CAD Manufacturability Prediction Using ML-Based Model Discretization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manufacturing companies face inaccuracies and inconsistencies in predicting production times and costs due to reliance on human experience without standardized methodologies, leading to increased overhead and a lack of real-time objective appraisal tools for products specified in CAD models.
Innovation Solution
A computer-based method using machine learning and artificial intelligence to discretize digital models, infer manufacture processes, and provide near real-time predictions by determining physical attributes and applying machine learning models to generate accurate and consistent predictions for manufacturing processes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If human experience alone is used to predict manufacture appraisals, then flexibility and adaptability are maintained, but accuracy and consistency deteriorate
Solution Approach 1:
The patent introduces machine learning models as an intermediary between CAD models and manufacture predictions. These models are trained on historical manufacture data and serve as mediators that process digital models to generate accurate, consistent predictions for manufacturing time, cost, and material usage, resolving the contradiction by providing objective appraisal without requiring complex human expert systems
Solution Approach 2:
The patent replaces the mechanical system of human expert judgment with an automated machine learning-based system. By substituting human estimators with trained algorithms that process CAD models and historical data, the system achieves improved accuracy and consistency while reducing the complexity associated with managing human expertise and subjective judgment variations
2Reliability
If standardized methodologies are implemented for manufacture predictions, then consistency and accuracy improve, but flexibility and adaptability to unique cases may worsen
Solution Approach 1:
The patent implements a dynamic prediction system where machine learning models are continuously trained and updated with new manufacture data. This allows the standardized methodology to adapt to unique product cases over time, as the models learn from diverse manufacturing scenarios while maintaining consistent prediction frameworks, thus resolving the contradiction between standardization and adaptability
Solution Approach 2:
The patent utilizes parameter changes in the machine learning models by adjusting training parameters, feature weights, and model configurations based on the specific characteristics of different products and manufacturing contexts. This enables the standardized system to adapt to unique cases by modifying model parameters rather than changing the fundamental prediction methodology
3Productivity
If machine learning models are used for real-time predictions, then prediction speed and objectivity improve, but computational complexity and data processing requirements worsen
Solution Approach 1:
The patent applies preliminary action by pre-training machine learning models offline using extensive historical manufacture data before deployment. This preliminary training phase separates the computationally intensive model development from the real-time prediction phase, allowing fast, simple inference at runtime while the complexity is managed during the offline training period, thus resolving the contradiction between real-time speed and computational complexity
Data Source
AI summary
The subject technology is related to methods and apparatus for discretization and manufacturability analysis of computer assisted design models. In one embodiment, the subject technology implements a computer-based method for the reception of an electronic file with a digital model representative of a physical object. The computer-based method determines geometric and physical attributes from a discretized version of the digital model, a cloud point version of the digital model, and symbolic functions generated through evolutionary algorithms. A set of predictive machine learning models is utilized to infer predictions related to the manufacture process of the physical object.