Additive Manufacturing Part Sorting via Predictive Tray Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In additive manufacturing, the manual sorting of parts fabricated in powder-based systems is inefficient due to similarity in characteristics among parts from different orders, leading to frequent errors and increased costs, as automated processes often result in unorganized piles making it difficult to identify and correctly sort parts.
Innovation Solution
The system predicts characteristics of parts from 3D computer models and assigns them to separate build trays based on these predictions, then uses sensors to measure and match fabricated parts with predicted characteristics for accurate sorting, reducing the likelihood of errors by minimizing the number of similar parts in the same build tray.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If parts from different orders are fabricated together in the same build tray to reduce costs, then manufacturing efficiency and cost-effectiveness improve, but sorting accuracy deteriorates due to similarity in characteristics among parts
Solution Approach 1:
The system segments parts into different build trays based on their predicted characteristics (mass, volume, dimensionality, shape, color, density). By dividing the manufacturing process into separate trays for similar parts, the system maintains sorting accuracy while still achieving batch production efficiency.
Solution Approach 2:
The system performs preliminary prediction of part characteristics from 3D computer models before fabrication. This advance classification allows parts to be assigned to appropriate build trays beforehand, preventing sorting errors before they occur and enabling efficient batch processing.
2Productivity
If automated sorting processes are used to improve speed, then productivity increases, but sorting accuracy deteriorates because parts end up in unorganized piles
Solution Approach 1:
The system replaces manual mechanical sorting with an automated optical/electronic measurement system using sensors. Sensors quickly measure actual characteristics of fabricated parts and compare them against predicted values from 3D models, enabling fast automated identification and sorting without creating unorganized piles.
Solution Approach 2:
The system implements feedback by measuring actual part characteristics with sensors and comparing them against predicted characteristics from 3D computer models. This feedback mechanism enables automated systems to accurately identify and sort parts based on verified measurements rather than relying on visual inspection or manual handling.
3Measurement precision
If manual sorting is used to maintain accuracy, then sorting precision improves, but productivity decreases due to time-consuming comparison processes
Solution Approach 1:
The system replaces time-consuming manual comparison and sorting with automated sensor-based measurement and computer-controlled sorting. Sensors rapidly measure part characteristics and computers automatically compare measurements against predicted values, achieving both high accuracy and fast sorting speeds.
Solution Approach 2:
The system uses 3D computer models as digital copies of the parts before fabrication. These models contain predicted characteristics that serve as reference templates for sorting. By comparing actual measured parts against these digital copies, the system achieves rapid automated sorting with high accuracy without manual intervention.
4Productivity
If similar parts are placed in the same build tray to maximize utilization, then resource efficiency improves, but error identification becomes more difficult
Solution Approach 1:
The system segments parts into different build trays based on their predicted characteristics to prevent similar parts from being mixed. This segmentation maintains high build tray utilization by fully loading each tray while ensuring that parts within each tray are sufficiently differentiated for easy identification and sorting.
Solution Approach 2:
The system uses feedback from sensor measurements and comparison against 3D model predictions to automatically identify and sort parts. This feedback mechanism makes error detection straightforward by objectively comparing measured characteristics against expected values, eliminating the difficulty of manually detecting similarities among parts.
Data Source
AI summary
In one example, an executable program on a computer-readable storage medium that instructs a processor to predict a characteristic of at least some parts in plural orders from a corresponding computer model of the part. The program also instructs the processor to assign, based on the predicted characteristics, models of at least some similar parts from different orders to different build trays usable to fabricate the parts. The program further instructs the processor to, for each build tray, sort the fabricated parts into the plural orders by matching a measured characteristic of a part to a predicted characteristic of unsorted parts from the build tray.