Calorie tracking and sleep monitoring using DynaSense
MetadataShow full item record
We are inundated with sensors and devices around us that capture snapshots of our daily lives. Sensors from various devices communicate with each other to share data. Some sensors can provide the same form of data, such as a step count, using different mechanisms or devices. It is often the case that each of them have a different way of accessing the data for our use. It can be seen that there is a need to access data in a way that the focus is shifted from what sensor is being used to what is being sensed. This thesis presents DynaSense, a dynamic sensing middleware system that provides a unified approach of accessing data from sensors. By focusing on data instead of data sources, DynaSense can fetch relevant data from any source that has the capability of providing a particular type of data. The source can then be a hardware sensor, a web URL, a database, or a composite source comprising of the above. Such a mechanism helps in writing a program due to conciseness of code and obsolescence of code tangential to the program's functionality. The design is then illustrated by an implementation of such middleware service, built on Android, along with a few applications that make use of it. The key applications used here are a Calorie Tracker and a Sleep Monitor. We then observe the difference between developing applications with and without DynaSense to discuss the merits and demerits of DynaSense.