A sensor package includes nine different types of sensors that monitor potentially dozens of activities involved in the Internet of Things. Source: Carnegie Mellon UniversityA plug-in sensor package has been developed by researchers at Carnegie Mellon University that monitors multiple phenomena in a room without the need for smart appliances.
The package can monitor activities such as sounds, vibration, light, heat, electromagnetic noise and temperature. Built with machine learning, the sensor suite can determine if a faucet’s left or right spigot is running, if the microwave door is open, how many paper towels have been dispensed or if the oven has been left on.
“The idea is you can plug this in and immediately turn a room into a smart environment,” says Gierad Laput, a Ph.D. student in CMU's Human-Computer Interaction Institute (HCII). “You don't have to go out and buy expensive smart appliances, which probably can't talk to each other anyway, or attach sensors to everything you want to monitor, which can be both hard to maintain and ugly. You just plug it in to an outlet.”
The suite includes nine sensors that are found commonly available in other smart home devices with the one exception being a camera for privacy concerns. Machine learning allows the sensors’ raw feeds to be combined to identify a wide range of events and objects such as distinguishing between a blender, coffee grinder and mixer based on sounds and vibrations. Even writing or erasing a whiteboard can be detected, researchers say.
The sensor suite can also track the state of a device to determine if cooking is interrupted or if a microwave has completed its cooking cycle. A faucet left running for a long time where no person is in the room might prompt the suite to send a warning message to a user’s smartphone.
“It can not only tell you if a towel dispenser is working, but it can also keep track of how many towels have been dispensed and even order a replacement roll when necessary,” Laput says.
CMU says human activity can be sensed such as when someone is sleeping, in the shower, streaming a video or has left the house. Because the processing on occurs mostly on the sensor platform itself, detailed or sensitive data is not transmitted or recorded.
Pre-training of the sensors is possible to cycle water heaters or air conditioning units or even to detect devices and brands of home or office products, allowing the sensor platform to detect these devices as soon as it is plugged in.
Learn more about the project by visiting the Iot Expedition, a partnership with Google and universities.
