Intelligent robots have seen significant advancements over the past few decades, with notable developments in recent years. The integration of artificial intelligence into robotics is transforming the manufacturing, agricultural, and healthcare industries, leading to increased efficiency, productivity, and the ability to perform complex tasks autonomously. In this paper, we explore four key technologies in intelligent robot systems: visual perception, decision-making, path planning, and control. Visual perception serves as a robot’s “eyes,” enabling it to recognize objects and their relationships within its environment, thereby facilitating rational decision-making and control of actions. Decision-making interprets mission objectives and environmental context to select high-level strategies or task sequences. Path planning then translates these decisions into feasible paths that serve as guidelines for robotic movement. Both decision-making and path planning can be regarded as the robot’s “cerebrum”. Control technologies function as the robot’s “cerebellum” and are responsible for regulating its actions and behaviors in the real world. This paper summarizes the cutting-edge work on intelligent robots in robotic visual perception, decision-making, path planning and control from the perspective of visual intelligence, offering insights into the state-of-the-art research in this area. Additionally, it introduces the current challenges of intelligent robots and explores future directions in the field.

By establishing continuous or transient physical contact with the environment, aerial manipulators have expanded from traditional information acquisition platforms to complex interactive systems with active operation capabilities, evolving from “passive sensing” to “active operation”. They have broad application prospects in complex interaction scenarios of large-scale energy and transportation infrastructure such as bridges, power grids, and oil and gas pipelines. Compared with non-contact flight operations, contact operations involve significant dynamic coupling, contact force constraints, and multi-modal motion switching. The introduced high-dimensional nonlinear constraints limit the feasible solution space of motion planning and restrict the autonomy and stability of the system. To address these challenges, this paper systematically reviews the research progress in the field of motion planning for aerial manipulators. Firstly, starting from the system architecture, this paper analyzes the influences of various flight platform configurations and operation mechanisms on the plannable space and operation capability. Secondly, it summarizes system modeling methods for motion planning, contact dynamics modeling, and various task planning constraints. On this basis, focusing on the planning requirements of contact operation tasks, three mainstream types of motion planning algorithms, namely sampling-based, optimization-based, and learning-based algorithms, are introduced respectively. Their development trends are analyzed, and their applicability and limitations in high-dimensional state search, dynamic constraint handling, and real-time replanning are compared. Finally, the challenges faced by aerial manipulators are summarized and future development trends are prospected.

Intelligent unmanned system combat is moving from concepts to actual combat applications. The military applications of intelligent unmanned systems are emphasized by focusing on the needs of the combat missions that they will undertake on future battlefields and summarizing the development statuses of land, air and maritime unmanned systems. Aiming at the key problems faced by intelligent unmanned systems and the practical difficulties that will be faced on the battlefield in the future, we analyze the key technologies required for intelligent unmanned systems, including autonomous perception and understanding in complex environments, behavioral decision-making and trajectory planning, autonomous navigation and positioning, multi-scenario autonomous skill learning and intelligent control, unmanned cluster cooperative control, natural human-computer interaction, etc. At the same time, aiming at the problems that will be encountered in the further development of intelligent unmanned systems on the battlefield in the future, we start with individual enhancement and cluster enhancement, and conduct a detailed analysis of the development trends of intelligent unmanned systems.