FIESTA: Fast Incremental Euclidean Distance Fields for Online Motion Planning of Aerial Robots (1903.02144v3)

Abstract: Euclidean Signed Distance Field (ESDF) is useful for online motion planning of aerial robots since it can easily query the distance and gradient information against obstacles. Fast incrementally built ESDF map is the bottleneck for conducting real-time motion planning. In this paper, we investigate this problem and propose a mapping system called FIESTA to build global ESDF map incrementally. By introducing two independent updating queues for inserting and deleting obstacles separately, and using Indexing Data Structures and Doubly Linked Lists for map maintenance, our algorithm updates as few as possible nodes using a BFS framework. Our ESDF map has high computational performance and produces near-optimal results. We show our method outperforms other up-to-date methods in term of performance and accuracy by both theory and experiments. We integrate FIESTA into a completed quadrotor system and validate it by both simulation and onboard experiments. We release our method as open-source software for the community.

• The paper presents FIESTA, a system that incrementally builds a global ESDF to reduce computational overhead in aerial robot motion planning.
• It employs tailored data structures and a BFS-based update mechanism to efficiently manage obstacle data, outperforming methods like Voxblox.
• Experimental integration on a quadrotor platform shows superior accuracy and an order-of-magnitude speed improvement in updating distance fields.

A Comprehensive Overview of the FIESTA System

The paper "FIESTA: Fast Incremental Euclidean Distance Fields for Online Motion Planning of Aerial Robots" presents a novel approach to enhancing onboard motion planning for aerial robots through the efficient computation and updating of Euclidean Signed Distance Fields (ESDF). This research tackles the significant challenge of real-time planning in dynamic and cluttered environments, especially for Micro Aerial Vehicles (MAVs), by focusing on improving the data structures and algorithms responsible for mapping environmental obstacles.

Methodology and Key Contributions

The primary contributions of the paper include the development of a system named FIESTA, which implements an incrementally built global ESDF with high computational efficiency, and the introduction of multiple data structures to support this development. Key innovations include:

1. Data Structures and Algorithms: The authors design tailored data structures using Indexing Data Structures and Doubly Linked Lists to efficiently manage and update map data. This involves two separate queues for handling obstacle insertion and deletion, resulting in a minimized need for updating nodes within the breadth-first search (BFS) framework used.
2. ESDF Update Mechanism: The paper provides an algorithmic procedure that reduces computational overhead by selectively updating only the necessary nodes. This approach enables the FIESTA framework to produce near-optimal ESDF when compared to other state-of-the-art methods.
3. Integration and Practical Application: The system is integrated into a complete quadrotor platform, demonstrating its applicability in both simulations and real-world onboard experiments. The results highlight the system's efficiency in motion planning tasks through the maintenance of an accurate ESDF map and its ability to update incrementally.
4. Open-source Availability: By releasing FIESTA as open-source software, the authors contribute significantly to the research community, encouraging further development and application of their approach.

Numerical Results and Analysis

The evaluation demonstrates that FIESTA outperforms existing methods like Voxblox in both performance and accuracy metrics. In terms of computational speed, FIESTA exhibits an order of magnitude improvement in updating time, and achieves more accurate distance field calculations with lower root mean squared (RMS) error across various dataset tests. The performance gains are attributed to the novel BFS-based ESDF update mechanism and the efficient management of voxel data using specialized data structures.

Implications and Future Directions

The research provides a robust framework for real-time onboard motion planning in autonomous aerial robots. The efficient updating of the global ESDF map directly translates to improved responsiveness and accuracy in navigating dynamic environments, thus having significant implications for the deployment of MAVs in complex scenarios such as search and rescue, monitoring, and mapping.

The paper paves the way for several future research avenues:

• Further enhancement of the scalability and adaptability of the FIESTA framework to handle larger environments or more dynamic obstacles.
• Exploration of hybrid methods combining ESDF with machine learning algorithms for predictive modeling and optimization of motion trajectories.
• Conducting extensive field trials under diverse environmental conditions to ascertain robustness and real-world operational limitations.

Overall, the work represents a notable advancement in the optimization of motion planning systems for aerial robotics, emphasizing the potential of efficient computational models to transform autonomous navigation capabilities.