ML Demand Forecasting with Disruption Vectors

Solution Overview

Problem

Conventional demand sensing models in supply chains are unable to accurately capture the impact of disruptions such as economic downturns and pandemics due to their reliance on historical data and lack of consideration for external variables, leading to misguided Supply Chain Management (SCM) decisions, resulting in high inventory costs, stockouts, and poor pricing strategies.

Innovation Solution

A method and system that utilizes a trained Machine Learning (ML) model to predict demand by feeding input vectors including intensity and duration vectors of potential disruption events, along with extrinsic data parameters, to provide accurate future demand forecasts.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional demand sensing models rely on historical demand time-series data, then the models can be implemented with existing data infrastructure, but the models are unable to capture impact of disruption events

Engineering Contradiction:

Improvedemand forecasting accuracyVSAvoidability to capture disruption impact

Core Design Contradiction:

ReliabilityVSAdaptability or versatility

Solution Approach 1:

The model segments demand prediction into two distinct components: (1) baseline demand from historical time-series data using ARIMA, and (2) disruption impact from external variables using regression analysis. This segmentation allows each component to be optimized independently while combining to provide comprehensive demand forecasting that captures both normal patterns and disruption effects.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The invention merges two different modeling approaches: ARIMA for capturing historical demand patterns and regression analysis for incorporating external disruption variables. By combining these models, the system achieves both reliability from historical data and adaptability to disruption events, resolving the contradiction between these two requirements.

Inventive Principle:

Principle #5Merging (Combining)

2Device complexity

If conventional demand sensing models do not consider external variables, then the models are simpler to implement, but the models are unable to capture impact of disruption events

Engineering Contradiction:

Improvemodel complexityVSAvoiddemand forecasting accuracy

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The model segments demand prediction into two distinct components: (1) baseline demand from historical time-series data using ARIMA, and (2) disruption impact from external variables using regression analysis. This segmentation allows each component to be optimized independently while combining to provide comprehensive demand forecasting that captures both normal patterns and disruption effects.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The invention merges two different modeling approaches: ARIMA for capturing historical demand patterns and regression analysis for incorporating external disruption variables. By combining these models, the system achieves both reliability from historical data and adaptability to disruption events, resolving the contradiction between these two requirements.

Inventive Principle:

Principle #5Merging (Combining)

3Ease of manufacture

If conventional demand sensing models rely on historical data only, then the models require minimal data collection infrastructure, but the models provide misguided SCM guidance during disruptions

Engineering Contradiction:

Improveease of model implementationVSAvoideffectiveness of SCM guidance

Core Design Contradiction:

Ease of manufactureVSEase of operation

Solution Approach 1:

The model segments demand prediction into two distinct components: (1) baseline demand from historical time-series data using ARIMA, and (2) disruption impact from external variables using regression analysis. This segmentation allows each component to be optimized independently while combining to provide comprehensive demand forecasting that captures both normal patterns and disruption effects.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The invention merges two different modeling approaches: ARIMA for capturing historical demand patterns and regression analysis for incorporating external disruption variables. By combining these models, the system achieves both reliability from historical data and adaptability to disruption events, resolving the contradiction between these two requirements.

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS20230342796A1Method and system for predicting demand for supply chain

Publication Date: 2023.10.26 WIPRO LTD

AI summary

A method and a system for predicting demand for a supply chain is disclosed. The method includes feeding input vectors to a trained Machine Learning (ML) model, for a future time-period. The input vectors include an intensity vector corresponding to an intensity of a possible disruption-event at each point of time within the future time-period and a duration vector corresponding to the duration of the possible disruption-event, and one or more extrinsic data vectors. The method further includes obtaining a demand for a target product in the future time-period from the trained ML model based on the input vectors.