Arduino, Raspberry Pi, and Smartphone Usage Comparison for Voice-based Virtual Assistant

Abstract

A virtual assistant or an artificial intelligence chatbot is one of the technological interventions to help individuals with borderline personality disorder (BPD). The virtual assistant must have a bank of knowledge to answer the questions from individuals with BPD to accompany and chat with them like a digital friend. As part of artificial intelligence software, Program-O can handle the interaction between users and the bank of knowledge in the server. The voice recognition and textto-speech features can be used to enhance the virtual assistant’s function to become even closer as a friend. There are three possibilities for implementing voice recognition and text-to-speech features on virtual assistants, which are using Arduino, Raspberry Pi, and smartphone devices. All these three devices can be connected to the speech recognition module or Application Protocol Interface (API) for voice recognition on the internet. This paper will explore and compare several aspects of the implementation of voice recognition and text-to-speech features in the virtual assistant. The result will be a recommendation for the application of voice recognition and text-to-speech features on a virtual assistant for individuals with BPD.