Image Analysis AI [TDSE Eye] Utilization Seminar - Streamlining Visual Inspection with AI!

Build AI models without programming! It is possible to visually confirm anomalies.

The use of AI requires specialized knowledge and tools, and in reality, it is limited to a few companies that have secured personnel knowledgeable in AI and IT. Our company has built an image recognition AI modeling platform called "TDSE Eye" and has started providing various products necessary for real business. AI models can be built without programming (Web interface), and AI will streamline visual inspections, such as maintenance tasks and product quality checks. **Features of TDSE Eye** - AI model building without programming (Web interface) - Continuously provides cutting-edge beneficial image AI technology in a cloud environment - Anomaly detection AI is built using only normal images - It is possible to visually confirm abnormal areas - Portable AI server enables anomaly degree calculation on edge devices like PCs *For more details, please refer to the PDF document or feel free to contact us.*

• Visual Inspection Equipment

Select only the products that require human inspection and automate the process! It's okay to inspect only a few percent of the total inspection count!

Visual inspection of defective products on production lines by humans involves a high workload and leads to inconsistencies in inspection quality. Therefore, we first use AI to identify products with a high likelihood of defects. Only those selected products are then visually re-inspected by humans, ensuring consistent quality while reducing manpower. Our anomaly detection service, 'TDSE Eye', implements advanced algorithms to streamline visual inspections for maintenance tasks and product quality verification. 【Case Overview】 ■Challenges - Time and labor costs are high - The work is dependent on individuals, leading to inconsistencies in inspection quality ■Benefits - AI inspection selects only the products that require human inspection, reducing manpower - AI ensures consistent inspection quality *For more details, please refer to the PDF document or feel free to contact us.

• Visual Inspection Equipment

Automatically detect cracks and other issues using images captured by drones with AI! This makes work safer and more efficient.

Regular inspections of concrete deterioration (such as cracks) in tunnels and bridges conducted by human eyes are very time-consuming and can sometimes be dangerous. Therefore, we use images captured by drones to automatically detect cracks and other issues with AI, making the inspection process safer and more efficient. Our anomaly detection service, 'TDSE Eye,' allows for the rapid construction of complex models in a cloud environment, enabling implementation and operation without specialized knowledge. [Case Overview] ■ Challenges - Time and labor costs are high, the work is dependent on individuals, making skill transfer difficult, and there is a risk of falling debris from deteriorated areas. ■ Benefits - Automation and reduction of inspection work through AI, maintaining consistent inspection quality with AI, and ensuring safety through drone imaging. *For more details, please refer to the PDF document or feel free to contact us.

• Visual Inspection Equipment

Improving work efficiency and reducing costs is possible! AI can detect inspections and checks instead of visual inspections.

"Veteran workers are relied upon, and inspections are becoming person-dependent." "Visual inspections take time and human costs." "We cannot introduce AI without specialized knowledge." All these concerns will be resolved by 'TDSE Eye.' AI will take over the visual inspections that have relied on humans until now, enabling efficiency and cost reduction in operations. It can be implemented and operated without specialized knowledge, and it can be built at low cost and in a short period. 【Features of TDSE Eye】 ■ Implementation of advanced algorithms ■ Streamlining visual inspections for maintenance tasks and product quality checks ■ By building AI models in the cloud, high-performance options are always available ■ Easy operation through a web interface to build AI ■ Offline execution of AI inference at the edge ■ Extensive track record in the field of image AI *For more details, please refer to the related links or feel free to contact us.

• Visual Inspection Equipment

Always providing cutting-edge and beneficial image AI technology in a cloud environment! There is an introductory video that fully explains the features!

"TDSE Eye" is an anomaly detection service that streamlines visual inspection tasks, such as equipment maintenance and product quality verification, using AI for appearance inspections. The specific areas of the identified images that are abnormal are quantified and can be visualized as a heatmap. Additionally, it is possible to run the AI model built in the cloud on edge devices, such as PCs that are disconnected from the network. 【Features】 ■ AI model construction without programming (Web interface) ■ Continuously provides cutting-edge beneficial image AI technology in a cloud environment - Anomaly detection AI is built solely on normal images - Allows for visual confirmation of abnormal areas ■ AI model can also be utilized on edge devices like PCs that are disconnected from the network *For more details, please refer to the PDF document or feel free to contact us.

• Visual Inspection Equipment

Carpets, leather products, bottles, etc.! Detecting any anomalies with unusual appearances (such as unknown foreign objects).

The appearance inspection system "TDSE Eye" determines abnormalities in various textures and objects. When it detects abnormalities such as chips, cracks, fraying, tears, or breaks, the degree of abnormality in the appearance of the image differs from the normal images used for model creation. It may also be capable of detecting any unusual appearance abnormalities (such as unknown foreign objects). Additionally, this product can be implemented and operated without specialized knowledge. It can build complex models in a short period within a cloud environment. 【Examples of Abnormality Detection (Partial)】 < Texture Types > ■ Carpet ■ Grid ■ Leather and leather products ■ Tile ■ Wood and more < Object Types > ■ Bottle ■ Cable ■ Capsule ■ Food ■ Wood ■ Tablet ■ Screw ■ Brush ■ Electronic components ■ Zipper and more * For more details, please refer to the related links or feel free to contact us.

• Visual Inspection Equipment

Train AI using only a small amount of normal images! Implementation and operation are possible without specialized knowledge.

"TDSE Eye" implements advanced algorithms to streamline visual inspections such as equipment maintenance and product quality checks. When building AI with normal and abnormal images, the emergence of various abnormal conditions can prolong the image collection process, resulting in significant time required before implementation and operation can begin. This product allows for AI training using only a small number of normal images, enabling quick and cost-effective AI development. Since there is no need to prepare abnormal images, image collection is simplified. *If there are variations in the normal state, all variations must be trained as normal. [Features] ■ AI built using only normal images ■ Implementation of advanced algorithms ■ Consistently high-performance anomaly detection AI ■ Easy implementation and operation without specialized knowledge ■ Offline execution of AI inference at the edge *For more details, please refer to the related links or feel free to contact us.

• Visual Inspection Equipment

By combining with a 360-degree camera, we can streamline internal inspections of products! This reduces the costs of visual inspections for appearance checks and quality confirmation.

In visual inspections, automating the inspection process is difficult due to the structure of iron pipes, resulting in high time and labor costs. By combining the abnormal detection service "TDSE Eye" with a 360-degree camera, we can detect abnormalities such as internal damage from the captured images and identify areas that require visual inspection. This eliminates the need to verify everything visually, achieving a reduction in labor and time costs. Additionally, our product allows for easy operation through a web interface to build AI, enabling implementation and operation without specialized knowledge. 【Overview】 ■BEFORE - Automating the inspection process is difficult due to the structure of iron pipes - High time and labor costs ■AFTER - Efficiently capturing the interior with a 360-degree camera - Reducing labor and time costs by inspecting camera images with TDSE Eye *For more details, please refer to the related links or feel free to contact us.

• Visual Inspection Equipment

Only a few percent of the total inspections are needed! Implementation and operation are possible without specialized knowledge.

In human visual inspections, it is necessary to conduct a full inspection, which incurs time and labor costs, and the work is subjective, leading to inconsistencies in inspection quality. By utilizing the anomaly detection service "TDSE Eye," we can automate the process by selecting only those products that require human inspection through AI inspections, thereby reducing manpower and ensuring consistent inspection quality. Additionally, this product can quickly build complex models in a cloud environment and execute AI inference offline at the edge. 【Overview】 ■BEFORE - Time and labor costs are high - Work is subjective, leading to inconsistencies in inspection quality ■AFTER - Automate the process by selecting only products that require human inspection through AI inspections - Ensure consistent inspection quality with AI *For more details, please refer to the related links or feel free to contact us.

• Visual Inspection Equipment

With drone images and TDSE Eye, we can conduct inspection work safely and securely!

In visual inspections of tunnel walls by humans, the work is subjective and it is difficult to pass on skills, and there is a risk of deteriorated areas falling. By utilizing the abnormality detection service "TDSE Eye" and conducting inspections with drone images and AI, it is possible to standardize inspection quality and automate and reduce manpower in the work. This product allows for easy operation through a web interface to build AI, and it can be implemented and operated without specialized knowledge. [Overview] ■ BEFORE - Time-consuming and costly in terms of manpower, work is subjective and skill transfer is difficult, with a risk of deteriorated areas falling. ■ AFTER - Automate and reduce manpower in inspection work with AI, standardize inspection quality with AI, and ensure safety with drone photography. *For more details, please refer to the related links or feel free to contact us.

• Visual Inspection Equipment

Build complex models in a cloud environment in a short time! It's also easy to use AI from individual applications.

The anomaly detection service "TDSE Eye" provides access to always new, high-performance AI models built in the cloud. Model construction can be done using only a few normal images, eliminating the need to collect abnormal images. Additionally, by using a portable AI server, the AI models built in the cloud can be utilized on edge devices such as PCs that are disconnected from the network. It is also easy to use AI from individual applications. 【Five Reasons to Choose Us】 ■ Always new, high-performance anomaly detection AI ■ Can be implemented and operated without specialized knowledge ■ Built at low cost and in a short period ■ Offline execution of AI inference at the edge ■ Extensive track record in the field of image AI *For more details, please refer to the related links or feel free to contact us.

• Visual Inspection Equipment

Image recognition on cloud edge devices! We also offer system implementation and various support tailored to your company's environment.

"TDSE Eye" implements advanced algorithms to streamline maintenance tasks and visual inspections of product quality. An AI model is built using only images of normal conditions. In the cloud, simply uploading images allows for the identification of abnormal images. On edge devices, abnormal image identification is performed by downloading the portable AI server and AI model. We also offer system implementation tailored to your company's environment and various support services, so please feel free to consult us when needed. 【Process until Implementation】 ■STEP 01: Prepare Images (If there are variations in normal conditions, images corresponding to all variations are required) ■STEP 02: Build AI Model ■STEP 03: Image Identification (Cloud and Edge Device Compatible) *For more details, please refer to the related links or feel free to contact us.

• Visual Inspection Equipment

Easily accessible AI from individual applications! Complex models can be built in a short time in a cloud environment!

"TDSE Eye" implements advanced algorithms to streamline visual inspections such as equipment maintenance and product quality checks. By building AI models in the cloud, high-performance solutions are always available. AI can be constructed easily through a web interface. Our company has handled numerous projects in the field of image AI, and this service reflects that experience. 【Features】 ■ Always high-performance anomaly detection AI ■ Can be introduced and operated without specialized knowledge ■ Built at low cost and in a short period ■ AI inference can be executed offline at the edge ■ Extensive track record in the field of image AI *For more details, please refer to the PDF document or feel free to contact us.

• Visual Inspection Equipment

Berry Inc. (Headquarters: Taito-ku, Tokyo; CEO: Hiroshi Nakano) will hold a free online seminar titled "Evolved with AI! - A Comprehensive Explanation of eQMS 'QMSmart' from Basics to Latest Updates," which will explain the overall picture and latest updates of the cloud-based eQMS "QMSmart" with a live screen demo, on August 21, 2026 (Friday) from 14:00 to 15:00. This seminar is aimed at those considering the introduction of eQMS, as well as quality assurance, design development, and manufacturing personnel at medical device manufacturers who feel limited by paper and Excel-based QMS operations, as well as management. Participants will be able to see the evolved QMSmart in its entirety in one hour, following this summer's major update.

Takaya Corporation will be exhibiting at the "36th Microelectronics Symposium (MES2026)" held at Osaka University Nakanoshima Center from September 16 (Wednesday) to 18 (Friday), 2026. MES2026 is one of Japan's leading technical symposiums in the field of electronics packaging. The symposium will focus on implementation technologies that support the high functionality, miniaturization, and high reliability of electronic devices, featuring the latest research results and technological trends from universities, research institutions, and companies. Additionally, technical exchanges and information sharing will take place through corporate exhibitions. At our booth, we will introduce the features of the Flying Probe Tester "APT Series" and solutions that contribute to quality improvement, as well as offer consultations regarding inspection processes. We invite you to stop by the Takaya booth during your visit.

In this session, we will thoroughly analyze case studies of three major chemical accidents that occurred in the past! When we unravel the root causes of these accidents, we find a common pattern of "lack of safety management systems" rather than just equipment failure or chemical reactions. We will introduce specific solutions to bridge the three gaps of "Visibility," "Control," and "Preparedness," and discuss Chemwatch's efforts to prevent risks before they occur. For any questions, please contact us at japan@chemwatch.net. We sincerely look forward to your participation.

☆☆☆ Urgent Visit to Japan & Emergency Event Announcement ☆☆☆ Dr. Yuval Levy, CEO of ISCFDC and developer of EPPIC, who has been involved in CFD research at NASA Ames Research Center for about five years, is coming to Japan! He will demonstrate and explain advanced CFD technologies, application cases, and the latest features of EPPIC himself. □ Date and Time □ July 31, 2026 (Friday) 13:00-18:00 (Registration starts at 12:30) <Part 1: Post-Processor Session> 13:00-16:40 <Part 2: Pre-Processor Session> 16:50-17:50 <Networking Event> 18:10-19:40 □ For more details □ Visit the Hulinks page https://www.hulinks.co.jp/seminar/eppic/

While the business utilization of generative AI (such as ChatGPT) is rapidly advancing, the "security risk of confidential information such as on-site data, proprietary manuals, and customer information being sent to the cloud" has become a significant issue. Until now, the high cost of GPU servers and the necessity of relying on the cloud for model intelligence have been challenges. However, this premise is now undergoing a major change. With the emergence of Google's latest open model, 'Gemma4', we have finally entered an era where it is possible to operate with "intelligence equal to or greater than that of the cloud" at "one-third the hardware cost of traditional setups," and within "a secure, company-exclusive server completely isolated from the internet." We will introduce the latest trends in AI implementation along with our company's latest DX solutions.