EMBBEDED SYSTEM DEVELOPMENT ON WECARE: WEARABLES FOR CALORIE REPORT

• Main Author: Faith Arthanto, Yashael
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/42128
• Institution: Institut Teknologi Bandung

The fitness tracker is one of the most popular and vastly used wearables as it helps people to track their daily activities by giving them information regarding steps walked and calories burned. However, in the year 2016, it was found that diet program participant who uses similar products did not obtain significant benefits, even the weight loss they experienced is less than those who did not use such products. So, the information that is given by these products is not perfect, especially for the total calories burned. As some sources point out, similar products count steps, although the user is not walking. Based on these researches and reports, errors in detecting activity contributes to reducing the calorie prediction accuracy. WECARE, Wearables for Calorie Report, is a solution proposed to solve this problem. The solution consists of two devices, worn both in hand and foot, and a mobile application, as the user interface. WECARE utilizes machine learning technique to recognize users' activities and based on the prediction, WECARE predicts the calories burned using MET, Metabolic Equivalent of Task. In this last project, the creation of WECARE product consists of three sections, which are: the development of an embedded system, for the device worn in hand and foot; the development of the mobile application, as the user interface together with the communication system; and the data processing function, including activity recognition and step counting algorithm. The embedded system in WECARE functions to measure acceleration and angular velocity on the users' opposite wrist and ankle as an input for the data processing function.