Fitness trackers estimate how much a person moves and how fast their heart beats by turning physical signals into measurements. To count steps, they detect patterns of motion that resemble walking or running. To measure heart rate, most use light to detect changes in blood volume beneath the skin. Small sensors collect these signals, and algorithms interpret them to produce the numbers displayed on a screen.
Although these measurements can look precise, they are not all equally direct. A heart rate reading is derived from changes in blood flow, while a step count depends on a device deciding which movements qualify as steps. Understanding how these systems work explains why fitness trackers are useful, why their readings sometimes differ from reality, and what their numbers can—and cannot—tell us about health.
How fitness trackers detect movement
Most wrist-worn fitness trackers and smartwatches use a sensor called an accelerometer to detect movement. An accelerometer measures acceleration: changes in an object’s speed or direction. Because acceleration can occur along three perpendicular axes, a small sensor can detect movement forward and backward, from side to side, and up and down.
The sensor also detects the effect of gravity. Even when a tracker is sitting still, its accelerometer registers a signal related to Earth’s gravitational pull. When the wearer moves an arm, the measured acceleration changes. Walking, running, climbing stairs, and swinging an arm produce different patterns of motion over time.
The challenge is that an accelerometer does not inherently know what a step is. It records physical movement, not activities or intentions. The device must interpret the resulting data to distinguish walking from other actions, such as waving, reaching for an object, or riding in a moving vehicle.
Some trackers also contain a gyroscope, which measures rotational motion. This can help a device understand how it is turning or changing orientation. Depending on the model and activity, combining rotational information with acceleration can improve motion classification. However, an accelerometer alone can provide enough information for many everyday step-counting tasks.
How a fitness tracker turns movement into steps
Walking produces a repeating pattern of body movement. Each step involves changes in the motion of the legs, torso, and arms. When a tracker is worn on the wrist, it can detect some of the acceleration associated with this repeating pattern, particularly when the arm swings naturally.
The tracker samples the accelerometer’s signal many times per second. Software then analyzes the measurements for features associated with steps, such as repeated peaks, changes in acceleration, and rhythmic timing.
Imagine a person walking at a steady pace. Their arm swings back and forth in a roughly regular pattern, creating a sequence of motion signals. A step-detection algorithm looks for recurring patterns that match the expected characteristics of walking. When the pattern meets its criteria, the device registers a step and updates the total.
The software may also use a person’s recent movement to judge whether a new signal is plausible. For example, a single sharp wrist movement might be ignored if it does not fit a walking pattern, while a sequence of regular movements may be counted as steps.
These algorithms differ among manufacturers and models. Some use relatively simple signal-processing rules, while others apply more sophisticated statistical or machine-learning methods to classify movement. In either case, the goal is to separate genuine steps from unrelated motion while avoiding missed steps.
Step counting is therefore an estimate based on recognizable movement patterns, not a direct measurement of how many times a foot touches the ground.
Why step counts are sometimes inaccurate
A fitness tracker can make mistakes because wrist movement is only an indirect indicator of walking. Activities that produce step-like movements may be counted as steps even when the wearer is not walking. Conversely, genuine steps may go undetected when the expected movement pattern is absent.
For example, washing dishes, gesturing during a conversation, or repeatedly moving an arm can create motion signals that resemble steps. A well-designed algorithm filters out many such movements, but it cannot eliminate every false count.
The opposite problem occurs when someone walks with little arm movement. Pushing a shopping cart, holding a stroller, carrying a heavy object, or walking with a hand in a pocket can reduce the wrist’s characteristic swinging motion. The tracker may miss some steps even though the legs are moving normally.
Running generally produces stronger and more regular motion than slow walking, but accuracy still depends on how the device is worn and how its software interprets the signals. Irregular movement, changes in pace, and activities that combine walking with other motions can make detection more difficult.
Device placement matters, too. A wrist-worn tracker observes arm motion, whereas a phone in a pocket or a sensor attached to the waist experiences a different movement pattern. Each placement requires an algorithm suited to the signals it receives. Two devices worn in different locations can therefore report different step totals for the same walk.
A step count is most useful as a consistent measure of activity over time. If the same device is worn in similar circumstances each day, it can reveal whether a person is generally moving more or less, even if the count is not perfectly accurate on every occasion.
How fitness trackers measure heart rate with light
Most modern wrist-worn fitness trackers measure heart rate using a technique called photoplethysmography, commonly abbreviated as PPG. The method uses light to detect changes in blood volume in the tissue beneath the device.
On the underside of a tracker, one or more light-emitting diodes shine light into the skin. Sensors measure how much of that light returns to the device after interacting with the skin and underlying tissue.
Blood absorbs light differently from surrounding tissue. As the heart contracts, it pushes blood into the arteries, changing the volume of blood in the small vessels near the skin. These changes alter the amount of light absorbed and reflected or scattered back toward the sensor.
The resulting signal varies over time. With each heartbeat, the amount of blood in the illuminated tissue changes, producing a repeating pattern in the detected light. The device’s software identifies these pulse-related fluctuations and estimates how frequently they occur.
In simple terms, the tracker shines light into the skin, detects changes associated with each pulse, and calculates the number of pulses per minute. That value is displayed as heart rate, measured in beats per minute, or bpm.
Many devices use green light because it works well for detecting pulse-related changes in blood volume near the skin. Green light is strongly absorbed by blood, which can help the sensor distinguish changes in blood volume from the surrounding tissue signal. Some devices also use red or infrared light, particularly for additional measurements such as estimates of blood oxygen saturation.
The exact arrangement of lights, detectors, and signal-processing methods varies by device. Regardless of the design, the central principle is the same: changes in the detected light provide information about the pulse.
How a tracker calculates beats per minute
The raw optical signal is not a clean sequence of perfectly distinct pulses. It can contain noise from arm movement, changes in sensor contact, ambient light, and other sources. The device must process the signal before estimating heart rate.
First, software identifies the portion of the signal that changes in a pattern consistent with the pulse. It may filter out fluctuations that are too rapid, too slow, or otherwise inconsistent with the expected physiological signal. It then searches for recurring pulse features, such as peaks or rising edges, and estimates the time between them.
If successive pulse peaks are separated by a short interval, the calculated heart rate is higher. If they are farther apart, the calculated rate is lower. For a reasonably regular rhythm, heart rate can be estimated from the average interval between detected pulses.
For example, if the time between successive pulse peaks is approximately one second, the estimated rate is about 60 beats per minute. If the interval is approximately half a second, the estimated rate is about 120 beats per minute.
Real heartbeats are not always perfectly evenly spaced, so trackers generally analyze multiple pulses rather than relying on one interval. They may also smooth the results to prevent the displayed number from jumping around in response to small fluctuations or noise.
This processing introduces a trade-off. More smoothing can make the display steadier, but it may also delay the detection of rapid changes in heart rate. Less smoothing can respond more quickly, but it may produce a noisier reading.
The number on the screen is thus a calculated estimate based on an optical pulse signal, not a direct view of the heart itself.
Why heart rate readings can be wrong
Optical heart-rate measurement works best when the device can maintain good contact with the skin and detect a clear pulse signal. Several factors can interfere with that process.
Movement is a major source of error. During exercise, the wrist and surrounding tissue move while the sensor is trying to detect subtle changes in blood volume. Those movements can create changes in the optical signal that resemble or obscure a pulse. Running, cycling over rough surfaces, lifting weights, and other activities can make accurate detection more difficult, depending on how the wrist moves and how the tracker fits.
Fit and contact matter. A tracker that is too loose may allow the sensor to shift or admit outside light. A very tight band can be uncomfortable and may alter local tissue conditions. A secure, comfortable fit that keeps the sensor in consistent contact with the skin generally helps produce a cleaner signal.
Blood flow near the skin varies. Cold temperatures can cause blood vessels near the skin to narrow, reducing the strength of the pulse signal at the wrist. Other factors, including differences in skin characteristics and circulation, can also affect how much useful light reaches the detector. Devices vary in how well their sensors and algorithms handle these conditions.
The exercise itself can complicate detection. During activities involving repeated wrist flexion, gripping, or vigorous arm movement, motion signals can overlap with the pulse signal. Some algorithms may briefly lock onto a movement pattern instead of the true pulse rhythm, producing a reading that is too high or too low.
These problems explain why a tracker may provide reasonably stable readings during quiet rest but become less reliable during intense or irregular exercise. Accuracy depends not only on the optical sensor but also on the quality of the signal-processing software and the circumstances in which the measurement is taken.
Heart rate is not the same as electrical heart monitoring
A wrist tracker measures a pulse-related change in blood volume. It does not ordinarily measure the heart’s electrical activity directly.
An electrocardiogram, or ECG, records electrical signals generated as the heart’s chambers activate. Clinical ECG equipment uses electrodes to detect these signals and can provide information about heart rhythm that an optical pulse sensor cannot supply on its own.
The distinction matters because electrical activity and the resulting pulse are related but not identical. An electrical heartbeat may sometimes produce a weak pulse that is difficult for an optical sensor to detect. Conversely, movement or changes in blood flow can distort an optical signal without representing an abnormal electrical rhythm.
Some smartwatches include an ECG feature that uses separate electrodes and a different measurement method. This is distinct from the continuous optical heart-rate function found on most wrist trackers. Even when a device offers ECG recording, its capabilities and clinical uses depend on its design and applicable regulatory authorization.
For everyday exercise, optical heart-rate measurements are often useful for observing broad changes, such as an increase during activity and a decrease during recovery. They should not be treated as a complete assessment of heart rhythm or as a substitute for medical evaluation when symptoms or health concerns arise.
How fitness trackers estimate calories burned
Many trackers use step counts and heart-rate measurements as inputs to estimate energy expenditure, or the amount of energy the body uses. They may also consider age, height, weight, sex, activity type, movement intensity, and the duration of an activity.
The underlying challenge is that energy expenditure cannot be measured directly by a step counter or wrist optical sensor. The device must infer it from signals associated with physical effort and from general models of human physiology.
Step counts can indicate how much a person walks, while acceleration patterns can help estimate activity intensity. Heart rate provides another clue because it often rises as the body’s demand for oxygen increases during exercise. Software combines these inputs to estimate how much energy the activity required.
However, heart rate is influenced by more than exercise intensity. Heat, stress, illness, hydration, medications, and individual differences in cardiovascular fitness can all affect the relationship between heart rate and energy use. Two people performing the same activity may have different heart rates and burn different amounts of energy.
Likewise, activities such as cycling, weight training, and carrying heavy loads may require substantial effort without producing a step pattern that fully captures that work. A tracker may recognize some of these activities through other motion signals, but its estimate still depends on the capabilities of its sensors and algorithms.
Calorie figures should therefore be understood as model-based estimates, not precise measurements of energy burned. They can be useful for observing broad patterns, but small differences between daily totals may not represent meaningful differences in actual energy expenditure.
What fitness trackers can reliably tell you
The usefulness of a fitness tracker depends partly on the question being asked. A device may be imperfect at counting every step yet still be helpful for identifying changes in daily movement. Similarly, an optical heart-rate sensor may occasionally produce an incorrect reading during vigorous exercise while still revealing the general pattern of a workout.
Repeated measurements are particularly valuable. When a person wears the same device consistently, trends in step counts can show whether daily activity is increasing or decreasing. Heart-rate records can reveal how the pulse changes during exercise, how it responds to different workloads, and how it returns toward resting levels afterward.
Comparisons are more meaningful when the measurement conditions are similar. A daily step total is easier to interpret when the device is worn in the same location and during comparable activities. Resting heart-rate trends are more useful when readings are taken under similar conditions, because recent exercise, stress, illness, sleep, and other factors can influence the result.
It is also important to distinguish measurement precision from measurement accuracy. A tracker may display a heart rate of 72 bpm or a step total of 8,413, but the number of digits does not establish how close the result is to the true value. A display can be precise in appearance while still containing measurement error.
Fitness trackers are best viewed as tools for monitoring patterns rather than perfect instruments for measuring every aspect of physical activity. Their step counts come from interpreting movement, their heart-rate readings come from interpreting changes in blood flow, and their estimates of calories and other metrics rely on additional calculations. Understanding these distinctions makes the information more useful—and helps keep the numbers in perspective.