MIT device measures walking speed with wireless signals to detect health problems

MIT researchers developed the WiGait so it measures walking speed, which can help predict potential health issues, by analyzing wireless signals. It is reportedly more accurate than a wearable such as Fitbit or GPS via a phone. The devices could help develop health-aware smart homes.

WiGait uses wireless signals to continuously measure a person's walking speed.
Jason Dorfman/MIT CSAIL

The way a person walks, or gait recognition, has long been used for biometric purposes, but MIT researchers say that since how fast you walk is supposed to help predict health problems, they’ve come up with a device that can detect walking speed, and thereby potential health issues, without being as invasive as a surveillance camera, a Kinect, or even a wearable. The device is dubbed “WiGait” and it uses wireless signals to monitor walking in a “continuous and unobtrusive” way.

WiGait is a white, picture-sized sensor which can be mounted on the wall. It emits low-power radio signals – emitting about one-hundredth the amount of radiation of a cellphone – and then analyzes how wireless signals reflect off a person’s body. The team from MIT Computer Science and Artificial Intelligence Lab (CSAIL) developed algorithms which can “distinguish walking from other movements, such as cleaning the kitchen or brushing one's teeth.”

You don’t have to be in the same room as the device, as it can measure how wireless signals bounce off bodies through walls as long as the person is within a 29- to 39-foot radius of the device. In fact, MIT researchers say WiGait can use wireless signals to measure the walking speed of multiple people with a 95 to 99 percent accuracy.

Why is the speed of walking important? “Many avoidable hospitalizations are related to issues like falls, congestive heart disease, or chronic obstructive pulmonary disease, which have all been shown to be correlated to gait speed,” said Dina Katabi, MIT Professor of Electrical Engineering and Computer Science. Researchers believe how fast you walk could help predict “cognitive decline, falls and even certain cardiac or pulmonary diseases.”

Since WiGait measures a person’s stride length with an 85 to 99 percent accuracy, it might also help researchers understand conditions such as Parkinson’s disease that is characterized bysmaller step size. While this research was conducted on healthy people, in the future, the researchers hope to use WiGait in tests on people with multiple sclerosis, Alzheimer’s, or other walking impairments.

The researchers claim that wearables such as Fitbit and Jawbone cannot measure stride length and provide a rough estimate of speed based on step count; GPS-enabled smartphones also give inaccurate walking speeds. Physical therapists use a stopwatch to measure walking speed, but that is hardly a full day’s worth of monitoring. Cameras and Kinects installed in homes are intrusive, but WiGait was developed to be more “privacy-minded” – data is anonymized and a person is “nothing more than a moving dot on a screen.”

WiGait, which could be installed in smart homes to monitor health, is described as being capable of measuring walking speed “with a high level of granularity,” without requiring any user interaction; there’s no need for a person to wear or carry a sensor.

MIT PhD student Chen-Yu Hsu, lead author of the research paper, explained, “By using in-home sensors, we can see trends in how walking speed changes over longer periods of time. This can provide insight into whether someone should adjust their health regimen, whether that’s doing physical therapy or altering their medications.”

“Extracting Gait Velocity and Stride Length from Surrounding Radio Signals” (pdf) will be presented this month at the ACM Conference on Human Factors in Computing Systems (CHI 2017).

The research paper concluded:

We believe our results [will] help develop smart homes that are health-aware and can monitor the safety and well-being of their occupants. Also, WiGait enables new interaction capabilities, and can be incorporated into user interfaces that adapt the environment as the user’s health changes, e.g., the environment may encourage the user to exercise more, or alert family and friends for health emergencies.

Copyright © 2017 IDG Communications, Inc.

Bing’s AI chatbot came to work for me. I had to fire it.
Shop Tech Products at Amazon