chain across items and nodes.