Automated Diagnostic System "Artemis"

We introduce an automatic equipment diagnosis system that implements multimodal deep learning methods.

"Artemis" is an automatic diagnostic system that enables the identification of abnormal factors/signs by learning from the medical record data of skilled technicians. Based on the actual waveform data and medical record data accumulated by skilled technicians over decades of diagnosis in various fields, we have constructed a system for automating diagnosis using multimodal deep learning techniques (Patent No. 6853986). Additionally, by using a newly developed dedicated wireless vibration sensor, we enable retrofitting for constant remote diagnosis of equipment. Applications can be selected from both cloud-based and edge systems. [Features] - Implementation of multimodal deep learning techniques - Direct cloud connection of wireless sensors for constant remote monitoring (cloud version) - Implementation of a trained model based on medical record data accumulated by skilled technicians - Email alerts sent upon abnormal detection *For more details, please download the file below and refer to it. *If you have any questions or concerns, please feel free to contact us.

basic information

【Other Features】 ■ Choose edge/cloud server and wired/wireless sensor configurations according to the equipment ■ Automate the identification of abnormal causes and signs from normal/abnormal judgment (maximum value, RMS, CF) ■ Deep learning models continuously evolve through online learning (under development) ■ Implement the experiential knowledge of skilled workers into a multimodal neural network (Patent 6853986) ■ The dashboard is browser-based, allowing freedom of use on devices such as PCs, tablets, and smartphones ■ Centralized management of multiple equipment via server (cloud version) ■ Compatible with general rotating machinery, pumps, electric motors, speed increasers/decreasers, etc. * For more details, please download and refer to the file below. * If you have any questions or concerns, please feel free to contact us.