Researchers from the University of California, led by electrical engineering professor Yasamin Mostofi, have been able to determine how many people are in a given space using WiFi devices.
This tool could have a variety of applications ranging from crowd size control, so necessary in times of pandemic, to help with planning in various areas.
All this without spending additional money on video equipment or image recognition systems. And it is that the proposal of the house of studies of Santa Bárbara can work with conventional WiFi routers and antennas.
The WiFi-based solution was presented at the MobiSys International Conference. This shows that it is possible to count the number of people thanks to small involuntary movements.
In testing, a WiFi transmitter and receiver were located on both sides of an area with people seated. Simple changes like changing position, crossing your legs, or coughing produced signal fluctuations.
WiFi as a tool to count people
Credit: University of California
As WiFi is nothing more than electromagnetic radiation, it was “disturbed” by obstacles or movements. Precisely, the analysis of these variations made it possible to determine how many people were in the place.
“This approach uses only the power measurements received from a WiFi link, it does not depend on people charging a device and it works through walls,” explains Professor Mostofi.
To analyze the WiFi data, the researchers developed a complex mathematical model. Interestingly, based on a problem known as “queuing theory”, that is, from an unrelated field.
“Queue theory is a branch of mathematics that studies waiting lines in systems that involve the arrival of customers, who require the service of an entity,” says Mostofi.
The results of the evaluation of the method using WiFi showed encouraging results. In groups with 10 people or less, 96.3% accuracy was achieved without walls and 90% with walls as an obstacle.
Everything seems to indicate that the WiFi technique is promising for crowd counting in real scenarios. The details about the experiments and the mathematical model used can be consulted in this publication of the aforementioned university.
On the other hand, the Wi-Fi standard 802.11bf, planned for 2024, will bring with it WiFi Sensing, a technology that will allow measuring distances and detecting movements without additional equipment. Telefónica has taken the first steps with its Remote Care solution.