Abstract
This report introduces a definition for robot task intelligence and explains the basis of the challenges associated with implementing robot task intelligence in real-world problems. Two fundamental elements of this intelligence, robot manipulation and navigation, are introduced herein. We describe the existing trends and industrial applications of the robot task intelligence in logistics and agricultural automation. Furthermore, as an underpinning technology for this intelligence, we review the existing three-dimensional position estimation techniques and summarize the difficulties associated with applying pure SLAM technology to real-world applications. Finally, we discuss the prospects of the robot task intelligence research and its potential for solving real-world problems.