A Comprehensive System for Indoor Assistance and User Interaction

The contemporary world is characterized by a growing integration of virtual assistants that simplify everyday life by supporting users in multiple activities. In recent years, we have witnessed the rising necessity to design and develop multi-agent systems, with which users can interact using natural language, allowing them to locate points of interest and carry out specific activities. With this research, we aim to provide a contribution to this setting by developing a practical and real application that allows an autonomous system to answer different questions, such as “Where is this user?” and “Bring documents to this specific user”. We present an architecture that allows users to interact with an autonomous system based on Neural Networks and equipped with advanced algorithms for indoor localization and path-finding implementation to achieve specific goals. The implemented system was subjected to various experiments, demonstrating the effectiveness of specific frameworks, such as DEMOCLE and Hermes, in achieving the ultimate goal. The solution also integrates the implementation of a mesh network and the use of the Bluetooth protocol to facilitate internal localization.