Percetpion, Localization and Navigation. For autonomous vehicles (AVs), including ground, aerial and underwater vehicles, environmental perception, localization, and navigation are of utmost importance for accomplishing complex autonoumous tasks, underpinning the feasibility of other high-level missions, such as atunomous driving. Roughly speaking, how to let an AV preciesly perceive its surrounding environment, be aware of where it is, and plan what to do next adhering to specified objectives is fundamental for unmanned vehicles.
Multi-Vehicle Cooperation. Cooperation among a group of AVs (even human in the loop), inlcuding both homogeneous and heterogeneous AVs, is significant for practical large-scale problems (e.g., cooperative search, reconnaissance, surveillance and rescue), due to the consideration of data privacy, capability limitation, attacking robustness, etc. In this respect, we are especially interested in the collaboration between UAVs and AGVs, such as autonomous docking of UAVs onto stationary platforms or moving ground/aquatic vehicles.
Adversarial Games. Aside from cooperation among AVs, a vast array of realistic problems exhibit the opposition hallmark, for example, drone racing. This type of problems can genereally be modelled as adversarial games, consisting of multiple players (e.g., UAVs and human), which inevitably entail cutting-edge techniques, including game theory, etc.
AI methods. Artificial intelligence (AI) has been one of active and momentous fields all over the world, which is deemed to possess puissant fortes especially for complex problems with massive data and unclear models. Reinforcement learning, deep neural networks, among others, are instances of AI methods. It is promising to leverage or come up with AI methods to conquer the aforementioned problems in AVs and other possible fields.