Layered Scheduling with Priority Partitioning for Faster Solves
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Complex scheduling problems in supply chain management often result in increased complexity and longer solve times, leading to suboptimal schedule quality due to the inability to effectively prioritize and sequence resource demands.
Innovation Solution
A layered scheduling approach is employed, where scheduling problems are partitioned into smaller, prioritized subsets, solved stepwise, and refined through a final solve pass to improve overall schedule metrics, utilizing a server and database system that includes a partition module, schedule optimizer, and user interface to manage and optimize scheduling across supply chain entities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional scheduling methods are used to solve complex supply chain scheduling problems, then comprehensive resource allocation is achieved, but solve times increase and schedule quality deteriorates
Solution Approach 1:
The patent partitions the complex scheduling problem into multiple priority-based subsets, where high-priority demands are separated from low-priority demands. This segmentation allows the system to solve critical scheduling decisions first without being overwhelmed by the entire problem space, thereby reducing solve time while maintaining schedule quality for urgent requirements.
Solution Approach 2:
The system performs preliminary scheduling actions by establishing high-priority schedules first before addressing lower-priority demands. This preliminary action ensures that critical resource allocation decisions are made early, reducing overall solve time while guaranteeing that time-sensitive scheduling requirements are met with high quality.
2Adaptability or versatility
If comprehensive scheduling of all demands is attempted, then complete resource allocation is achieved, but complexity increases and manageability decreases
Solution Approach 1:
The scheduling problem is divided into priority-based subsets that can be managed independently. High-priority demands are scheduled separately from low-priority demands, reducing the complexity of each individual scheduling task while maintaining comprehensive coverage of all demands through iterative processing of multiple subsets.
Solution Approach 2:
The patent introduces a priority dimension to organize and manage scheduling demands. By adding this hierarchical dimension, the system transforms a flat, complex scheduling problem into a structured multi-level problem that is more manageable and adaptable, allowing comprehensive scheduling coverage without overwhelming complexity.
3Reliability
If high-priority demands are scheduled first with iterative locking, then schedule quality for critical demands improves, but overall scheduling flexibility may be reduced
Solution Approach 1:
The system performs preliminary scheduling for high-priority demands and locks these schedules early to ensure quality critical path decisions. This preliminary action guarantees that time-sensitive and high-value scheduling requirements are optimized first, while subsequent lower-priority scheduling can still adapt to available resources without compromising overall flexibility.
Solution Approach 2:
The scheduling system dynamically adjusts between locked high-priority schedules and flexible low-priority scheduling. The iterative locking mechanism allows the system to maintain rigid constraints where quality is critical while preserving flexibility in less critical areas, adapting the level of constraint based on demand priority and resource availability.
Data Source
AI summary
A system and method are disclosed for layered scheduling. The method includes partitioning a scheduling problem into ordered subsets based on a prioritization scheme, applying a scheduling algorithm to optimize a first subset of the ordered subsets and freeze a corresponding schedule, determining whether there are any remaining subsets that have not been optimized, in response to determining that there are remaining subsets that have not been optimized, loading a next subset ordered according to the prioritization scheme, optimizing the loaded subset without disturbing the frozen schedule, and in response to determining that there are no remaining subsets to optimize, running a final pass of the scheduling algorithm to improve the global schedule metrics. The method further includes where the prioritization scheme is based on a relative priority of tasks to be performed, a value of finished goods that are to be produced or requirements regarding a use of resources.