State-Based Decision Diagram Optimization Solvers for Dynamic Routing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing decision science-based optimization software tools face challenges in accommodating dynamic routing, scheduling, and assignment problems due to static decision models that cannot be easily updated to reflect fluid conditions, leading to sub-optimal solution sets and deployment issues across different execution environments.
Innovation Solution
The development of state-based decision diagram optimization solvers and simulators that automate and simplify testing and deployment across various software execution environments and hardware platforms, enabling dynamic updates and flexible configuration of optimization models.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If static decision models are used, then deployment simplicity is maintained, but adaptability to dynamic conditions deteriorates
Solution Approach 1:
The patent implements dynamic decision models that can be updated and reconfigured at runtime to adapt to changing conditions. The system allows modification of model parameters, constraints, and objectives without requiring complete model redesign, enabling real-time adaptation to dynamic routing, scheduling, and assignment problems while maintaining manageable complexity through structured model architecture.
Solution Approach 2:
The system enables dynamic adjustment of model parameters such as optimization objectives, constraints, and decision variables based on changing operational conditions. This allows the same model framework to handle various dynamic scenarios by modifying parameters rather than restructuring the entire model, thus improving adaptability without proportionally increasing complexity.
2Reliability
If dynamic optimization models are implemented, then solution quality improves, but deployment complexity increases
Solution Approach 1:
The patent segments the optimization system into modular components including model definition, parameter configuration, solver execution, and result interpretation. This modular architecture allows dynamic models to be built and deployed in manageable sections, reducing overall deployment complexity while maintaining high solution quality through systematic model construction and validation.
Solution Approach 2:
The system performs preliminary model validation, parameter checking, and feasibility analysis before full optimization execution. This preliminary action ensures that dynamic models are properly configured and will produce reliable solutions, reducing deployment failures and simplifying the overall deployment process by catching issues early in the model setup phase.
3Reliability
If comprehensive testing across multiple environments is performed, then deployment reliability improves, but testing time increases
Solution Approach 1:
The patent implements virtual environment copies that replicate production conditions for testing purposes. These virtual copies allow comprehensive testing of dynamic optimization models across multiple scenarios and configurations without requiring physical deployment to each environment, significantly reducing testing time while maintaining deployment reliability through thorough virtual validation.
Data Source
AI summary
Technologies for creating and executing a runnable that includes a decision diagram (DD)-based optimization function are capable of configuring the runnable based on runnable options data passed by the client program to a runner function, creating an executable version of the configured runnable based on context data associated with the client program, using input data read from the data source, creating a decision model and executing the executable version of the configured runnable using the decision model, and writing output produced by the executing of the executable version of the configured runnable to the data source.