This article is part of our exclusive IEEE Journal Watch series in partnership with IEEE Xplore.
In a recent study, researchers have unveiled a new “smart floor tile” that passively monitors the number of pedestrians and their traffic flow, using no external power source. Instead, the tile is powered by the mechanical energy of people stepping on it. The advance could help make it easy and cheap to monitor crowds for public safety.
“Public safety accidents associated with large-scale pedestrian gatherings occur periodically across the globe, particularly in metropolises,” says Junrui Liang, an associate professor at the School of Information Science and Technology at ShanghaiTech University in China. He points to the COVID-19 pandemic—a time when it was particularly useful to understand pedestrian traffic flow and how close individuals were to one another in real-time.
Liang notes that while cameras can help monitor crowds, these involve privacy concerns (such as being recorded in public spaces without consent) and high costs to deploy. “Therefore, we conceived some alternative solutions for pedestrian traffic flow monitoring, including the battery-free floor tile,” he says. The researchers published their results in November in IEEE Sensors Letters.
Smart Floor Tile Technology Explained
The smart floor tile designed by Liang and his colleagues is just 15 millimeters thick, and contains quasi-static-toggling (QST) electromagnetic harvesters. The harvesters convert the mechanical energy from a person’s footstep into electrical energy by using the person’s footstep to trigger a special spring-like mechanism. When the mechanism is compressed or released beyond a 1-millimeter threshold, it creates a magnetic flux through a coil, inducing a sharp voltage pulse. That electrical energy is then used to transmit data—which includes a timestamp and information on the direction of the footstep—via Bluetooth low-energy transmission to a cloud server for real-time monitoring.
Specifically, the smart tile is able to determine the direction in which a person is traveling thanks to the way in which we step—the heel strikes the tile first, followed by the ball of the foot and toes. The timestamped data from different QST harvesters in a single tile can therefore be used to determine a person’s direction by the orientation of the different parts of the footstep hitting the tile.
In controlled experiments with a single person walking slowly or quickly over the smart tile, the tile could accurately determine the direction of the traveler with a success rate of 97.4 percent. “Even during running, the success rate remained at 92.6 percent,” Liang says.
But how does the tile perform in a more realistic scenario with a crowd?

The researchers placed their smart tile at the entrance of their own lab and used video recordings to verify how accurately the tile monitored traffic flow through the entrance over the course of 120 consecutive hours, or five days. In the field—or office, so to speak—the tile still achieved a 94.8 percent overall success rate in counting people and determining their direction. In total, it achieved 714 correct counts, 35 missed counts, and 2 opposite counts, where the wrong direction was recorded.
Most failures occurred during high-traffic periods, for example when there were fast consecutive footsteps, when someone stepped near the tile’s edges, or when a footstep didn’t completely trigger the tile’s four sensor modules.
Liang says he is surprised that the tile achieves such high accuracy across different walking speeds and foot positions, especially given the system’s extreme simplicity and battery-free design. He attributes the success in large part to the sensitivity of the QST energy harvesters, as well as the model the researchers used to process data, which is adept at filtering out false positives.
The current design of the tile offers many benefits, Liang adds, noting that it is low-cost, easy-to-deploy, and has full energy autonomy. However, it currently cannot count pedestrians who are walking side-by-side—only those traveling in single file. As well, the Bluetooth transmission range is limited to 10 meters, requiring that the data receiver be placed within close proximity of the tile.
The research team has launched an IoT startup, called ViPSN Ltd., and plans are underway to scale up and commercialize their smart tiles. In the long-term, they plan on expanding the tiles’ applications beyond pedestrian monitoring, such as security and virtual reality.
- How to Run a Successful Mobile Crowdsourcing Project ›
- AI Tool for COVID Monitoring Offers Solution for Urban Congestion ›
- Crowdsourcing Smartphone GPS Could Improve Earthquake Warnings ›
Michelle Hampson is a freelance writer based in Halifax. She frequently contributes to Spectrum's Journal Watch coverage, which highlights newsworthy studies published in IEEE journals.



