WiFi Motion Sensor Uses WiFi to Detect Movement

September 29. 2018. 7 mins read

It’s a question that has plagued humanity for nearly 80 years since the first Marvel superhero debuted way back in 1939: What superpower would you want to have? The emergence of certain technologies has made that question easier to answer: Exoskeletons can give us super strength, while brain-computer interfaces enable us to move isht with our minds. X-ray vision would be one of those cool, minor superpowers if you had a full suite of god-like abilities like Superman. It turns out we ordinary humans can also have X-ray vision by turning the wireless signals that allow us to surf for porn art on Etsy into a WiFi motion sensor.

WiFi Motion Sensor Research

For at least the last five years, researchers at MIT have been working diligently on a system to track human movement, from even behind walls, using WiFi technology rather than something like radar or sonar. Dubbed Wi-Vi (see video here), the original system employed two transmitter antennas and a single receiver. The signals from the transmitters are set up so that they interfere with each other and cancel each other out. Static objects, such as a wall, create identical reflections from the transmitters, so they too are cancelled out. Only reflections that change between the two signals, such as the motion of Superman leaping a tall building, would arrive back at the receiver.

More recently, the same big-brained scientists are tweaking their WiFi motion sensor technology to detect walking speed, as demonstrated in the video above. That can’t be good news for holders of Fitbit (FIT) stock who have seen the wearable tech company flatline since its IPO more than three years ago. In theory, the WiFi signal is at least as highly accurate as GPS, without the need to geolocate somebody from outer space. The scientists found that they can analyze the signals reflected off bodies to measure everything from breathing to specific emotions.

Even more recently, we came across some research out of the University of California, Santa Barbara, that shows how WiFi signals can be harnessed to estimate the number of people inside a room from outside. You can imagine the future possibilities: One day you could use an app to find out whether the scene inside your favorite hipster dive bar is dead or just mostly dead on any given night. In this case, the WiFi motion sensor uses only one transmitter and one receiver, which measures the received signal strength from the transmitter to give a guesstimate on how many people are behind the green door.

The Other Kind of WiFi Motion Sensor

One can start to imagine the commercial applications of this kind of technology, such as smart energy management (no reason to heat the basement if there’s just one guy down there babysitting his stapler) or even as a security system sans for the need of any additional hardware. For example, last year we covered 10 home security startups that relied on the more traditional use of WiFi, which is as the road network that enables us to drive various smart devices in our homes or offices through one of those talking heads, whether it’s Alexa or some other smart voice assistant in the emerging IoT universe. We highlighted companies like Denver-based Notion, which has raised $16 million to date, for motion sensors that can be used on objects in the home so you know when the kids are getting into your jar of special Girl Scout Cookies.

The Notion Home Security System
The Notion Home Security System

But back to the more technologically innovative use of WiFi: Tracking people for profit. We’ve found a few startups that are leveraging the work of academia, as well as the MBAs, computer nerds and engineer-types in their own companies, to commercialize these emerging discoveries. Not surprisingly, the applications skew toward security and the smart home of almost-tomorrow.

Turning the Home into a WiFi Motion Sensor for Security

Let’s first talk about security. Founded in 2014, Silicon Valley-based Radiomaze uses WiFi equipment like routers and repeaters, as well as WiFi-enabled devices, as the “nodes” in a WiFi motion sensor detection system. It spies movement based on the way WiFi radio waves behave as they bounce around us. That’s because, as the company says, “human beings are just electromagnetically large bags of water,” so that the system detects when a person walks by an area covered by domestic WiFi, triggering a notification.

Dashboard for Radiomaze WiFi motion sensor detection system.
Dashboard for Radiomaze WiFi motion sensor detection system. Credit: Radiomaze

The latest version of the Radiomaze motion detection system allows users to customize notifications, such as “away mode” when no one is home, alerting the laser-armed sharks to open fire on intruders. Then there’s “home mode,” which alerts you to when the kids sneak back into the house late at night. The whole idea is to ditch the hardware, including the creepy hidden webcams.

WiFi Motion Sensor Powered by Artificial Intelligence

Click for company websiteFounded in 2015, Montreal-based Aerial has raised $4.3 million, including $1.9 million in a Seed round in February of this year. Like Radiomaze, Aerial uses existing WiFi architecture to create a WiFi motion sensor detection system. However, the startup emphasizes its use of cloud-based machine learning to process the distortions in the WiFi signals to “add context and meaning to motion, allowing the recognition of presence, motion, activity and identity.” The company claims it can get to know a home’s occupants – even pets – based on factors like mass and gait.

The Aerial Wi-Fi motion sensor system learn to identify occupants through machine learning based on radio signal behavior. It might also tell you that you're fat.
The Aerial Wi-Fi motion sensor system learns to identify occupants through machine learning based on radio signal behavior. It might also tell you that you’re fat. Credit: Aerial

This apparently allows Aerial to offer solutions outside of security, such as in-home elderly care without the need for a live-in caregiver or even a robot companion. For instance, a user could customize alerts for an elderly parent living at home if he stays in bed for more than three hours at a time or stays in the bathroom for more than 15 minutes (unless he brings in the Wall Street Journal, of course). Aerial has reportedly tested Aerialytix (the name for its AI-powered motion detection system using WiFi signals) over the course of seven simulated years of operation without one false positive. It recently announced a partnership with Quantenna Communication (QTNA), which makes hardware for WiFi systems, to market Aerial’s WiFi motion detection system to internet service providers and their customers.

Click for company websiteAnother Canadian startup, Cognitive Systems was founded in 2014. Last year, it launched its Aura Home Security system. The system also leverages radio waves from a home’s WiFi network to detect motion, reputedly using AI to process and interpret the signals. However, it does rely on additional hardware, consisting of one hub and at least one beacon, depending on the coverage area. The starter kit retails for $199 and covers up to 700 square feet with one beacon. And like Aerialytix, Aura is continuously learning to recognize the source of motion, but it also comes with an app that syncs with a recognized smartphone, so that you know exactly who is arriving or leaving, creating a geofence around the home.

A sensor plugged into the wall socket is part of the system that captures radio signals from a home WiFi network.
A sensor plugged into the wall socket is part of the system that captures radio signals from a home WiFi network. Credit: Cognitive Systems.

Aura isn’t just about physical home security: Its custom-made chipset also provides network security for a smart home’s IoT devices. After this article was published, Cognitive Systems reached out to let us know the following additional information about how their business model is evolving:

Through collaborations with top Wi-Fi Chipset manufacturers like Qualcomm, Marvell and Cypress, Cognitive’s Aura WiFi Motion stack is licensed for use in smart home ecosystems products such as mesh routers. The Aura WiFi Motion stack is built to work with any 802.11ac platform and can be pushed as an over-the-air upgrade; no need for additional hardware or changes to existing designs. The Cognitive Systems team has developed its own machine learning and predictive analytics algorithms to robustly identify and localize motion and to provide contextual awareness.

WiFi Motion Sensor to Detect Breathing

Finally, Origin Wireless out of Japan, also uses AI for its WiFi motion sensor technology, which it calls the Time Reversal Machine (TRM), which we assume comes from their fascination with Hot Tub Time Machine the movie. The system works on the Qualcomm Mesh Networking Platform, with one router serving as the transmitter and any others in the network acting as receivers. Origin maintains that TRM offers indoor tracking accuracy of less than one inch. The technology is so precise, according to Origin, that it can even detect a person breathing. In fact, Origin is marketing the TRM for vital signs detection and monitoring, in addition to home and office security.

Update 02/22/2021: Origin Wireless has raised $14 million in funding to commercialize its WiFi sensing products. This brings the company’s total funding to $32.3 million to date.


The ability to detect and interpret WiFi radio signals in order to detect an intruder or whether your grandfather has taken a nasty fall in the house seems like an amazing breakthrough. The lack of additional hardware (with one exception), particularly sensors or cameras that pose privacy concerns, makes these solutions particularly intriguing, especially as the smart home of the future grows ever more complex. We’re not surprised that most of these systems employ AI or some sophisticated algorithms to make sense of the signals, as we’ve said before that artificial intelligence is the new electricity, and every technological device will soon be smarter than the average bear.

WiFi motion sensor technology seems to be pretty niche at this point, as we’re just now seeing commercialization from companies like Cognitive Systems. It’s definitely one of those technologies that will be in the early-adopter phase for a while. However, we wouldn’t be surprised if a company like Amazon, which reports say may be getting into the networking hardware business, would be interested in snatching up one of these startups to keep Alexa company.


Leave a Reply

Your email address will not be published.