Imagine having a highly qualified predictive maintenance assistant in your factory, permanently available, listening to the machines, and able to warn you of any abnormal behavior before a problem impacts the production line! It is now possible: Bob fulfills this mission and performs a real-time diagnosis of the health of the equipment to be monitored.
The assistant is made up of vibration and temperature sensors which, thanks to its powerful embedded machine learning algorithms, allows it to understand its environment in a completely autonomous way. The sensor builds up his reference knowledge, and then monitor the behavior of the equipment on which it has been installed 24 hours a day.
At regular intervals, Bob sends by LoRaWan radio a report on the activity and health of the monitored equipment, as well as alerts in case of suspected anomalies (present or future). Bob is thus able to recognize the signature of normal or abnormal vibrations, just as Shazam* would do for music. As soon as spots a problem or a drift, Bob immediately transmits an alert, i.e. the result of its analysis, to a supervision console, an IoT platform on the Internet or an operator’s smartphone. Bob does not transmit data, only communicates the result of its analysis. So production data is never sent to the cloud.