By training AI with normal manufacturing operation data, we can detect unforeseen troubles and provide a system that supports on-site engineers.
【Do you have any of these concerns?】 - No abnormal data (or very little) - Many false detections (low accuracy) - Not used on-site (becoming a black box) At SAILESS, to prevent your valuable system from becoming a black box, we prioritize the voices from the field and provide usable AI services to understand the signs of impending disasters. 【What SAILESS can do】 - Build systems using normal data - Detect signs of trouble even with no prior experience - User-operated (easy to retrain detection models) and scalable 【Because we can predict abnormalities, we improve profits by solving challenges in manufacturing and maintenance】 By implementing SAILESS, you can avoid emergency stops on production lines and reduce production losses. Additionally, in the maintenance department, you can incorporate condition-based maintenance (CBM), contributing to the reduction of maintenance costs and prevention of serious accidents, leading to optimal plant and factory operations.
Inquire About This Product
basic information
【Service Content (BTO Model)】 1. Issue Hearing and Data Analysis (PoC) To provide a just-fit anomaly detection system for various on-site challenges faced by customers, we conduct hearings on objectives, issues, anomalies to be detected, and the current data acquisition status. We can respond through data analysis. *Even if the target data is limited to the site and not accumulated, we will support the construction of a database. 2. Selection of Anomaly Detection Algorithms The optimal algorithm for anomaly detection varies depending on the anomalies to be detected. At SAILESS, we select the best method from various algorithms necessary for anomaly detection based on data analysis results. 3. Construction of Detection System Using the optimal algorithm, we calculate and visualize a new metric called [Anomaly Degree]. The monitoring screen can be constructed using various PIMS (Plant Information Management System) tools or OSS (Open Source Software). We also support the development of custom monitoring screens for customers. To prepare for false detections due to aging of target equipment, we provide a system that supports the re-learning of AI models, allowing users to operate and evaluate results themselves.
Price information
Please contact us.
Delivery Time
Applications/Examples of results
【Steady-State Plant】 Overall Process: Revenue improvement through reduction of chocolate stops, Cost reduction in maintenance through early detection of failures Distillation Tower, Reactor: Prevention of quality degradation through early detection of abnormalities Sensors: Revenue improvement by avoiding sudden failures 【Cement Plant】 Kiln: Prevention of quality degradation through early detection of abnormalities 【Assembly】 Control Devices: Efficiency improvement in maintenance *Also applicable to other non-steady batch plants.
catalog（5）
Download All Catalogs![AI Anomaly Detection System Construction Service [SAILESS]: Case Study Interview](https://image.mono.ipros.com/public/catalog/image/01/c33/719730/IPROS11198032261643745617.png?w=120&h=170)
![Abnormal Signal Detection System Construction Service [SAILESS] Service Introduction Materials](https://image.mono.ipros.com/public/catalog/image/01/10d/692649/IPROS17298787988696157787.png?w=120&h=170)
Company information
Our company was established in 1983 to develop the information systems department of UBE Corporation, which operates in a wide range of fields including chemicals, construction materials, industrial machinery, and energy. We have extensive experience not only in managing functions unique to the manufacturing industry, such as research and development, design, production, and quality control, but also in building and operating information systems for procurement, sales, logistics, accounting, and human resources. Some of these are offered as system products (packaged software).