Optimización de la toma de decisiones para la gestión de tareas del robot Pepper dentro de un entorno doméstico

Abstract

In this proposal, the process for the implementation of an algorithm based on game theory for optimal decision-making in a domestic environment, with high uncertainty and using the Pepper robotic platform, is proposed within the structure of the humanorobot interaction process ( HRI). This arises from the need to solve problems presented in the development and implementation of solutions after the participation of the GED group in the @home league of RoboCup version 2019, as well as improving the performance of task management in robotics that is presented in several of the projects within the research group and in general, in any project where a robotic platform is involved. Among the interests are: determining scenarios and tasks that frame a real situation of the use of the Pepper robot in a domestic environment. Model the reception, assignment and execution dynamics of tasks from the perspective of game theory, as well as define a performance index to validate and evaluate the system, and design and implement an algorithm for optimal decision-making in a domestic environment. of human-robot interaction using the Pepper robot. With this project, the GED research group seeks to achieve different objectives, including counting the goals defined in the framework of the SINFONÍA alliance of which the Santo Tomás University is part; giving continuity to the topic related to social robotics, specifically the HRI, and through this topic strengthening the research lines of the GED group, robotics and computational intelligence. Likewise, it impacts on the teaching exercise and contributes to the curriculum of the Electronic Engineering program.