Abstract: A system, method and computer program product for operating a low-voltage Internet-of-Things sensor device. The method includes sensing of the temperature dependence at each voltage condition in addition to the actual temperature and voltage. A programmed machine learning model uses the information to decide when it is appropriate to test the device functionality and use the results of different tests to determine when the system should run synchronously or asynchronously through a machine learning predictive algorithm. Based on said one or more sensed operating conditions, the system uses the model to detect a mode of operation of said IoT device indicating IoT device meets an expected level of performance, or a mode indicating said IoT device is not operating according to the expected level of performance. Based on the detected operating condition, the IoT device automatically adapts its operation to ensure a desired level of IoT sensor device performance.

Abstract: A system, method and computer program product for operating a low-voltage Internet-of-Things sensor device. The method includes sensing of the temperature dependence at each voltage condition in addition to the actual temperature and voltage. A programmed machine learning model uses the information to decide when it is appropriate to test the device functionality and use the results of different tests to determine when the system should run synchronously or asynchronously through a machine learning predictive algorithm. Based on said one or more sensed operating conditions, the system uses the model to detect a mode of operation of said IoT device indicating IoT device meets an expected level of performance, or a mode indicating said IoT device is not operating according to the expected level of performance. Based on the detected operating condition, the IoT device automatically adapts its operation to ensure a desired level of IoT sensor device performance.