The engineers at MIT have developed a smart electrical outlet that can learn what appliances are plugged in and tell dangerous electric spikes from spikes that are benign. MIT says that nuisance trips are things like your vacuum turning off unexpectedly or a lamp going off when your AC unit turns on. Those trips often happen when a detector installed inside the wall senses something that could be an arc-fault.
Arc-faults could create sparks that lead to a fire. MIT says that current arc-fault detectors are overly sensitive and shut off the power at an outlet in response to signals that are harmless. The team is working on a solution to eliminate those false positives with a new smart power outlet. The device can analyze electrical current usage from single or multiple outlets and distinguish between benign arcs that pose no risk and those that could cause sparks and lead to a fire.
The device is also able to train itself to identify what might be plugged into the outlet. The design can process electrical current in real-time and features software able to analyze data via a neural network. The neural network, in this case, is focused on determining if a captured signal poses risk or not.
MIT’s smart power outlet is part of the internet of things and would work with an app to share data with users on electricity use. Thanks to the neural networks, the more data that is gathered and shared from various users, the smarter the outlets get until people experience very few of these nuisance trips of their power while using appliances at home.
MIT used a Raspberry Pi Model 3 computer in their design, which is low cost and easy to use. The device uses an inductive current clamp and a USB sound card to read the incoming current data. The device can pick up small amounts of data to analyze and build its database. MIT offers no indication of when or if the technology might be commercialized.