The AI-based visual inspection package (OKIKAE for AI visual inspection) has been updated to support chemical products. Inspection and production of various uneven materials, which was difficult to judge with conventional image processing inspection (rule-based) […]

In chemical plants, there are variables such as material input amounts, operating parameters, and conditions such as the surrounding environment in each process. We collected those data and applied them to BayoLink while considering causal relationships. There are multiple processes such as the 1st process and the 2nd process, and they […]

Data acquired by a robot and images of cracks detected by AI are superimposed on a 3D point cloud of a structure acquired by a 3D scanner. By superimposing the 3D point cloud of the building/structure acquired by the robot and the image of the detected cracks, the damage […]

Access RBM (Risk-Based Maintenance) SaaS equipped with predictive maintenance AI via the cloud, enter necessary data into the software, perform risk assessment, and implement risk-based maintenance (RBM). (Sa […]

We have developed a real-time risk assessment system based on probabilistic inference (AI technology that probabilistically analyzes and infers complex causal relationships) and TAG information obtained by detecting process anomalies using invariant analysis technology. , accidents at manufacturing sites […]

We have developed a prototype of a sensor that measures pipe thinning without contact using the magnetic field around the pipe, and conducted a demonstration experiment. As a result of calculating the position and depth of thinning from the magnetic field around the pipe measured by the sensor and the magnetic field, we […]

Build and verify a predictive model with data on operating conditions when an abnormality in the indicated value of a liquid level gauge occurs in a plant, examine threshold values, and use the machine learning scheme to relearn the current operating data. , implemented an anomaly detection model for liquid level control. Based on the indicated value of the liquid level gauge […]

Before undertaking the work, the person in charge, the engineer, and the person in charge of analysis gathered together and held a workshop. Through the workshop, we formed a common understanding of the events to be analyzed, the quality and quantity of data, the policy of analysis, the final output, etc. Prospects for productivity improvement and development […]

It collects and analyzes a large amount of time-series data collected from many sensors, automatically models the invariant relationships between sensors that hold in normal times, and creates relational expressions. It automatically learns and monitors the behavior of the system and detects signs of anomalies […]