Fill Rate Prediction System Using ML for Supply Chain Visibility
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for predicting fill rates in supply chains are costly, require common platforms and active participation from sellers and vendors, and fail to accurately predict future orders, leading to potential shortfalls and increased inventory costs.
Innovation Solution
A fill rate prediction system using machine learning models that analyze order attribute data, vendor rank data, and recency data to determine the probability of in-full fill rates, allowing for proactive remedial actions and reducing the need for extensive technological integration with vendors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If complex information technology systems are implemented for communication between sellers and vendors, then communication efficiency improves, but system cost and complexity increase
Solution Approach 1:
The patent introduces a fill rate prediction system that acts as an intermediary between sellers and vendors. This system uses machine learning models to predict fill rates and automatically communicate predictions to relevant parties, eliminating the need for complex direct communication systems while maintaining operational efficiency. The prediction system mediates information flow between sellers and vendors through automated algorithms rather than complex IT infrastructure.
Solution Approach 2:
The fill rate prediction system operates autonomously by automatically collecting data, running predictions through machine learning models, and generating outputs without requiring active participation from sellers or vendors. The system serves itself by autonomously performing data collection, analysis, and communication functions that would otherwise require complex interconnected IT systems between multiple parties.
2Loss of information
If integrated information technology systems are implemented between sellers and vendors, then supply chain visibility improves, but up-front costs increase
Solution Approach 1:
The patent introduces a fill rate prediction system that acts as an intermediary between sellers and vendors. This system uses machine learning models to predict fill rates and automatically communicate predictions to relevant parties, eliminating the need for complex direct communication systems while maintaining operational efficiency. The prediction system mediates information flow between sellers and vendors through automated algorithms rather than complex IT infrastructure.
Solution Approach 2:
Instead of implementing expensive, permanent integrated IT systems between sellers and vendors, the patent employs a cost-effective machine learning-based prediction system that processes information independently. This approach uses computationally efficient models that can be deployed without substantial infrastructure investment, providing supply chain visibility at minimal cost compared to traditional integrated systems.
3Measurement precision
If active participation by sellers and vendors is required to input and revise order information, then data accuracy improves, but operational burden increases
Solution Approach 1:
The fill rate prediction system operates autonomously by automatically collecting data, running predictions through machine learning models, and generating outputs without requiring active participation from sellers or vendors. The system serves itself by autonomously performing data collection, analysis, and communication functions that would otherwise require complex interconnected IT systems between multiple parties.
Solution Approach 2:
The system incorporates feedback mechanisms where prediction results are communicated back to sellers and vendors, allowing them to verify and correct data if needed. This feedback loop maintains data accuracy by enabling participants to review predictions against their actual data while minimizing the need for active data input, as the system primarily operates on automatically collected information.
4Measurement precision
If machine learning models are used to predict fill rates automatically, then prediction accuracy improves, but computational requirements increase
Solution Approach 1:
The patent applies machine learning models selectively to predict fill rates only for specific orders or vendor-seller combinations where predictions provide the most value. Rather than continuously running complex models for all possible predictions, the system employs partial action by targeting predictions to high-impact scenarios, thereby reducing overall computational resource consumption while maintaining high prediction accuracy where it matters most.
Data Source
AI summary
A computing device is configured to obtain order attribute data characterizing at least one order placed and to obtain rank data characterizing a supply performance versus other supply performances. The computing device can also be configured to obtain recency data characterizing a past supply performance and to determine a probability of an in-full fill rate of the at least one order using a fill rate prediction model. The computing device can also send the probability of the in-full fill rate to a supply partner.