For any business that relies on machinery, regular maintenance is crucial. These days, machines are monitored by sensors and it’s possible to predict anomalous behavior and machine failure from patterns in the data. Whilst it may be straightforward to start collecting sensor data, transforming and enriching that data to unlock its value can be challenging. Achieving accurate prediction requires a robust data pipeline in place to serve forecasting algorithms and ensure timely alerting when a possible problem is detected.
To help make predictive maintenance more accessible to newcomers, we’ve created a simple reference architecture that simulates a fleet of 3D printers. In this webinar we will show you how to use it, how it works and provide the context on how it is enabled by a streaming data pipeline and time series database. We will also share the full source code so you can customize it for your own use.