An Interactive Voice Controlled Humanoid Smart Home Prototype Using Concepts of Natural Language Processing and Machine Learning

Currently available interactive technologies and infrastructure such as open source hardware, increased processing power, artificial intelligence (AI), machine learning (ML), Natural Language processing (NLP) and connectivity through the internet, have led to a new definition of human living. Smart homes, smart streets, smart cities and eventually a smarter planet, is what humans are progressing towards. The boon in things being smart is that it involves automation and interactive technologies, something that we as humans prefer over introvert and monotonous characteristic of machines. However, most present systems have a physical entity like a remote or a personal device assistant (PDA) which is used to control our homes. When the access to these physical devices is lost, you can no longer control them. This paper addresses this very issue and proposes a solution which eliminates the need for any physical device. It proposes a multi- functional `Smart Home Automation System' (SHAS) controlled using an interactive and humanoid central processing unit which is powered by features of AI and NLP. Voice-commands to control the devices of the house such as fans, lights, doors, etc. can be given by the residents. The system is also intelligent enough to search things which it doesn't quite understand using the internet. The system has a camera integrated for continuously monitoring the home and also uses image processing techniques to identify the people trying to access the house. The system also has an automated mailing system which e-mails photos captured by the camera based on valid/invalid users. The system is aimed at being easy, cost-effective and non-physical.