For most of us, WiFi means Netflix, email, and a dozen smart gadgets clamoring for bandwidth. But engineers at UC Santa Cruz have shown that the same invisible waves could also pick up something far more intimate: your heartbeat.

Their system, dubbed Pulse-Fi, uses low-cost WiFi chips and a clever dose of machine learning to measure heart rate with clinical accuracy, no smartwatch or chest strap required.

The work, presented at the 2025 IEEE International Conference on Distributed Computing in Smart Systems and the Internet of Things, turns a household technology into a kind of invisible stethoscope. Heart rate is a basic but powerful metric of health, tied to stress, activity, hydration, and illness. The idea that your router could quietly track it in the background sounds futuristic. It also raises the question: what happens when everyday infrastructure doubles as medical equipment?

As Ph.D. student and lead author Nayan Bhatia explained:

“The signal is very sensitive to the environment, so we have to select the right filters to remove all the unnecessary noise,” Bhatia said.

WiFi devices constantly send out radio waves, which shift ever so slightly when they encounter movement. Pulse-Fi trains a neural network to detect the faintest variations caused by a beating heart, while ignoring the distractions of someone shifting in a chair or pacing the room. To build their dataset, the researchers set up equipment in UC Santa Cruz’s Science and Engineering library, pairing WiFi readings with the ground-truth of a fingertip oximeter. The resulting ESP-HR-CSI dataset is, as far as they know, the first of its kind.

The team tested Pulse-Fi on 118 participants across 17 different positions: sitting, standing, lying down, even walking. After just five seconds of monitoring, the system delivered heart rate readings within half a beat per minute of error. Longer measurements were even sharper. That level of accuracy is comparable to hospital-grade tools, though here it came from chips costing less than a fast-food lunch.

“What we found was that because of the machine learning model, that distance apart basically had no effect on performance,” said visiting researcher Pranay Kocheta. “The other thing was position—all the different things you encounter in day to day life, we wanted to make sure we were robust to however a person is living.”

The researchers were deliberate about testing with ultra-cheap ESP32 modules (about $5-10 each) and slightly pricier Raspberry Pi boards ($30), showing that accuracy didn’t depend on premium gear. In fact, commercial-grade routers would likely improve performance even further. They also showed Pulse-Fi worked at distances up to three meters, with unpublished data hinting at even longer ranges.

The potential applications are obvious. Non-intrusive monitoring could be a boon for elder care, at-home sleep studies, or health tracking in places where medical infrastructure is thin. And yes, there are commercial implications too: your next WiFi router might come bundled with a pulse sensor by default. Whether consumers welcome that dual use is another matter. We’ve already accepted that our smart speakers listen; are we ready for routers that watch our physiology?

For now, the group is extending Pulse-Fi to track breathing rates and even flag sleep apnea. They’ve filed no consumer product yet, but the contact information for UC Santa Cruz’s tech transfer office is right there in the paper. It’s not hard to imagine venture capitalists already sniffing around.

And tucked inside the author list is a detail that speaks volumes about where innovation comes from: one of the lead contributors, Pranay Kocheta, is still in high school. He joined the project as a visiting researcher, proving that breakthroughs don’t always wait for Ph.D. defenses. Sometimes they happen in a university library, with a $5 chip, and the nerve to ask whether WiFi can feel the same heartbeat we do.

WiFi heart-rate monitoring works because radio waves are slightly altered when they pass through the body. Each heartbeat changes the chest’s shape and blood flow, creating tiny disturbances in those waves. Pulse-Fi’s algorithm filters out background noise and learns to spot those regular fluctuations. The beauty is that no wearable or direct contact is needed: a router and receiver can capture the data through the air, turning ordinary connectivity hardware into a health sensor.

Published in: 2025 IEEE International Conference on Distributed Computing in Smart Systems and the Internet of Things (DCOSS-IoT). DOI: 10.1109/DCOSS-IoT65416.2025.00037