4_2. SmartPhone
Smart Watch primarily handles Data Process.
Smartphones also detect sensors. This is because of the special circumstances occurring in the bicycle environment. If only a smart watch can be distinguished without any other equipment, it is a problem that can not distinguish between 'Turn to the left' and 'Stretch arms to the left'. To solve these issues, data is collected and used later in the 'Data subtraction' process.
In this process, data from multiple sensors is combined to represent a single value.
Typically, sensors' data have lots of noise value and in the case of a accelerometer, data has unstable values. Also, in the case of a gyro sensor, it occurs 'drift' effect that the value is raised to one side. To solve these problems and gain stable values, it passes through filtering process. The accelerometer, gyro, geomagnetic sensor values are filtered through the same process as shown in the figure, and are finally shown as a single orientation value.
The Orientation values come from smart phones and smart watches, and be subtracted from each other. Because as mentioned above, it is a problem that can not distinguish between 'Turn to the left' and 'Stretch arms to the left'. If user turns to the left, both the smartphone and smartwatch will lean to the left and subtract these two values, the result values show data in straight line form.
The resulting data will be decided through an internal motion recognition algorithm. Compare the data through sliding windows to determine the phase like up, down, left and right movement. Then, events occur for the corresponding behavior. In the current development status, direction-markings can be displayed.