Social robots: an open-source architecture for personal assistant robots

Abstract

With the ever-increasing progress of robotics and machine learning and their vivid presence in everyday life, researchers in these fields have been very active in the last few years. However, the presence of robots in different environments still has much for improvement, both software-wise and hardware-wise. For example, personal assistant robots, to be available and useful for everybody, should be cheap, and by having features like affective computing, their interaction with humans may be more engaging. Furthermore, they should be designed so that updating their software and adding new features be an effortless task.In this paper, the authors aim to design and implement an intelligent personal assistant system on a low-cost robot. This system consists of an architecture that can be implemented on any robot with an operating system(OS) and connection to the Internet. Any part of this architecture is replaceable and easy to change. This robot will be able to recognize its owners and their feeling and will respond accordingly with different reactions in real-time. For instance, It can play a suitable song to the owner’s feelings upon their request or show various body movements to imitate emotions. Furthermore, users can interact with the robot using voice commands.