Edge computing is in some sense the “opposite” of cloud computing in that computations are performed on a device in the field rather than on a central server. The cloud has unparalleled ability to synthesize different data sources and access computing power, but the logistics of transporting and storing data still impose heavy limitations.  A typical setup transmits only one read every 15 minutes back to the cloud. The edge has a 3.6 million-fold data advantage, and when this matters for a particular application or AI model, the relevant computation must be performed on the edge.  

Edge Computing

Edge computing is in some sense the “opposite” of cloud computing in that computations are performed on a device in the field rather than on a central server. The cloud has unparalleled ability to synthesize different data sources and access computing power, but the logistics of transporting and storing data still impose heavy limitations.  A typical setup transmits only one read every 15 minutes back to the cloud. The edge has a 3.6 million-fold data advantage, and when this matters for a particular application or AI model, the relevant computation must be performed on the edge.  

Application: AI-based EV Detection 

As EVs hit the exponential part of their adoption cycle, managing the increasing demands of EV charging is a principal concern of distributors. It is one thing to know that people are buying EVs, but understanding who is charging, when they are charging and how much they are charging allows for better capacity planning and off-peak incentive program management. AI algorithms can extract this information from consumption data, but the sample rate makes an enormous difference. With 15-minute interval reads, the AI algorithm must wait for (the equivalent of) statistical significance and guess. However, with live data on the edge it can produce immediate and accurate identification of individual charging sessions.  

Application: Service Impedance  

Service impedance monitoring is another application that benefits from living on the edge. When a large load comes online inside a household, voltage dips ever so slightly. The Stratus IQ+ profiles these dips and can assess the health of the feeder lines, connections and other service infrastructure upstream from the meter. Problems such as corroding connections, degrading wires and temperature-driven loosening can be spotted before the service fails. This capability hinges on the instantaneous measurement of voltage dips and bumps. 15-minute interval reads simply do not have the time resolution required, making service impedance monitoring a quintessential edge computing application.  

Application: Custom  

The decision to write an application for the cloud or for the edge is very particular to the application. But what if you are a utility and don’t have the ability to write an edge application? Are you simply out of luck? Not anymore. The Stratus IQ+ supports utility-written, 3rd party written or Sensus written applications.  Now customers can customize their edge applications for a specific need by writing it themselves or contracting with a third party or Sensus.  

The Stratus IQ+ packs an upgraded processor that supports powerful edge computing capabilities ranging from AI-based EV detection all the way to custom python scripting. With this powerful meter in hand, you no longer have to choose between cloud and edge: you can simply have the best of both. To learn more, reach out to Sensus to discuss the “art of the possible” with edge computing.