Manufacturing Workflow Planning With Tolerance-Based Fallback Execution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing workflow planning methods in manufacturing environments face challenges in balancing the efficiency of offline planning, which lacks flexibility to handle real-world deviations, and the computational cost of online planning, which is flexible but not verifiable.
Innovation Solution
An advanced planning approach that generates a nominal plan with tolerance models and fallback strategies, allowing for deviations within acceptable limits and enabling seamless execution by incorporating preplanned strategies to handle deviations dynamically.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If offline planning is used to generate workflows in advance, then execution efficiency is improved and no time-consuming online planning is required, but the system cannot handle deviations from assumed initial or final states resulting in plan termination
Solution Approach 1:
The system performs preliminary actions by generating tolerance models and fallback strategies offline alongside the nominal plan. These preparatory elements are embedded in the plan structure before execution, enabling the system to handle deviations without requiring online planning while maintaining execution efficiency.
Solution Approach 2:
The system applies beforehand cushioning by pre-defining fallback strategies and tolerance models that act as protective measures against potential deviations. These cushioning elements are prepared in advance and automatically activated when deviations occur, preventing plan termination while maintaining efficient execution.
2Adaptability or versatility
If online planning is used to generate workflows during production, then flexibility and robustness to real-world situations are improved, but costly planning during production is required and plans cannot be verified beforehand
Solution Approach 1:
The system performs preliminary action by generating tolerance models and fallback strategies offline before production begins. This preliminary preparation provides the flexibility to handle real-world deviations without requiring costly online planning during production, as all necessary planning elements are pre-computed and verified.
3Reliability
If manual scripting of exception handling is used, then specific error cases can be addressed, but the approach requires a lot of experience, is error-prone, and relies on trial-and-error due to difficulty of predicting all possible deviations
Solution Approach 1:
The system applies self-service by automatically generating fallback strategies and tolerance models through systematic analysis of the nominal plan and deviation possibilities. This eliminates the need for manual scripting by experienced operators, reducing configuration complexity and errors while comprehensively addressing potential deviations through algorithmic generation.
Data Source
Figure 1A~1B
Figure 2
Figure 3
AI summary
Embodiments of the present disclosure provide a method and system for planning and executing a workflow process of a manufacturing facility. The method may comprise receiving, from a user, a planning task describing the workflow process; generating, by a planning engine, based on the planning task, a nominal plan including a first plurality of procedures, wherein each procedure includes a predefined initial and final state; generating, by the planning engine, for each procedure of the first plurality of procedures in the nominal plan, a tolerance model defining acceptable deviations for each one of the predefined initial and final state; generating, by the planner engine, based on each tolerance model, at least one other plurality of procedures; and adding by the planner engine, the at least one other plurality of procedures to the nominal plan to generate an advanced plan. The method further comprises executing, by an execution engine, a first part of the advanced plan to perform the workflow process of the manufacturing facility; and continuously monitoring, by the execution engine, a current state of currently executed procedures in the first part of the advanced plan and potential state deviations of a second upcoming part of the advanced plan by predicting, in a receding horizon manner, whether the potential state deviations expected in the second upcoming part of the advanced plan are acceptable deviations.