Just like disaggregation, also the detection of energy events can be done in the cloud as well as partly at the edge, distributed on the middleware of a NET2GRID IoT device.
NET2GRID has developed pre-trained detection models that recognizes appliances from the power curves in the household based on limited history. In this way appliances with a clearly identifiable patterns can be recognized. A clear example is EV charging and washing machine detection. The NET2GRID platform can be configured to send these appliance detection events to the cloud for further processing in particular use cases.
Other alerts that can be set up in middleware are energy usage alerts. In this case, particular treshholds are set by the user and monitoring by the IoT middleware to send a notification when the threshold is reached.
A few examples are discussed in the Energy Usage notifications section:
- Power level alert. In case a certain power level is reached that is set by the user
- Homestate hyperactive alert. A power level alert for three automatically generated activity levels of the household.
- Consumption threshold alert for consumed energy over a period, e.g. a day, week, month or year.
- Capacity peak alert that triggers when consumed energy for a particular quarter hour exceeds an earlier peak in a month. The capacity peak can be set by the capacitypeaks/config endpoint.
Updated 5 months ago