Gerad Laput, a researcher from the Carnegie Mellon University, introduced a universal device capable of tracking various events in a room and turning an ordinary house into a smart one without fitting any sensors to existing devices.
The device, called the Synthetic Sensor, is a small board in a transparent case. On the board, there are a controller and several sensors that detect various things, like temperature, color, illumination, magnetic field strength and atmospheric pressure. A regular USB-adapter is used for powering the device.
When switched on, the gadget starts tracking events, occurring in the room, and transmitting data about them to a computer with special software installed on it. The software uses machine learning and recognizes events based on the readings from Synthetic Sensor.
The program can determine whether a person is in the room and what kind of event is happening around. For example, sensors can detect if water is flowing from the faucet, if gas is on, if the blender or other device is running. More thorough data research can determine the operating time of different appliances, water flow rate, and so on.
In a long-term perspective, data from several sensors processed by the program can be used to run a smart house.