A state-of-the-art review on mobile robotics tasks using artificial intelligence and visual data
Nowadays, the field of mobile robotics has experienced an important evolution and these robots are more commonly proposed to solve different tasks autonomously. The use of visual sensors has played an important role in mobile robotics tasks during the past few years due to the advances in computer vision hardware and algorithms. It is worth remarking the use of AI tools to solve a variety of problems in mobile robotics based on the use of images either as the only source of information or combining them with other sensors such as laser or GPS. The improvement of the autonomy of mobile robots has attracted the attention of the scientific community. A considerable amount of works have been proposed over the past few years, leading to an extensive variety of approaches. Building a robust model of the environment (mapping), estimating the position within the model (localization) and controlling the movement of the robot from one place to another (navigation) are important abilities that any mobile robot must have. Considering this, this review focuses on analyzing these problems; how researchers have addressed them by means of AI tools and visual information; and how these approaches have evolved in recent years. This topic is currently open and a large number of works can be found in the related literature. Therefore, it can be of interest making an analysis of the current state of the topic. From this review, we can conclude that AI has provided robust solutions to some specific tasks in mobile robotics, such as information retrieval from scenes, mapping, localization and exploration. However, it is worth continuing to develop this line of research to find more integral solutions to the navigation problem so that mobile robots can increase their autonomy in large, complex and heterogeneous environments.