Your phone can use tiny skin-colour changes to measure your heart rate

Your smartphone could soon measure your heart rate using nothing more than its front-facing camera—by detecting microscopic skin color changes caused by blood flow. A team at Google, led by researcher Ming-Zo Po, has developed PHRM (Passive Heart Rate Monitoring), a system that uses machine learning to analyze brief video clips captured during normal phone use, like unlocking your device. The technology works by detecting subtle shifts in light absorption in the face—changes too small for the human eye to see but measurable by digital sensors. In a real-world study involving over 160,000 videos from 107 people across 26 phone models, PHRM achieved clinical accuracy within five beats per minute, meeting medical standards for resting heart rate tracking. However, the system struggled more with darker skin tones, showing lower measurement success rates—a gap that highlights urgent equity concerns in AI health tools. While the promise of invisible, universal health monitoring is compelling, the team emphasizes strict privacy safeguards: data would stay on-device, require user consent, and only activate during face authentication. Meanwhile, the episode also revisits a 1970s thought experiment by Richard Feynman on restaurant ordering, revealing a mathematical solution to the 'optimal stopping problem'—a model now validated by a new study involving over 2,000 participants.