Home Vision

      The project aims to automate home activities and appliances using a raspberry Pi and the Microsoft Azure Cloud Platform.

      I love integrating things together, hoping to see things in terms of the 'Grand Design'. This idea was born during my commute back home, whilst I was wishing for a SECURE way to automate my home, to literally 'welcome' me back, after a tiresome day. Thus, the idea of my own automation system, with a common-man's minimalistic approach, was born.

      The various modules of the project and their functionality are as follows:

Raspberry Pi : The Central control module will be the Raspberry Pi, on which the basic cloud access and algorithms will run on.

Microsoft Azure : The data storage via PHP and the analytics will be performed in Azure. Also, the central server to communicate to remote devices will run via it.

Machine Learning : The machine learning algorithm is, at this stage, envisioned to be a parametric, regression-based algorithm that can measure and after the learning stage, predict the user's preferences and automate according to the user's profile. The algorithm will be in python, making use of the 'pybrain' package.

Mobile : The user's mobile will be used as a user interface via an app that can be created via visual studio or eclipse. The application will also relay info, with the user's permission, to the central cloud server running in Azure, relaying data such as travel conditions, interests, notifications, actions to be completed, etc.

Arduino : Direct control of the hardware will be effected using the Arduino platform that will communicate serially with the raspberry Pi module. Prototyping will be with the Arduino Uno and higher variants can be used to scale up the device.

Future Improvements : With sufficient time and resuources, Image processing in Matlab to identify users and gestures can be implemented. Also, a DSP processor running a speech recognition algorithm, based on Mel's Cepstral Analysis can also be implemented.