Factory Task Ranking for Event-Responsive Manufacturing Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for operating a factory lack robustness and adaptability to unexpected events such as machine failures, worker absences, and security emergencies, as they rely on theoretical schedules that fail to account for real-time perturbations.
Innovation Solution
A computer-implemented method that determines an operating mode for a factory by representing manufacturing tasks with evolution laws, calculating propensities based on constraints and events, and dynamically assigning machines and resources to prioritize tasks, allowing for real-time adaptation and robust operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If theoretical scheduling algorithms are used to optimize machine and worker schedules, then production efficiency is improved, but robustness to unexpected events deteriorates
Solution Approach 1:
The patent implements a dynamic scheduling system that continuously monitors the factory state and recalculates task priorities in real-time. Instead of using static theoretical schedules, the system adapts its scheduling decisions based on current events such as machine failures, worker absences, or new urgent orders. This dynamic approach resolves the contradiction by maintaining high productivity through continuous optimization while simultaneously improving robustness by responding adaptively to unexpected disruptions.
Solution Approach 2:
The system incorporates real-time feedback mechanisms that monitor the actual state of the factory and compare it with the planned schedule. When deviations occur due to unexpected events, the feedback loop triggers recalculation of task propensities and re-ranking of manufacturing tasks. This feedback-driven approach allows the system to maintain productivity by quickly adjusting to new conditions while inherently providing robustness through continuous adaptation to actual factory conditions.
2Adaptability or versatility
If complex optimization algorithms are implemented to handle multiple constraints and events, then adaptability to manufacturing events is improved, but system complexity increases
Solution Approach 1:
The patent implements a self-service scheduling system where the factory management system automatically monitors events, recalculates task propensities, and re-ranks manufacturing tasks without human intervention. The system serves itself by having built-in capabilities to detect changes in factory state, compute new schedules based on updated constraints, and implement the revised scheduling decisions autonomously. This self-service approach improves adaptability to manufacturing events while controlling system complexity by automating the entire response chain rather than requiring complex manual coordination.
Solution Approach 2:
The system introduces an intermediary scheduling layer that sits between the raw factory events and the execution of manufacturing tasks. This intermediary component (the propensity calculation and task ranking mechanism) translates complex multi-constraint optimization problems into simplified priority rankings that can be easily implemented. By using this intermediary approach, the system achieves high adaptability to various manufacturing events while keeping the overall system complexity manageable through the use of a standardized mediation layer.
3Speed
If real-time monitoring and dynamic re-scheduling are implemented, then responsiveness to events is improved, but computational load increases
Solution Approach 1:
The patent implements periodic monitoring and re-scheduling cycles rather than continuous real-time processing. The system monitors factory events and triggers re-scheduling calculations at defined intervals or upon occurrence of specific triggering events. This periodic approach improves responsiveness by ensuring regular updates while reducing computational load by avoiding constant recalculation. The system balances the need for timely response with energy-efficient computation by using event-driven or time-based periodic triggers instead of continuous processing.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
The invention notably relates to a computer-implemented method for operating a factory. The method comprises providing one or more manufacturing tasks. Each manufacturing task is represented by an evolution law. The evolution law describes a manufacturing step of a product by one or more manageable machines using resources. The method further comprises providing one or more manufacturing constraints. The method further comprises providing one or more manufacturing events. The method further comprises determining an operating mode of the factory based on the one or more manufacturing constraints and on one or more constraints on a product to manufacture. The determining includes one or more iterations. Each iteration comprises computing, for each manufacturing task, a respective propensity of the manufacturing task. The propensity represents a frequency with which the task is chosen for manufacturing the product given the constraints and/or a previous occurrence of one or more of the manufacturing events. The iteration further comprises ranking the one or more manufacturing tasks according to a descending order of their respective propensities. The iteration further comprises visiting the one or more manufacturing tasks according to the ranking and, for each visited task, affecting to the task one or more manageable machines and resources for executing the task. The iteration may further comprise executing the one or more tasks until an occurrence of one or more of the manufacturing events. This constitutes an improved method for operating a factory.