Voice Assistant for an Autonomous Vehicle Aimed at Elderly

In an era where autonomous vehicles are becoming increasingly relevant, older age groups often encounter significant technological barriers. This paper outlines the design of a voice assistant tailored for an elderly audience, implemented within an autonomous vehicle in view of the facilities provided by a voice control method for elderly people. The system features a modular structure composed of four key components: audio capture using Voice Activity Detection (VAD) algorithms, audio transcription and processing through Speech To Text (STT) engines, response generation using a text generation engine based on data obtained from the vehicle through Robot Operating System (ROS) and various APIs that provide the car with external information, and finally, a Text-To-Speech (TTS) engine in charge of generating an auditory response. Additionally, a user interface (UI) has been developed to further reduce the generational digital barrier. Furthermore, the research includes a comprehensive comparative review of current text generation and speech synthesis technologies, with the proposed design validated through simulations and tested by real users.