Advancing Injection Molding Monitoring using Object Detection

Based on Patent Research | CN-111098464-B (2024)

Injection molding machines often struggle to maintain steady operational states during production. These inconsistencies cause defects like short shots or warped parts, which usually require slow manual checks. Object detection uses camera systems to find and locate specific flaws in real time. This automated approach identifies errors immediately to reduce scrap and stabilize machine performance. Operators can now ensure consistent quality across every batch while cutting down on costly downtime and wasted materials.

SOLUTION OVERVIEW

Upgrading Manual Monitoring to AI Detection

Object detection technology provides a direct solution for plastics manufacturing by identifying specific structural flaws in real time. This vision system starts by capturing high resolution images of molded parts as they exit the machine. It then processes these visuals using specialized algorithms to locate anomalies like flash or short shots. Finally, the system provides an immediate signal to the production controller, allowing for the rapid sorting of defective components before they reach the assembly stage.

By integrating this automated vision into existing assembly lines, manufacturers can maintain constant oversight without relying on periodic manual inspections. This approach is like having an expert quality inspector with a magnifying glass watching every single cavity during every cycle. These smart systems enable smoother workflows and more reliable machine uptime. Ultimately, adopting such technology empowers producers to refine their molding processes, ensuring that every batch meets rigorous quality standards while significantly reducing material waste.

HOW IT WORKS

Unlocking Production States in Video

Capturing High Resolution Part Images

The system begins by utilizing high resolution cameras to take clear photos of every plastic part as it is ejected from the molding machine. These images capture the surface details of the components, providing the necessary visual data for a comprehensive quality assessment.

Detecting Structural Anomalies and Flaws

The software processes the live visual feed using specialized algorithms to pinpoint specific defects such as short shots, flash, or warpage. By scanning each part in real time, the system accurately locates structural inconsistencies that are often difficult for human inspectors to catch quickly.

Communicating Real Time Quality Data

After identifying a defective piece, the system sends an immediate electronic signal to the production controller to initiate the sorting process. This rapid communication allows for the removal of faulty items and provides operators with the data needed to stabilize machine performance.

Potential Benefits

Immediate Waste Reduction

By identifying short shots and flash in real time, the system prevents defective parts from reaching later assembly stages. This rapid detection significantly lowers scrap rates and saves valuable raw materials.

Improved Production Consistency

Automated monitoring stabilizes machine performance by providing constant oversight of the molding cycle. This ensures that every batch maintains high quality standards without the variability inherent in manual checks.

Enhanced Operational Efficiency

The vision system works alongside existing lines to minimize costly downtime caused by unrecognized molding errors. Faster error identification allows operators to address machine inconsistencies before they lead to major failures.

Data Driven Quality Control

High resolution images offer precise insights into material flow and mold filling across every cavity. This objective data helps manufacturers refine their molding processes and make more informed decisions about equipment maintenance.

Implementation

1 Camera Hardware Installation. Mount high resolution cameras at the molding machine discharge area to capture detailed images of parts as they exit.

2 Lighting and Environment setup. Install consistent industrial lighting to eliminate shadows and ensure clear visibility of potential surface defects like flash or warpage.

3 Detection Model Configuration. Configure the object detection algorithms to recognize specific structural anomalies and set threshold levels for acceptable quality standards.

4 Production Controller Integration. Establish a digital connection between the vision system and the machine controller to enable automated sorting of defective components.

5 System Calibration Protocol. Perform initial test runs with known defect samples to verify detection accuracy and refine the automated signaling process.

6 Operator Dashboard Launch. Deploy a real time interface for machine operators to monitor detection results and track performance data across production batches.

Source: Analysis based on Patent CN-111098464-B "State determination device and method" (Filed: August 2024).