Скачать или смотреть AC Power Monitoring and Inference with Edge Impulse and the Particle Photon 2

This demo measures the current and voltage supplied to a power strip with a variety of devices plugged in. An ML model has been trained with Edge Impulse to determine which devices are powered on based on signatures in the current and voltage waveforms, without having to monitor each device independently. A display shows which devices are running. This data is also reflected in the cloud via a dashboard.

Given the data output by the model, we are able to determine whether a device is consuming a typical amount of energy, or is active for a typical amount of time. Deviations from the norm in either case could indicate a failure or dangerous condition that requires attention. This can be used to trigger an alert to a user via SMS or email if tripped.

The demo is intended to show off an approachable demo of Particle's new Photon 2 module and how Edge Impulse can take advantage of the powerful onboard processor. Viewers can easily extrapolate the use case to many different industries and applications, such as a whole-home energy monitoring system or industrial asset monitoring in a factory where monitoring multiple devices at a single point is economical, and additionally monitoring for deviations from typical use patterns may provide value.

Learn more:
https://edgeimpulse.com
https://particle.io