Industrial Internet Application Development

Machine learning and the IoT

Most of the use cases for the IoT today are centered around predictive maintenance and better servicing of digital and physical assets. This requires connectivity of these assets through sensors and simple telemetry, along with threshold-based messaging and alerting. As the use cases become more complex and advanced, a higher level of precision and predictability is required from the data that is captured. An example would be the use of clustering algorithms to group devices and assets into healthy or defective, based on the data they are emitting. If the vibrations coming from windmills or power generators fall in a certain range along with other parameters, they can be classified as defective or requiring servicing. ...