Kiwi | what you do, how you feel



Inside Fitness Trackers: What Makes the Magic Happen

What makes your wearable tick? How does your fitness tracker do anything at all? Want to know what's inside a FitBit? To understand, you’ll need to know what's inside the device and what those things do. 

 

We’ll take you through the data flow inside the fitness tracker, and how it finally gets to you in a form you can process. Just a caveat, some wearables depending on their complexity, may include more stages than are described here. However, by and large, most fitness trackers perform all these steps.

There are five main components of a fitness tracker, three of which are critical in collecting and transferring the data, and two that are critical for keeping the sensor going.

 

Stage 1 - Sensor Detection 

In the beginning, there was nothing. Nothing but data. Then you moved, and instantly, tiny mechanical systems begin to collect, calibrate, and convert, beginning the first step of the activity/motion recognition process. 

 

The muscles of the device, motion sensors detect an analog signal and convert it to a digital signal through an analog to digital converter (ADC). Sensors can include accelerometers, gyroscopes, magnetometers, barometers, heart rate sensors and more.

 

In an accelerometer, one method of detecting the signal consists of measuring the difference in the electrical charge between a mass that has moved due to acceleration and a stationary object.

 

Accelerometers and gyroscopes are present in almost all wearables, with heart rate monitors popular at the higher-end, while magnetometers and barometers, though less common, are rapidly gaining in popularity.

 Labelled diagram of an accelerometer

 

Stage 2 - Microcontroller Calculations

The brains of the group, the microcontroller houses the software that allows your movements to be tracked. The sensor feeds the microcontroller fresh motion data, which then digests the data using its motion algorithms and spits out information in a form that can be processed by us. This usually takes the form of frequency graphs.

Running Peak Detection

In more complex wearables, the microcontroller may contain sensor fusion capabilities. These allow it to combine the signals from two different sensors, like an accelerometer and a magnetometer and provide more accurate data. 

 Atmel Microcontroller

 

Stage 3 - Bluetooth Data Transmission 

Your wearable’s going to need some way to transmit data for it to be visualized, and currently, the best way is through a Bluetooth transmitter. Bluetooth’s become popular because of its low power usage and flexibility in how much data can be transferred.

 

Bluetooth chips come in two types: regular and LE (low energy). LE is more common due to its lower power requirements. Regular Bluetooth is used when a device requires a large amount of data to be transferred and battery life is not an issue. 

 

Stage 4 - Bluetooth Data Reception

A receiving device, like a smartphone or a computer, will then take the bluetooth signal and run it through software in order to display usable and workable data visualisations.

 

Bluetooth receivers are susceptible to interference from the microcontroller and other external signals  such as WiFi signals, radio signals, and even things like air conditioning.

 

The aim is to have as high a SNR (signal-to-noise ratio) as possible; the higher the ratio, the better the signal. As a result, you should look carefully before selecting your device's bluetooth circuit.

 

Stage 5 - Battery Power and Charging

Tracking the body’s movements and converting them into computer-readable signals consumes a lot of energy. Your tracker needs a battery that can generate enough energy to power the device for at least a day, and can recharge quickly.

 

Manufacturers balance higher battery lives with cost, with capabilities ranging from a few hours to more than 6 months. However, improvements in the capabilities of the battery are incremental, with major advances coming from improvements to processors or optimization of battery usage.