Physics-Assisted ML Diagnosis for Additive Manufacturing Components
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing additive manufacturing devices face challenges in diagnosing aborted or failed builds and identifying performance issues, requiring significant manual effort and time to determine the root cause of failures.
Innovation Solution
A physics-assisted machine learning model is used to diagnose components of additive manufacturing devices by obtaining parameters, generating physics features, and classifying conditions to determine the health of the components, enabling rapid diagnosis without manual analysis and enhancing accuracy by considering wear and tear.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual diagnosis is used to identify failures and performance issues in additive manufacturing devices, then diagnostic accuracy can be achieved through expert analysis, but the process requires significant time and human labor
Solution Approach 1:
The patent replaces manual mechanical diagnosis with an automated machine learning system that uses sensor data and algorithms to diagnose additive manufacturing device failures. The system substitutes human expert analysis with computational models that automatically process operational parameters, sensor readings, and build data to identify root causes of failures, thereby eliminating the time-consuming manual diagnosis process while maintaining or improving diagnostic accuracy
Solution Approach 2:
The patent introduces a machine learning model as an intermediary between raw sensor data and diagnostic conclusions. This intermediary system processes and interprets complex operational data, transforming it into actionable diagnostic information. The model acts as a bridge that automatically translates device performance data into failure mode identification, replacing the need for direct human expert intervention while preserving diagnostic quality
2Loss of information
If manual expert analysis is used to determine root cause of failures, then accurate root cause identification can be achieved, but the process is difficult and time-consuming
Solution Approach 1:
The patent replaces manual expert analysis with an automated machine learning system that systematically evaluates operational parameters, sensor data, and build information to identify root causes. The system substitutes human cognitive processes with computational algorithms that can simultaneously analyze multiple data streams, thereby maintaining root cause identification accuracy while dramatically improving diagnosis efficiency and reducing the difficulty of the process
Solution Approach 2:
The patent implements preliminary action by pre-training machine learning models with historical failure data and operational parameters before deployment. The system performs preliminary analysis of device performance trends and anomalies in real-time, preparing diagnostic conclusions before full failure occurs. This preliminary processing enables rapid root cause identification when failures happen, eliminating the need for time-consuming post-failure manual analysis
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system for diagnosing an additive manufacturing device is provided. The system includes a first module configured to: obtain one or more parameters for a digital twin of a component of the additive manufacturing device based on raw data from the component of the additive manufacturing device; and generate physics features for the digital twin of the component of the additive manufacturing device based on the one or more parameters and one or more transfer functions, a second module configured to obtain one or more classifiers for classifying the component as a first condition or a second condition based on physics features; and a third module configured to: determine a health of the component based on the generated physics features of the first model and the one or more classifiers.