Additive Manufacturing Data Fusion for In-Process Error Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Additive manufacturing processes face challenges in detecting and correcting qualitative printing errors in real-time, which can lead to scrap production and subsequent issues in critical products like automotive, railway, and aerospace components.
Innovation Solution
A system and method for in-process quality assurance in additive manufacturing using a data fusion component to merge control and image data streams, a geometrical classification component to assess geometric shapes, and an inspection quality component to compare and correct errors, with optional machine learning algorithms for automated corrections.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If real-time monitoring and quality assessment are implemented in additive manufacturing, then workpiece quality and error detection capability are improved, but system complexity and processing time increase
Solution Approach 1:
The system segments the quality assessment process into distinct functional components: data fusion component that integrates sensor data, geometrical classification component that identifies workpiece features, and inspection quality component that evaluates quality metrics. This modular segmentation allows each component to specialize in specific tasks, improving overall precision while managing system complexity through organized modularity.
Solution Approach 2:
The system performs preliminary quality assessment during the additive manufacturing process itself, rather than after completion. By continuously monitoring and assessing quality in real-time during production, the system detects errors early when they occur, enabling immediate correction before defects propagate through subsequent layers or production steps.
2Reliability
If continuous quality monitoring is performed during additive manufacturing, then error detection capability is improved, but processing time and computational load increase
Solution Approach 1:
The system implements rapid data processing by skipping unnecessary intermediate steps in the quality assessment pipeline. The geometrical classification component directly processes sensor data to identify critical features, and the inspection quality component immediately evaluates these features against quality criteria, bypassing redundant computations to minimize processing time while maintaining detection capability.
Solution Approach 2:
The system performs self-assessment by automatically comparing real-time sensor data with the digital twin model without requiring external intervention. The inspection quality component autonomously evaluates workpiece quality metrics and generates assessments, enabling continuous monitoring with minimal external processing overhead.
3Measurement precision
If multiple data streams from control device and image sensor system are integrated, then quality assessment accuracy is improved, but data processing complexity increases
Solution Approach 1:
The data fusion component merges multiple data streams from the control device and image sensor system into a unified quality assessment framework. By integrating these diverse data sources through the digital twin model, the system achieves comprehensive quality monitoring that leverages the complementary strengths of each sensor type, improving measurement precision through multi-source data correlation.
Solution Approach 2:
The digital twin model serves as an intermediary that mediates between raw sensor data and quality assessment results. It transforms complex multi-source data into standardized geometric representations that the geometrical classification component can process, simplifying the integration of heterogeneous data streams while maintaining assessment accuracy.
4Manufacturing precision
If geometric shapes are decomposed and compared with printing steps, then printing error detection is improved, but computational requirements and processing time increase
Solution Approach 1:
The geometrical classification component segments the workpiece geometry into discrete geometric shapes that correspond to individual printing steps. This segmentation allows the inspection quality component to compare each decomposed shape with its corresponding printing step parameters, enabling precise error detection at the level of individual print operations rather than requiring analysis of the entire workpiece.
Data Source
AI summary
A system for assisting an additive manufacturing process of a workpiece comprises an interface configured to receive a first and second data stream generated while the workpiece is being manufactured. The first data stream comprises first data, provided by a control device, representative of a first form of the workpiece, and the second data stream comprises second data, provided by a sensor system, representative of a second form of the workpiece. Machine-executable components comprise a data fusion component configured to receive the first and second data stream and to generate a merged data stream by associating the first data stream with the second data stream so that there is a common time reference between the first and second data stream in the merged data stream, a geometrical classification component configured to determine the second form, and a quality inspection component configured to compare the second form with the first form.