AI Appearance Inspection [RoxyAI] Function Upgrade
From human to robot for visual inspection! Introduction of features added with the version upgrade.
Would you like to start automating visual inspections, which are currently judged by human hands, eyes, and experience? Eliminate the black box! Understand by seeing and touching! Usable on-site! Introducing the AI visual inspection product 【RoxyAI】, which pursues simplicity for "on-site usability": - Determines whether defects are included in the inspection area / identifies defect types / detects unknown defects - Learns features from a small amount of training data / utilizes a unique high-speed, high-precision AI algorithm (patent pending) - Features a unique visualization technology that is easy to understand even for those unfamiliar with AI (patent pending) - Allows users to grasp weaknesses by touching and confirming AI sensitivity, making it easy for anyone without AI knowledge Our company leverages its experience in designing and manufacturing dedicated machines, inspection machines, in-factory logistics, and automation systems in the FA field to provide proposals and support for AI verification, actual machine evaluation, and automation of visual inspection processes. We also support the automation of production processes, including pre- and post-inspection processes, so please feel free to contact us first. Sanki Co., Ltd. (Atsuta-ku, Nagoya) is the official distributor of RoxyAI.
basic information
**Features Added in the Version Upgrade** 1. Hint Function If the AI does not capture the characteristics of defects, accuracy will not improve. The hint function added in V1.14 makes it easier to learn the characteristics of defects by masking areas other than the defective parts. 2. Enhanced Image Augmentation Additional augmentation methods have been added to enrich the types of augmented images. 3. Strengthened Learning Function A learning method has been added that is expected to improve generalization performance at the cost of learning speed. Learning time will be approximately 1.5 times longer (varies depending on learning settings). You can specify whether to use this function in the learning settings screen. 4. Expansion of Base Models for Type 3 Base models that can be used with Type 3 have been added. You can select base models according to your needs, such as speed-oriented or accuracy-oriented. 5. Addition of Sub-Windows Images filtered by purpose are displayed in a sub-window, reducing the effort required to search for images. 6. Batch Learning Result Comparison You can view and compare various metrics and settings from the results of batch executions. 7. Others Improvements in usability, speed enhancements, and bug fixes, etc.
Price information
The three machines are proposal-based and support our customers' "troubles." 【Purchase Method】 In addition to cash lump-sum purchases, we can propose payment plans that meet our customers' needs in collaboration with partner leasing companies. Recently, there are also leases that can ultimately be converted into company assets. Additionally, we can assist with various subsidies. If you have any concerns, please feel free to contact us.
Delivery Time
Applications/Examples of results
【Other Features】 - The AI automatically extracts suspicious data from the learning data. - Focused checks on intervals where normal and defective data are mixed on the histogram. - Human judgment on the handling of suspicious data by reviewing enlarged images and overall images. - If issues are found, the corresponding images are deleted or normal/defective classifications are swapped. 【Examples of Achievements】 - Scratches (line scratches, dents) on metal products, processing defects (surface inspection after welding and burr removal). - Scratches (line scratches, dents, sink marks) on resin molded products, processing defects (burrs, chips). - Sensory inspection for shape defects such as weld beads and soldering. *This does not guarantee the detection of all defects.
