Applications of data mining in robotics

From a high-level viewpoint, robotics is all about data. In other words, each robot agent for performing its designated tasks should deal with data in some form - from collecting data from the environment to generating internal data. For this reason and due to broad range of possible applications, managing and dealing with such data is an important issue in the field. On the other side, data - in its different types - is known as an invaluable treasure of knowledge and insight within any information system. In this regard, extracting such hidden knowledge through data mining has been in the center of focus for more than two decades. In fact, data mining could be considered as a super-influential phenomenon on all scientific fields and robotics is not an exception. Over the recent years, many studies have been conducted on different aspects of leveraging data mining techniques in the robotic domain to improve performance of robots in various ways. However, despite the current disparate and unorganized studies in different directions, there are some potential areas that have not received enough attention. In this paper, we provide a high-level overview of the field and try to organize the current studies as a reference point and road map for future works. Moreover, challenges and future possible directions will be studied.