Editorial: Applications of AI in autonomous, surveillance, and robotic systems

摘要

The rapid advancement of Artificial Intelligence (AI) has ushered in a new era of intelligent systems capable of autonomous decision-making, real-time perception, and adaptive learning. Across robotics, unmanned systems, and surveillance technologies, AI is transforming how machines interact with the world-enhancing precision, e iciency, and autonomy. This editorial provides an overview of key research contributions in this domain, highlighting the pivotal role of AI in shaping the future of autonomous navigation, robotic manipulation, and multi-agent surveillance systems.The selected papers in this issue address critical challenges in visual perception, swarm coordination, and secure data management, demonstrating how AI-driven solutions are pushing the boundaries of what autonomous systems can achieve. From Visual SLAM in robotics to blockchain-secured UAV swarms, these studies showcase both the current state of the art and future directions for research and deployment.AI in Autonomous Robotics: Perception, Navigation, and ManipulationThe paper "A Review of Visual SLAM for Robotics: Evolution, Properties, and Future Applications" provides a comprehensive survey of how Simultaneous Localization and Mapping (SLAM) has evolved with deep learning and computer vision. Traditional SLAM systems relied heavily on geometric algorithms, but modern AI-powered approaches leverage neural networks for feature extraction, dynamic object tracking, and semantic understanding. This shift enables robots to operate in complex, unstructured environments-from industrial warehouses to disaster zones. Future research directions include lightweight SLAM for edge devices and multi-agent collaborative mapping.In "Image-based Robot Navigation with Task Achievability," the authors explore AI-driven navigation strategies that go beyond simple path planning. By integrating reinforcement learning and vision-based perception, robots can assess environmental constraints and adjust their trajectories to ensure task completion. This is particularly valuable in logistics, where mobile robots must dynamically reroute around obstacles while maintaining operational e iciency.A major challenge in robotic manipulation is the discrepancy between simulated training data and real-world execution. The study "6IMPOSE: Bridging the Reality Gap in 6D Pose Estimation for Robotic Grasping" introduces a novel AI framework that improves 6D pose estimation-critical for precise grasping in industrial automation and healthcare robotics. By combining synthetic data augmentation with domain adaptation techniques, the proposed method significantly enhances robotic dexterity in unstructured settings.Unmanned Aerial Vehicles (UAVs) are increasingly used for surveillance, delivery, and infrastructure inspection, but their operational endurance is limited by battery life. The paper "Vision-based Safe Autonomous UAV Docking with Panoramic Sensors" presents an AI-powered docking system that allows UAVs to autonomously land on charging stations using panoramic vision. This innovation enhances mission longevity while reducing human intervention, paving the way for fully autonomous drone fleets.The integration of UAVs, Unmanned Ground Vehicles (UGVs), and Unmanned Marine Vehicles (UMVs) into a cohesive surveillance network is explored in "UAV-UGV-UMV Multi-Swarms for Cooperative Surveillance." AI-driven coordination enables these systems to share sensory data, optimize coverage, and adapt to dynamic threats. Applications range from border security to environmental monitoring, where multidomain swarms provide unparalleled situational awareness.As UAV swarms become more prevalent, ensuring data integrity and preventing cyber threats is paramount. "Towards a Blockchain-Based Multi-UAV Surveillance System" proposes a decentralized AI framework where blockchain technology secures communication between drones. This approach mitigates risks such as spoofing and data tampering, making autonomous sur