AI analyzes radio waves to map your sleeping pattern
Researchers at Massachusetts Institute of Technology (MIT) and Massachusetts General Hospital have developed an artificial intelligence system that's capable of monitoring a person's sleep cycle by analyzing the radio signals surrounding him/her. The new AI engine extracts relevant data from ambient radio waves and translates them into sleep stages such as light, deep, or rapid eye movement (REM). Here's more!
No need of body sensors
In this system, there is no need to have any sensors, like trackers or smart watches, attached to one's body to know if he/she has been sleeping properly or not. MIT professor, Dina Katabi, who led the study said, it is akin to a Wi-Fi router that knows when one is dreaming and can tell if one is getting enough sleep or not.
What's the plan?
Katabi said the vision was to develop health sensors that are discreet in nature, yet capable of capturing physiological signals and critical health metrics. These sensors would not meddle with the users' normal behavior by making him/her wear a device. Radio-based sensors, capable of remotely measuring vital signs and behavior, which act as health indicators, have been explored by researchers for quite some time.
How is it possible?
These sensors carry a wireless device that's equivalent to the size of a laptop and capable of emitting low-power radio frequency (RF) signals. When it reflects off the user's body, even the slightest movement of a body alters the frequency of those reflected waves. Upon analyzing these waves, vital signs, such as pulse and breathing rate, can be revealed.
In simpler terms
Katabi simplified it further by saying, "It's a smart Wi-Fi-like box that sits in the home and analyses these reflections and discovers all of these changes in the body, through a signature that the body leaves on the RF signal." These can be used to substitute monitors such as electroencephalography (EEG) machines that are currently used for monitoring sleep patterns in a lab.
Another leap for AI
Previously, this wouldn't have been possible, as existing algorithms would just get confused with all the data in hand. However, advances in AI now allow training of deep neural networks to extract and analyze significant information from complex datasets. MIT researchers, therefore, have now created a new AI algorithm to eliminate unnecessary information from radio signals and analyze only those, relevant to sleeping patterns.
