Ontology-Based Capability Checking for Manufacturing Process Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for checking whether a technical process can be implemented using a cyber-physical manufacturing system fail to reliably determine if the system has the necessary capabilities and if production steps can be optimized for minimized production time and resource use, often requiring a complete change in solution strategy.
Innovation Solution
A method using ontologies to describe the technical system and process, combining symbolic knowledge representation with constraint satisfaction problems to dynamically compare capabilities and optimize production steps, including filtering to reduce the search space for efficient solutions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing methods for checking technical process implementation are used, then the checking can be performed, but the methods fail to reliably determine system capabilities and cannot optimize production steps for minimized production time and resource use
Solution Approach 1:
The patent transforms the capability checking problem into a constraint satisfaction problem with optimized parameters. By defining capabilities as formal parameters with specific constraints (time, resources, system states), the method enables both reliable determination and simultaneous optimization of production parameters through constraint solving algorithms.
Solution Approach 2:
The patent introduces an intermediary constraint satisfaction problem formulation that mediates between system capabilities and production optimization. This intermediary layer translates capability queries into constrained optimization problems, allowing reliable capability assessment while inherently optimizing production time and resource usage through the constraint solving process.
2Adaptability or versatility
If specialized procedures are adapted to specific tasks, then the solution can be tailored, but changes and extensions require a complete change in solution strategy
Solution Approach 1:
The patent creates a universal constraint satisfaction problem framework that can handle multiple task types and system configurations through a single unified approach. The same CSP formulation and solving methodology applies whether checking basic capability, optimizing production schedules, or evaluating system reconfiguration, eliminating the need for complete strategy changes when tasks evolve.
Solution Approach 2:
The patent employs dynamic constraint satisfaction where constraints and variables can be modified to reflect changing tasks and system states. The framework adapts to new requirements by adjusting constraint parameters and adding/removing variables rather than requiring fundamental strategy changes, enabling flexible response to task evolution.
3Reliability
If comprehensive checking of all production conditions is performed, then feasibility can be determined, but the checking process becomes computationally intensive
Solution Approach 1:
The patent performs preliminary constraint propagation and filtering before full constraint solving. By pre-processing the constraint satisfaction problem to eliminate obviously infeasible paths and simplify constraints, the method maintains complete verification of all production conditions while significantly reducing the computational time required for the exhaustive checking process.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a method for checking whether and/or how a technical process can be carried out using a technical system with two or more units which have technical capabilities that are potentially relevant to the technical process. At least one ontology is used which describes the technical system and the technical process, and a constraint satisfaction problem is generated using the at least one ontology. In the method according to the invention, solutions are found for the constraint satisfaction problem and are preferably output.