The key idea behind predictive maintenance is to predict the unexpected failures of the equipment and to take action early. Unexpected failures result in unplanned downtime which could last a long time; especially for some key equipment, resulting in a loss of revenue. As the name implies, predictive maintenance is achieved through continuous monitoring of the equipment for any potential signs of failure and issuing a server alert. The monitoring itself is achieved using the low-power digital accelerometers that come with embedded advanced signal filtering capabilities that detect early signs of equipment failures. Since many of the monitoring devices are battery-operated, ROHM offers highly efficient voltage regulators and discrete components that help to extend the life of battery-powered monitors themselves, thus further reducing the operational cost and downtime due to maintenance. When a fault is detected, an alert is issued to the host microprocessor. By monitoring faults autonomously, ROHM’s accelerometers save power and computational resources of the main system’s processors. The job of the main microprocessor is to process the alert from an accelerometer and issue a server alert. In the case of a battery-operated device, the alert is issued wirelessly. ROHM’s low-energy wireless radios (Wi-SUN and Bluetooth) allow transmission of alerts to the server quickly and reliably.