Bright Machines. has been granted a patent for a vision analytics and validation (VAV) system designed to enhance robotic assembly inspections. The system utilizes a trained neural network to classify components as good, bad, or uncertain, and includes features for operator review and retraining based on inspection outcomes. GlobalData’s report on Bright Machines gives a 360-degree view of the company including its patenting strategy. Buy the report here.
Smarter leaders trust GlobalData
According to GlobalData’s company profile on Bright Machines, Pose estimation was a key innovation area identified from patents. Bright Machines's grant share as of June 2024 was 53%. Grant share is based on the ratio of number of grants to total number of patents.
Vision analytics system for robotic assembly inspection
The granted patent US12026865B2 introduces a Vision Analytics and Validation (VAV) system designed to enhance the inspection processes within robotic assembly systems. Central to this system is an image combiner that integrates multiple images from various camera-based inspection machines to produce a unified image of a circuit board. The system employs grid and Region of Interest (ROI) logic to segment this image into specific areas, focusing on components and their associated pads. A trained neural network functions as a three-way classifier, categorizing each component as "good," "bad," or "do not know," based on the analysis of these regions. Additionally, a clustering system is utilized to establish a model of "good" components, allowing for the classification of any object outside this cluster as "bad."
Further enhancements to the VAV system include the ability to assign probabilities to classifications, with adjustable thresholds set between 80% and 98%. The system can process images under varying conditions, such as different lighting and angles, and even includes X-ray images. Optical character recognition (OCR) is integrated to assist in identifying components, while masking logic filters out irrelevant information before data is classified. The system also features a synthetic training data generator to create additional training data for components not present in the original dataset. An overtagging trigger and alert system is included to monitor classification rates, allowing operators to override classifications when necessary. This comprehensive approach aims to improve the accuracy and reliability of inspections in robotic assembly environments.
To know more about GlobalData’s detailed insights on Bright Machines, buy the report here.
Data Insights
From
The gold standard of business intelligence.
Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.
GlobalData, the leading provider of industry intelligence, provided the underlying data, research, and analysis used to produce this article.
GlobalData Patent Analytics tracks bibliographic data, legal events data, point in time patent ownerships, and backward and forward citations from global patenting offices. Textual analysis and official patent classifications are used to group patents into key thematic areas and link them to specific companies across the world’s largest industries.
