Software engineering research on the Robot Operating System: A systematic mapping study

The Robot Operating System (ROS) has become the de-facto standard framework for robotics software, and a great part of commercial robots is expected to have at least one ROS package on board in the coming years. For good quality, robotics software should rely on strong software engineering principles. In this paper, we perform a systematic mapping study on several works in software engineering on ROS, published at the top software engineering and robotics venues. Our goal is to analyze and evaluate such state-of-the-art regarding its relevance to the robotics software industry. The potentially-relevant studies are subject to a rigorously defined selection process. This results in a set of 63 primary studies on software engineering research on ROS. Those primary studies are then qualitatively analyzed according to a rigorously-defined classification framework. The results are of interest to both researchers and practitioners: (i) we provide an up-to-date overview of the state of the art on software engineering research on ROS and its potential for industrial adoption, (ii) a broad discussion of the research area as a whole, and (iii) point out routes of action for a better alignment between research and industry.