Principles of operation and calculation of steps in Xiaomi fitness bracelets

Many users of activity trackers often wonder about the reliability of the displayed data, especially when the numbers on the screen are at odds with their personal feelings. Understanding how the microcontroller processes signals from sensors helps to better interpret the results and adjust their workouts. At the heart of all devices in the Mi Band line, a complex set of hardware and software solutions that constantly analyze the movement of your body.

Inside the compact case, there's a miniature accelerometer, which is the main tool for tracking movement in three-dimensional space, and it's this sensor that detects the slightest changes in acceleration and transmits raw data to an embedded algorithm that distinguishes step from a simple hand-wake. Modern models like the Mi Smart Band 7 or 8 use advanced versions of these sensors, which greatly improves the accuracy of measurements.

Importantly, the software of the device plays no less a role than hardware, because it filters noise and false positives. Machine learning algorithms embedded in the firmware allow the gadget to adapt to the individual gait of the owner, minimizing errors when walking or running. Let's take a detailed look at every aspect of this technology so that you can make the most of your gadget.

The role of accelerometer in motion detection

At the heart of the activity tracking system is a three-axis accelerometer that responds to any change in the speed of the device. When you take a step, your body's center of gravity shifts, and the wrist bracelet experiences a characteristic acceleration and deceleration, which is converted into electrical signals that are then analyzed by the processor to match the step pattern.

The key here is the amplitude of the vibrations and the frequency of the vibrations, which must be timed to be counted as a step. If you just wave your hand holding an object or type on a keyboard, the pattern of acceleration will be different from rhythmic walking, and a smart algorithm will ignore these movements. However, with very sharp or specific hand movements, the system can mistake them for steps.

⚠️ Warning: Wearing the bracelet too loosely on the wrist can lead to an increase in the amplitude of the oscillations, which sometimes causes false reading of steps during normal gestures.

Modern sensors in Xiaomi devices have high sensitivity, allowing you to capture even very slow walking. However, there is the concept of a “dead zone” or sensitivity threshold, below which movements are not recorded to avoid constant growth of the meter from hand trembling in transport. This is the balance between sensitivity and noise protection, which engineers adjust for each model separately.

Software Algorithms and Noise Filtering

The raw data from the accelerometer itself says little unless it is processed using complex mathematical models. The embedded software uses digital filters that cut off high-frequency vibrations (like tapping your fingers on the body) and low-frequency drifts. Only after this initial cleaning does the signal arrive at the pattern analyzer.

The algorithm looks for repeating acceleration cycles that are characteristic of a person’s step, given that the hand moves differently from the legs when walking. In more advanced models such as the Mi Band 6 and later, machine learning is used that analyzes thousands of hours of data about users’ movements, which allows the system to “understand” the difference between climbing stairs, jogging and walking on a flat surface.

Particular attention is paid to the situation when a person walks holding a handrail in a transport or pushing a wheelchair, in which case the hand can remain relatively still, and the accelerometer will not detect the characteristic bursts. In such cases, the gyroscope (if it is in the model) or analysis of overall activity helps to correct the data, although it is difficult to completely avoid underestimating the result in such scenarios.

Effects of the gyroscope and additional sensors

Unlike the simple pedometers of the past, modern Xiaomi fitness bracelets often feature not only an accelerometer but also a gyroscope.This sensor measures the device's angular rotation speed in space, allowing you to determine the orientation of the wrist. The combination of data from the accelerometer and gyroscope gives a much more accurate picture of movements, especially during difficult exercises or running.

Having a gyroscope allows the algorithm to know if you flipped your wrist to look at the screen, or just waved your hand when you were talking, and this is critical for the proper function of the wrist lift function and for more accurate counting of steps in non-standard body positions. Without the gyroscope, the system would rely solely on the acceleration vector, which would increase the error.

• 📍 Accuracy of orientation: The gyroscope helps to determine which plane the movement is taking place, distinguishing the step from the rotation of the hand.
• 🏃 Running analysis: Joint operation of sensors allows you to more accurately assess cadence (step frequency) and step length when running.
• 💤 Sleep Tracking: Micro-movement and arm position data help sleep algorithms determine restful rest phases.

It’s worth noting that integrating data from multiple sensors requires more computing resources, but modern Nordic or Dialog chips used in wristbands can easily do this without significantly affecting autonomy.

Synchronization and processing of data in the application

Once the bracelet has collected and processed data, it is transmitted to the smartphone via a Bluetooth connection. Mi Fitness (or Zepp Life for older models) acts as the second level of analytics, where the final calibration and preservation of the history takes place, and this is where the user sees the final figures, which may be slightly different from those displayed on the screen of the gadget in real time.

When synchronized, the app can adjust data based on your profile information: height, weight, step length and age, which affect the calculation of distance traveled and calories consumed, although the step counter itself usually remains unchanged after writing on the device. Cloud algorithms can also make corrections if they detect anomalies in the data.

It is important to understand that late synchronization does not mean data loss, since the wristband memory can store statistics for several days without connecting to the phone. However, it is recommended to synchronize regularly to correctly display daily rhythms and timely notifications of achievements.

Factors affecting the accuracy of the calculation

Despite the high technology, there are a number of factors that can affect the objectivity of readings, and understanding these nuances will help you avoid disappointment and correctly assess your activity, a margin of error of 5-10% is considered normal for consumer electronics in this class.

One of the main factors is where the device is worn: a bracelet worn too high or low on the wrist can give different readings due to changes in the amplitude of the hand movement, and the type of activity affects accuracy: when riding a bicycle or working at a computer, the meter may behave differently than when walking.

⚠️ Warning: Vibration from tools (drill, jackhammer) or shaking in long-distance transport may be misinterpreted by the accelerometer as active walking.

Also worth mentioning is the individual gait, which means that people with very short or very wide strides may encounter small discrepancies because the algorithm is set to average values, and in these cases, regular use and synchronization help the system to get used to your unique style of movement.

Calibration and accuracy of measurements

While the automatic calibration in Xiaomi devices works well, the user can independently influence the accuracy of the measurements, especially when filling out the profile correctly in the application, indicating the real height and weight allows algorithms to more accurately calculate the length of the step and energy consumption.

For Android users, there is a possibility of more deep customization through the system services Google Fit, which can act as an intermediary between the bracelet and other applications. Sometimes resetting statistics and re-pairing the device helps to eliminate accumulated software errors in the operation of the accelerometer.

If you notice a systematic error (for example, the bracelet always lowers steps by 20%), try reinstalling the application or resetting the bracelet to factory settings, which will clear the temporary files and force the device to recalibrate the sensors when you first use, in most cases, this returns the accuracy of measurements to normal.

Comparison of Mi Band models and their sensors

The Mi Band line has evolved in parallel with the development of sensory input, where early models relied on simple basic filtration accelerometers, modern versions (6, 7, 8, 9) use multi-component sensor systems, which not only count steps, but also analyze sleep quality, blood oxygen (SpO2) and stress.

The difference between generations is the frequency of the sensors and the complexity of the algorithms. New models are able to distinguish finer nuances of movement, which is especially noticeable when swimming or yoga classes, where traditional step counting does not apply. The sampling rate in new chips is higher, which gives a smoother activity curve.

• 🚀 Mi Band 1-3: Basic step count, high probability of errors during active gestures.
• ⌚ Mi Band 4-5: Implementing improved algorithms, the emergence of automatic activity detection.
• 🔥 Mi Band 6-9: Use of advanced accelerometers, support for multiple sports modes and high accuracy.

When choosing a device, you should consider that even the old model, when calibrated correctly, can show results comparable to new ones in normal walking mode. However, for professional training and analyzing complex movements, it is better to focus on devices of the latest generations with an expanded set of sensors.

💡

For maximum accuracy when running, try to keep your hand relaxed without squeezing your fist too hard - this changes the accelerometer readings due to the tension of the muscles of the forearm.

💡

The accuracy of the Xiaomi pedometer depends on a combination of factors: the quality of the sensor, the density of the adhesion of the bracelet and the correctness of the anthropometric data entered by the user.

Frequently Asked Questions (FAQ)