my name is iSlam, industrial designer from Egypt, it’s my last year and the project i want to make something like smart home system with voice control, can anyone help me with any advice.

  • $\begingroup$ The first step in realizing a smart home system is to program a simulator. That is a 2d computergame which contains of a groundplan in which the user can play around with buttons. The logic of the smart home system consists of an action model which are high-level-tasks, events and reactions of the systems. Github has some projects with the keyword “smart home simulator”. Don't be shy ... $\endgroup$ – Manuel Rodriguez Jun 4 '18 at 5:48
  • $\begingroup$ Welcome to SE:AI! Unfortunately, this is not a suitable question at present because it is so general. A big part of this question is software and hardware architecture, not specifically related to AI. I'm closing provisionally pending an edit with more specific details. (My thought is you are meaning to ask "which current AI methods would be most suitable for a smart home system [add details about the scope and capabilities you hope to produce, and the type of hardware.]) $\endgroup$ – DukeZhou Jun 4 '18 at 19:43

I would advise looking into NLP(natural language processing) for voice activated input. Then Object recognition and detection.

RNN (Recurrent Neural Networks) are most ideal in this case, you can code your google home pod using TensorFlow or Baidu DeepSpeech (not compatible with the google home pod) thats if you wanna start buiding your own.
Amazon is doing something with Alexa and apparently you can download alexa and make your own DIY home pod.

Object detection and recognition
CNN (Convolutional Neural Networks) are best suited for this use case, there already some predefined API's from TensorFlow. You can implement facial recognition software and other types of fancy features for security in your project.


Not the answer you're looking for? Browse other questions tagged or ask your own question.