Online Whole-body Motion Planning for Quadrotor using Multi-resolution Search (2209.06761v2)

Abstract: In this paper, we address the problem of online quadrotor whole-body motion planning (SE(3) planning) in unknown and unstructured environments. We propose a novel multi-resolution search method, which discovers narrow areas requiring full pose planning and normal areas requiring only position planning. As a consequence, a quadrotor planning problem is decomposed into several SE(3) (if necessary) and R^3 sub-problems. To fly through the discovered narrow areas, a carefully designed corridor generation strategy for narrow areas is proposed, which significantly increases the planning success rate. The overall problem decomposition and hierarchical planning framework substantially accelerate the planning process, making it possible to work online with fully onboard sensing and computation in unknown environments. Extensive simulation benchmark comparisons show that the proposed method is one to several orders of magnitude faster than the state-of-the-art methods in computation time while maintaining high planning success rate. The proposed method is finally integrated into a LiDAR-based autonomous quadrotor, and various real-world experiments in unknown and unstructured environments are conducted to demonstrate the outstanding performance of the proposed method.

• The paper introduces a multi-resolution search strategy that decomposes the motion planning problem into SE(3) and ℝ³ sub-problems.
• It employs a hierarchical planning framework that speeds up computation by focusing intensive calculations on narrow, constrained spaces.
• Extensive simulations and real-world experiments demonstrate orders-of-magnitude improvements in computation time and robust trajectory execution.

Online Whole-body Motion Planning for Quadrotor using Multi-resolution Search

The paper presents a comprehensive approach to the challenging problem of online whole-body motion planning for quadrotors operating in unknown and unstructured environments. The authors introduce a novel multi-resolution search method that aids in dynamically navigating through environments requiring both SE(3) and $\mathbb{R}^3$ planning. This approach leverages the ability of quadrotors to perform aggressive maneuvers by considering the full spatial degrees of freedom, including both orientation and position, within narrow passages.

Methodological Contributions

The central contribution of this paper is the development of a multi-resolution search strategy. This strategy effectively differentiates between narrow areas, which necessitate SE(3) planning due to their constraints on both position and orientation, and broader environments suitable for the less complex $\mathbb{R}^3$ planning. This decomposition approach not only maintains the computation within feasible time limits but also enhances the robustness of the planning process by uniquely addressing the environment's structural demands:

• Parallel Multi-resolution Search: The novelty lies in concurrently using high and low-resolution map searches to delineate areas pressing on quadrotor orientation from those concerning only position. This dual-path search effectively segments the planning problem into manageable sub-problems requiring different planning intricacies, thereby aligning computational effort with the problem's nature.
• Hierarchical Planning Framework: The integration of the hierarchical planning strategy allows substantial acceleration of the planning process. By focusing computationally intense SE(3) planning strictly on areas identified as narrow, and employing faster $\mathbb{R}^3$ planning otherwise, the proposed method achieves remarkable improvements in computation time compared to existing state-of-the-art techniques.
• Corridor Generation Strategy: For effective passage through narrow gaps, the authors propose a corridor generation technique that increases planning success rates by providing a feasible path through seemingly non-traversable spaces. This approach actively avoids over-conservatism in spatial constraints, prevalent in previous methods.

Experimental Validation

The paper provides an extensive suite of simulations and real-world experiments, reinforcing the proposed method's capability to outperform existing approaches significantly in both computational efficiency and success rate. Notably, simulations demonstrate computation time improvements by up to several orders of magnitude over competing methods. Real-world tests involving a LiDAR-equipped autonomous quadrotor further validate the approach's viability in practical applications.

Theoretical and Practical Implications

The implications of this work are manifold:

• Theoretical: The decomposition of 3D motion planning into focused sub-problems based on environmental characteristics sets an important precedent for future research. It encourages further exploration into hybrid planning strategies that can intelligently adapt the computational load depending on environmental complexity.
• Practical: The capability to plan and execute trajectories in real-time using onboard sensors and computation promotes the practical deployment of quadrotors in dynamic environments. The demonstrated LiDAR-based implementation enhances their utility in applications such as search and rescue, where operating conditions are unpredictable.

Future Directions

Looking forward, the enhancement of perception strategies to optimize narrow area detection and expanding the method's applicability to varied environments are natural extensions of this work. Additionally, integrating more sophisticated models of environmental interaction, which include dynamic obstacles and moving targets, could further augment the robustness and applicability of this planning method.

To conclude, the method presented combines a multi-resolution search strategy with a hierarchical planning framework, establishing a robust and computationally efficient approach for quadrotor navigation. This paper successfully advances the field of motion planning by marrying theoretical insights with practical implementation, resulting in a significant stride towards fully autonomous aerial systems.