In brief: Taiwanese power management company Delta unveiled some fascinating R&D work that shows Wi-Fi networks may soon be able to monitor our vital signs. At the Computex trade show in Taipei this week, Dr. Tzi-cker Chiueh from Delta's research center demonstrated how their algorithms can detect breathing rates and even heartbeats just by analyzing disruptions in Wi-Fi signals.
It turns out that the channel state in modern Wi-Fi is quite sensitive to tiny movements and environmental changes. By precisely measuring the time of flight and angle of arrival of Wi-Fi signals bouncing around a room, Delta has crafted algorithms that can track breathing with great precision.
As reported by The Register, Chiueh claims their breathing rate estimation algorithm achieves 95% accuracy in measuring breathing rates over 5 meters. Even the tiny movements from a heartbeat can be spotted, with 83% accuracy in calculating pulse rates from up to one meter away. He showed a video of an experiment where the tech could distinguish sleep states for two people by their breathing patterns and body movements using nothing but Wi-Fi waves fired from a pair of smartphones.
It'd be revolutionary if this sort of tech arrived on smartphones to possibly do away with the need for wearables for basic health tracking. Google took a stab at 'contactless' health tracking with its Pixel smartphones, but it was a tad clunky, requiring the user to record a small clip of themselves to measure breathing and heartbeat. Perhaps Wi-Fi may help streamline this process.
However, Dr. Chiueh seems to be casting a wider net with the technology. He suggested Wi-Fi access points could be used to monitor patients in hospitals or elderly residents in care homes, without needing dedicated biomedical sensors or ECG equipment.
Kids and pets being left in cars during sweltering heat is another widespread problem that this technology can potentially detect.
Beyond healthcare and safety, Delta believes their Wi-Fi analysis could optimize network performance in challenging environments like warehouses by detecting physical changes that may be degrading signals. By automatically adjusting access point configurations, they aim to achieve throughput rivaling private 5G deployments.
The keynote also touched on Delta's other forward-looking projects enabled by advances in AI and machine learning. These include automated 3D modeling to help drones inspect infrastructure like bridges without GPS, generating an expanded set of test scenarios for self-driving cars beyond what can be reasonably simulated by hand, and even training robots to navigate environments like using elevators.
As of now, no solid timelines have been provided on when or if these nascent projects might turn into real products, but the demo alone deserves credit.