Footsteps Counter Using Variances Analysis of Three-Dimensional Accelerometer Data

The unhealthy diet then overcomes various diseases such as hypertension, diabetes, obesity, heart disease, and others. To avoid the risk of being overcome by the diseases, the community must familiarize a healthy lifestyle, such as making a healthy and regular diet with diligent exercise. One type of exercise that can be carried out by all levels of society without spending is walking exercise. Ministry of Health of Republic of Indonesia, as well as experts, recommends exercising as many as 1,000 steps to walk as a minimum target, and 10,000 steps everyday as a long-term target. So, a device is needed to be able to count the number of footsteps. The device is built by using an accelerometer sensor where the data is processed within Arduino Nano microcontroller to be sent to the user’s mobile phone to be displayed. The data processing is done by calculating the variance from each accelerometer sensor’s axis, and add them up. Then, the microcontroller calculates the difference of the sum from its average. The result of this calculation determines the threshold value to count footsteps. From the total of 18 tests, it was found the average accuracy of the device was 91,48%.