The outlets currently in your house are more than likely arc fault circuit interrupters (AFCIs). These are those sensors that cut the power when you try to run the microwave and the vacuum cleaner at once. They’re designed to trip an outlet’s circuit when they detect a potentially dangerous spark in a power line, but they’re also hopelessly outdated for today’s power grid and the power needs of today’s Internet of Things-equipped smart homes.
To fix this inadequacy, MIT’s scientists rebuilt a power plug from scratch — one equipped with functions that can be boosted with artificial intelligence. A Raspberry Pi microcomputer is driving the technology, running a machine learning algorithm that can learn more about electrical patterns the more data it absorbs. An inductive power clamp monitors the current from the adjacent magnetic field, while a simple USB sound card — designed to detect small signals at high data rates — reads the current data.
“We create fingerprints of current data, and we’re labeling them as good or bad, or what individual device they are,” researcher Josh Siegel told MIT News. “There are the good fingerprints, and then the fingerprints of the things that burn your house down. Our job in the near term is to figure out what’s going to burn down your house and what won’t, and in the long term, figure out exactly what’s plugged in where.”
Siegel’s team tested the data from a fan, an iMac, a stovetop burner, and an air purifier. As the algorithm collected data from the diverse power needs and functions, it was able to identify potentially deadly power arcs with 99.95 percent accuracy, a much higher rate than existing AFCIs. Future versions could integrate networking functions to wireless report on energy usage throughout an IoT aquarium, allowing the system to learn power patterns and adjust usage accordingly.
“By making IoT capable of learning, you’re able to constantly update the system, so that your vacuum cleaner may trigger the circuit breaker once or twice the first week, but it’ll get smarter over time,” Siegel says. “By the time that you have 1,000 or 10,000 users contributing to the model, very few people will experience these nuisance trips because there’s so much data aggregated from so many different houses.”
Siegel is an award-winning inventor and parallel entrepreneur whose research has been awarded the Lemelson-MIT Student Prize and the MassIT Government Innovation Prize. Siegel’s co-authors on these results are Shane Pratt, Yongbin Sun, and Sanjay Sarma, the Fred Fort Flowers and Daniel Fort Flowers Professor of Mechanical Engineering and vice president of open learning at MIT.
“This is all shifting intelligence to the edge, as opposed to on a server or a data center or a desktop computer,” Siegel said. “I think the larger goal is to have everything connected, all of the time, for a smarter, more interconnected world. That’s the vision I want to see.”