Deep Learning Supplier Selection and Order Allocation Under Uncertainty
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional deterministic and stochastic programming-based methods for supplier selection and order quantity allocation fail to account for uncertainties in supply chain operations, leading to suboptimal performance and computational inefficiencies, hindering proactive decision-making.
Innovation Solution
An end-to-end deep learning approach that leverages a trained neural network to generate optimal supplier portfolios and order quantities directly from input features, enabling rapid adaptation to changing business environments and facilitating sensitivity and what-if analyses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If stochastic programming-based methods are used to incorporate uncertainties into supplier selection and order quantity allocation, then the reliability of decision-making is improved, but the computational complexity and time consumption increase exponentially
Solution Approach 1:
The patent replaces the traditional stochastic programming mathematical optimization system with a deep learning neural network system. The neural network learns the mapping from problem parameters to optimal decisions through training, substituting the computationally intensive stochastic programming solver with a trained model that provides rapid predictions, thereby reducing computational complexity while maintaining decision reliability
Solution Approach 2:
The patent performs preliminary training of the neural network model using historical data and stochastic programming solutions to pre-learn the optimal decision patterns. This preliminary action allows the model to make rapid predictions without executing complex stochastic programming calculations in real-time, resolving the contradiction between reliability and computational complexity
2Manufacturing precision
If stochastic programming-based methods are used to generate optimal supplier portfolios, then the quality of supplier selection is improved, but the response time to parameter changes increases
Solution Approach 1:
The patent substitutes the time-consuming stochastic programming optimization process with a pre-trained neural network that can predict optimal supplier portfolios in milliseconds. The neural network maintains high selection quality by learning from comprehensive training data while providing rapid responses to parameter changes
Solution Approach 2:
The patent creates a dynamic decision-making system where the neural network can quickly adapt to changing parameters by making real-time predictions. Unlike static stochastic programming solutions that require re-optimization, the neural network dynamically responds to parameter changes through its learned patterns, reducing response time while maintaining selection quality
3Adaptability or versatility
If deep learning-based methods are used for supplier selection and order allocation, then the adaptability to changing environments is improved, but the model training complexity increases
Solution Approach 1:
The patent creates a universal neural network model that can handle multiple supply chain scenarios and parameter variations through a single trained model. The model learns general patterns from diverse training data, enabling it to adapt to changing environments without requiring scenario-specific models, thereby improving adaptability while managing training complexity through unified model architecture
Data Source
AI summary
Systems and methods described herein can involve, for an input of supplier features associated with one or more suppliers, supply chain network features and predicted demand features, processing the input through a trained deep learning model configured to intake the input and output primary supplier from the one or more suppliers, backup supplier from the one or more suppliers, order quantity for the primary supplier, and reservation capacity from the backup supplier; and executing a contract with the primary supplier and the backup supplier based on the order quantity and reservation capacity.