Supply Chain Action Binning for Decision Speed
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Supply chain networks face inefficiencies due to the large action space when dealing with numerous products, leading to disruptions and severe real-world consequences.
Innovation Solution
A system that generates binned actions for each product in a supply chain network, compressing the action space by representing closely related actions with a single binned action, and using multi-task reinforcement learning to determine these actions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional decision-making methods are used for each product individually, then decision accuracy can be maintained, but computational cost and time increase significantly due to the large action space
Solution Approach 1:
The patent combines multiple similar actions into a single binned action. Instead of treating each ordering decision (order 10 units, order 11 units, etc.) as separate actions, the system bins these into consolidated action categories, reducing the overall action space while maintaining decision quality
Solution Approach 2:
The patent segments the large action space into smaller, manageable binned action categories. By dividing the continuous action space into discrete bins, the system makes the decision problem more tractable while preserving the essential decision-making capabilities
2Measurement precision
If detailed actions are considered for each product, then decision precision is maintained, but processing time increases due to the large number of products
Solution Approach 1:
The patent merges multiple similar actions into binned actions, reducing the number of decisions that need to be evaluated. This consolidation maintains the precision of action selection by preserving the essential decision categories while eliminating redundant evaluations
3Reliability
If the system processes decisions for all products comprehensively, then decision quality is maintained, but computational resources are excessively consumed
Solution Approach 1:
The patent combines multiple similar actions into binned actions, reducing the computational burden while maintaining decision reliability. This approach allows the system to process supply chain decisions for all products without exhaustively evaluating every possible action, thus conserving computational energy
Data Source
AI summary
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for generating actions for a supply chain network. One of the methods includes receiving a request to generate an action in a supply chain network for a particular product based on current state information; providing a request to an action model to generate a respective probability distribution for one or more actions for one or more products; receiving, from the action model, the respective probability distributions for the one or more products; determining, for each product, a binned action from the respective probability distribution; providing a request to a sequence model to generate a respective correction for the one or more binned actions; and receiving, from the sequence model, the respective correction for the respective binned action.