Graduation T20

Home Automation Using Machine Learning

The Project is based on the idea to automate Home using the algorithms of Machine Learning. The projects covers two major parts:

1. Neural Network Based Lobby
2. Command Operated Electronic Appliances

1. Neural Network Based Lobby:

Aim:

Make an energy efficient lighting system for places such as lobbies and corridors such that energy at each point is above a minimum threshold, wasting minimum energy. A human analogous model was designed using the neural network algorithm.

Approach:

The network learns the pattern of bulbs in the lobby according to the intensity values obtained by LDR sensors placed at various points and the atmospheric conditions. Accordingly the trained network gives the pattern of bulbs according to the atmospheric condition such that energy at each point is approximately equal to the threshold.

2. Command Operated Electronic Appliances

Aim:

To run electronic appliances through command sent from an Android phone, so that we can control all the appliances from your fingertips.

Approach:

The keywords for the use cases are identified. Then the voice command is processed to get the command and accordingly the value for the each appliance is set. This is then transferred to the micro-controller through blue-tooth which further turns on or off the appliances accordingly.

Code Repo: https://github.com/sourabhvarshney111/Home-Automation

While doing the project, we ran into many issues including data collection, feature extraction and all Machine Learning and NLP related issues. It took us a really great effort to get out from each of these issues and come out stronger. But we, finally made it.

It was the hard work of all my team members to produce a gem project like this. Kudos to my team members Harnish Rajput, Mansi Sampat, Aanchal Handa, and Vasu Eranki for being the best team-mates we needed for this project.