Programming Nao as an Educational Agent: A Comparison Between Choregraphe and Python SDK

Abstract

Programming a humanoid robot for educational purpose is a demanding task for a beginner with little experience. Several studies are available in which humanoid robots such as NAO, are used in educational settings to move, recognize objects and hold conversations similar to a human. These studies usually incorporate third party libraries and advanced deep-learning methods making it difficult for a beginner to follow. This paper aims to work as a getting-started guide for someone starting out with programming the NAO robot using Choregraphe and the Python SDK. In this study, NAO robot is used to implement four scenarios based on - dialog, movement, object recognition and obstacle avoidance - using the available components that come with the robot. The paper focuses on comparing the Choregraphe and NAO Python SDK during this process by considering the advantages and limitations of both approaches. The results show that both Choregraphe and the Python SDK have their nuances and their usage depends on the use case. However, for a beginner just starting out, Choregraphe is easier to get things done without writing a single line of code. Python, on the other hand is useful for low-level functionalities and provides rather more flexibility.