Additive Manufacturing Defect Prediction for Real-Time Process Stopping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current additive manufacturing quality control methods are limited to detecting defects after they occur, leading to resource wastage and potential safety issues, as they do not effectively anticipate and prevent defects in real-time.
Innovation Solution
A supervision system that predicts defects by analyzing product geometry data, production history, and real-time measurements from the manufacturing device, allowing for proactive process stopping or corrective actions to prevent defects, using predictive models like machine learning and decision trees to determine defect types and seriousness levels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional defect detection methods are used, then defects can be detected after they occur, but resources are wasted and manufacturing time is lost
Solution Approach 1:
The supervision system performs preliminary defect prediction by analyzing multiple data sources (product geometry data, production data, production history data, and real-time measurements) before defects actually occur. This allows the system to anticipate potential defects and trigger corrective actions in advance, preventing the need to discard partially manufactured parts and avoiding loss of manufacturing time.
2Reliability
If conventional defect detection methods are used, then defects can be detected after they occur, but material and resources are wasted
Solution Approach 1:
The system predicts defects before they occur by continuously monitoring production parameters and comparing them against historical data and geometric requirements. When a potential defect is predicted, the manufacturing process can be stopped or corrected, preventing waste of materials that would otherwise be consumed in producing defective parts.
3Reliability
If real-time defect prediction is implemented, then defects can be prevented, but system complexity increases
Solution Approach 1:
The supervision system is designed as a multi-functional platform that handles multiple data types (geometry data, production data, history data, sensor measurements), performs various analytical functions (data processing, defect prediction, severity assessment), and executes different corrective actions. This universal system consolidates multiple functions into a single integrated platform, managing complexity through functional integration rather than proliferation of separate systems.
Solution Approach 2:
The supervision system acts as an intermediary layer between the additive manufacturing device and the control system. It receives data from the manufacturing device, processes this information through predictive algorithms, and generates appropriate control commands or alerts. This intermediary architecture isolates the complexity of defect prediction algorithms from both the manufacturing device and the final decision-making processes.
4Measurement precision
If comprehensive data analysis is performed for defect prediction, then prediction accuracy improves, but processing time increases
Solution Approach 1:
The system performs preliminary processing and analysis of product geometry data, production data, and production history data before the actual manufacturing process begins. This pre-processing establishes baseline expectations and prediction models in advance, allowing the real-time monitoring phase to focus primarily on comparing current measurements against pre-established criteria, thereby maintaining both accuracy and speed during critical manufacturing operations.
Data Source
AI summary
Method for supervising an additive manufacturing process performed by an additive manufacturing device (1), comprisingproviding a set of data to a supervision system (6), said set of data comprising product geometry data, production data, production history data, and measurements (7) from said additive manufacturing device (1), and, at said supervision system (6):predicting the occurrence of a defect (S1) from said set of data,when an occurrence of a defect is predicted, determining a type of defect (S2), and deciding for stopping (8, 5) or not (S3) said additive manufacturing process according to said type of defect and of a seriousness level.