Industrial machine health monitoring has been around for around for decades in various semi-automatic forms, but AI tools at the edge are beginning to add sophistication to the process. Various steps are being integrated for detecting potential faults and reporting them to plant operators in time for repairs that can reduce the cost of replacing equipment.
A new approach announced Monday comes from Tokyo-based TDK through its subsidiary TDK SensEI. Its new product line, edgeRX, uses AI on edge sensors. With AI algorithms and edge computing via an ARM Cortex microprocessor, the system can give operators real-time health monitoring and actionable alerts.
The company said edgeRX is designed to help reliability engineers, maintenance techs and plant managers quickly deploy the product to then find problems before they escalate, thereby enhancing overall operational efficiency.
“Edge AI is very hard to do, but out of the box we do AI for you. We ship a gateway and sensors devices and you plop down and start collecting data. We’ll build models up into the cloud,” said Kyle Arnold, vice president of global marketing at TDK SensEI, in an interview with Fierce Electronics. A dashboard that is “very intuitive” helps operators detect problems.
The edgeRX relies on vibration and temperature sensing in the initial version, but will likely add magnetometer, ultrasonic, pressure and humidity in the future, he said. The entire system has already been deployed in 100 different factories in a soft launch.
Customers make a one-time purchase of the hardware and enter into a multi-year subscription agreement for the software. Pricing was not disclosed.
Arnold said edgeRX faces limited competition in a fragmented market. “Ours is embedded AI and out of the box, and not many companies do that,” Arnold added. “We’re displacing solutions mostly in the cloud.” Seimens, AWS and Monitron are among the companies addressing a similar market. Some competitors only provide vibration sensing, while edgeRX has both vibration and temperature. “We think we are much more advanced with a really low touch implementation,” Arnold said.
At installation, edgeRX is set at ISO standards for a particular machine but relies on AI with sensors to interpret conditions inherent in the setting around it, including the impact of a nearby passing train, for example. Train track vibration might initially look bad for a given industrial machine, but AI helps determine the detected vibration is still normal for operation in that particular setting.