Hierarchical Trajectory Planning of Floating-Base Multi-Link Robot for Maneuvering in Confined Environments

A research team from multiple institutions has published a breakthrough in robotics planning, introducing a hierarchical framework that enables floating-base multi-link robots—essentially snake-like drones—to autonomously navigate complex, confined environments. The work, accepted to IEEE Transactions on Automation Science and Engineering (T-ASE), addresses the critical challenge of planning in high-dimensional, constraint-rich spaces where collision avoidance must be balanced with kinematic limits and dynamic feasibility. This represents a significant step toward practical deployment of articulated aerial robots for applications like autonomous inspection in industrial settings or search and rescue in disaster zones.

The framework operates in two key stages: first, it exploits the robot's dual nature (a rigid root link for guidance and articulated joints for flexibility) to generate global anchor states, decomposing the complex planning problem into manageable segments. Second, a local trajectory planner optimizes each segment in parallel using differentiable objectives and constraints, systematically enforcing kinematic feasibility and avoiding control singularities to maintain dynamic stability. Crucially, the complete system processes raw point-cloud data directly, removing the dependency on pre-defined or handcrafted obstacle models that limit adaptability in real-world scenarios. The researchers confirmed the system's effectiveness through extensive simulations and, notably, real-world experiments on a physical robot—a first for this class of planning problem—demonstrating continuous, collision-free, and dynamically feasible trajectories.