Ever flipped your smartwatch over and noticed those little green lights blinking against your skin? That's not decoration. That's your watch literally interrogating your blood with photons — dozens of times every second.

Here's what's actually going on under that sensor panel.

Heart Rate: Catching the Pulse with Light

The core technology behind heart rate monitoring is called Photoplethysmography — or PPG, because nobody wants to say that word twice.

The concept is beautifully simple. Your heart beats. With every beat, a small wave of blood surges through your arteries and the tiny capillaries in your wrist. That surge changes the volume of blood sitting right beneath your skin. Your smartwatch exploits this in a clever way.

Here's the trick: hemoglobin — the protein in red blood cells that carries oxygen — absorbs green light particularly well. So your watch fires green LED light into your skin, and a photodiode sensor right next to it measures how much light bounces back.

More blood flow (during a heartbeat) → more light absorbed → less light returns to the sensor.
Less blood flow (between beats) → less absorption → more light returns.

This creates a pulsing wave signal. Each "dip" in reflected light corresponds to one heartbeat. The watch counts these dips over time and gives you a beats-per-minute (BPM) number.

It sounds straightforward, but the engineering challenge is enormous. Your wrist is a terrible place to measure pulse compared to, say, your chest or fingertip. The signal is weak. Motion artifacts from walking, typing, or scratching your nose create noise that can dwarf the actual pulse signal. Modern smartwatches use accelerometers to detect motion and then apply algorithms — often driven by machine learning — to subtract movement noise from the optical data. Some watches run multiple LEDs at different positions and wavelengths simultaneously, cross-referencing signals to filter out junk data.

That's why your heart rate reading might lag a few seconds during intense exercise. The watch isn't slow — it's thinking, trying to separate your actual heartbeat from the chaos of your flailing arms.

Blood Oxygen (SpO2): Now Add Red Light

If green light tells you how fast your heart beats, red and infrared light tell you how well your blood is doing its job.

Blood oxygen monitoring — or SpO2 measurement — relies on one key biological fact: oxygenated and deoxygenated hemoglobin absorb light differently.

Oxygenated hemoglobin (HbO₂) absorbs more infrared light and lets more red light pass through.
Deoxygenated hemoglobin (Hb) absorbs more red light and lets more infrared light pass through.

Your smartwatch shines both red (~660 nm wavelength) and infrared (~940 nm) LEDs into your skin. By measuring the ratio of absorption between these two wavelengths, the watch calculates what percentage of your hemoglobin is carrying oxygen.

A healthy reading sits between 95% and 100%. Below 90% is clinically concerning — though your smartwatch will gently remind you that it's "not a medical device" in the fine print you never read.

This is the same core principle used by the pulse oximeter your doctor clips onto your finger. The difference? A fingertip clip measures light passing through your finger (transmission), while your watch measures light bouncing back from your wrist (reflectance). Reflectance-based SpO2 is inherently noisier and less accurate, which is why most watches ask you to stay still during a reading and why they won't replace hospital-grade equipment anytime soon.

Stress: Where Biology Meets Math

This one's the most fascinating — and the most indirect.

Your watch can't measure cortisol. It can't scan your brain. It doesn't know you just read a terrible email from your boss. So how does it claim to know you're stressed?

Heart Rate Variability (HRV).

Here's something counterintuitive: a healthy heart does not beat like a metronome. The interval between beats naturally fluctuates — maybe 0.82 seconds between one pair of beats, then 0.79 seconds, then 0.85 seconds. This variation is called HRV, and it's controlled by your autonomic nervous system (ANS), which has two branches:

Sympathetic nervous system ("fight or flight") — speeds up heart rate, reduces variability. Your heart locks into a more rigid rhythm, ready for action.
Parasympathetic nervous system ("rest and digest") — slows heart rate, increases variability. A relaxed body allows the heart more freedom to fluctuate.

When you're stressed — physically or mentally — sympathetic activity dominates, and your HRV drops. When you're calm, parasympathetic activity takes over, and HRV rises.

Your smartwatch continuously tracks the tiny time gaps between heartbeats using the same PPG sensors described above, then performs time-domain and frequency-domain analysis on the data. Common metrics include:

RMSSD (Root Mean Square of Successive Differences) — a statistical measure of beat-to-beat variation.
LF/HF ratio (Low Frequency to High Frequency power) — a frequency-domain metric where higher ratios suggest sympathetic dominance (more stress).

These numbers are fed into proprietary algorithms — Garmin calls theirs "Body Battery," Samsung uses a stress score from 0 to 100, Apple folds it into general health insights — that translate raw HRV data into something a human can glance at and understand.

Some watches also supplement HRV with electrodermal activity (EDA) sensors, which measure tiny changes in sweat gland activity on your skin. Stress triggers micro-sweating — even when you don't feel sweaty — which alters your skin's electrical conductance. The Fitbit Sense and some Garmin models use this as an additional stress signal.

The Limits (Because Honesty Matters)

It's worth noting what your watch can't do well:

Tattoos, darker skin tones, and poor watch fit can interfere with optical sensors by affecting how light penetrates and reflects.
Cold weather constricts blood vessels in your wrist, reducing signal quality.
SpO2 readings during sleep can be noisy and inconsistent.
Stress scores are probabilistic, not diagnostic. A high stress reading might just mean you sprinted up the stairs, not that you're having an existential crisis.

These devices are screening tools and trend trackers, not clinical instruments. But they're getting better — fast.

The Bigger Picture

What makes all this remarkable isn't any single sensor. It's the fusion. A few LEDs, a photodiode, an accelerometer, maybe an EDA sensor — all sampling data constantly, feeding it into layered algorithms that cross-reference motion, light absorption ratios, beat-to-beat timing, and skin conductance to construct a surprisingly coherent picture of what's happening inside your body.

You're essentially wearing a tiny, underfunded hospital on your wrist. And every year, it gets a little smarter.

So next time you see that faint green glow on your wrist at 2 AM, know this: your watch isn't sleeping either. It's counting photons, doing math, and quietly keeping tabs on the one muscle you can't afford to ignore.