AvoidBench: A high-fidelity vision-based obstacle avoidance benchmarking suite for multi-rotors

Hang Yu; Guido C.H.E. De Croon; Christophe De Wagter

doi:10.1109/ICRA48891.2023.10161097

AvoidBench: A high-fidelity vision-based obstacle avoidance benchmarking suite for multi-rotors

Hang Yu, Guido C.H.E. De Croon, Christophe De Wagter

Control & Simulation

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

2 Citations (Scopus)

55 Downloads (Pure)

Abstract

Obstacle avoidance is an essential topic in the field of autonomous drone research. When choosing an avoidance algorithm, many different options are available, each with their advantages and disadvantages. As there is currently no consensus on testing methods, it is quite challenging to compare the performance between algorithms. In this paper, we propose AvoidBench, a benchmarking suite which can evaluate the performance of vision-based obstacle avoidance algorithms by subjecting them to a series of tasks. Thanks to the high fidelity of multi-rotors dynamics from RotorS and virtual scenes of Unity3D, AvoidBench can realize realistic simulated flight experiments. Compared to current drone simulators, we propose and implement both performance and environment metrics to reveal the suitability of obstacle avoidance algorithms for environments of different complexity. To illustrate AvoidBench's usage, we compare three algorithms: Ego-planner, MBPlanner, and Agile-autonomy. The trends observed are validated with real-world obstacle avoidance experiments. Code is available at: https://github.com/tudelft/AvoidBench

Original language	English
Title of host publication	Proceedings - ICRA 2023
Subtitle of host publication	IEEE International Conference on Robotics and Automation
Publisher	IEEE
Pages	9183-9189
Number of pages	7
ISBN (Electronic)	9798350323658
DOIs	https://doi.org/10.1109/ICRA48891.2023.10161097
Publication status	Published - 2023
Event	2023 IEEE International Conference on Robotics and Automation, ICRA 2023 - London, United Kingdom Duration: 29 May 2023 → 2 Jun 2023

Publication series

Name	Proceedings - IEEE International Conference on Robotics and Automation
Volume	2023-May
ISSN (Print)	1050-4729

Conference

Conference	2023 IEEE International Conference on Robotics and Automation, ICRA 2023
Country/Territory	United Kingdom
City	London
Period	29/05/23 → 2/06/23

Bibliographical note

Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care
Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

Access to Document

10.1109/ICRA48891.2023.10161097

AvoidBench_A_high-fidelity_vision-based_obstacle_avoidance_benchmarking_suite_for_multi-rotorsFinal published version, 4.39 MB

Cite this

Yu, H., De Croon, G. C. H. E., & De Wagter, C. (2023). AvoidBench: A high-fidelity vision-based obstacle avoidance benchmarking suite for multi-rotors. In Proceedings - ICRA 2023: IEEE International Conference on Robotics and Automation (pp. 9183-9189). (Proceedings - IEEE International Conference on Robotics and Automation; Vol. 2023-May). IEEE. https://doi.org/10.1109/ICRA48891.2023.10161097

@inproceedings{75c3f28dc3b04b29b1a55dee7e293c6c,

title = "AvoidBench: A high-fidelity vision-based obstacle avoidance benchmarking suite for multi-rotors",

abstract = "Obstacle avoidance is an essential topic in the field of autonomous drone research. When choosing an avoidance algorithm, many different options are available, each with their advantages and disadvantages. As there is currently no consensus on testing methods, it is quite challenging to compare the performance between algorithms. In this paper, we propose AvoidBench, a benchmarking suite which can evaluate the performance of vision-based obstacle avoidance algorithms by subjecting them to a series of tasks. Thanks to the high fidelity of multi-rotors dynamics from RotorS and virtual scenes of Unity3D, AvoidBench can realize realistic simulated flight experiments. Compared to current drone simulators, we propose and implement both performance and environment metrics to reveal the suitability of obstacle avoidance algorithms for environments of different complexity. To illustrate AvoidBench's usage, we compare three algorithms: Ego-planner, MBPlanner, and Agile-autonomy. The trends observed are validated with real-world obstacle avoidance experiments. Code is available at: https://github.com/tudelft/AvoidBench",

author = "Hang Yu and {De Croon}, {Guido C.H.E.} and {De Wagter}, Christophe",

note = "Green Open Access added to TU Delft Institutional Repository {\textquoteleft}You share, we take care!{\textquoteright} – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public. ; 2023 IEEE International Conference on Robotics and Automation, ICRA 2023 ; Conference date: 29-05-2023 Through 02-06-2023",

year = "2023",

doi = "10.1109/ICRA48891.2023.10161097",

language = "English",

series = "Proceedings - IEEE International Conference on Robotics and Automation",

publisher = "IEEE",

pages = "9183--9189",

booktitle = "Proceedings - ICRA 2023",

address = "United States",

}

Yu, H , De Croon, GCHE & De Wagter, C 2023, AvoidBench: A high-fidelity vision-based obstacle avoidance benchmarking suite for multi-rotors. in Proceedings - ICRA 2023: IEEE International Conference on Robotics and Automation. Proceedings - IEEE International Conference on Robotics and Automation, vol. 2023-May, IEEE, pp. 9183-9189, 2023 IEEE International Conference on Robotics and Automation, ICRA 2023, London, United Kingdom, 29/05/23. https://doi.org/10.1109/ICRA48891.2023.10161097

AvoidBench: A high-fidelity vision-based obstacle avoidance benchmarking suite for multi-rotors. / Yu, Hang ; De Croon, Guido C.H.E.; De Wagter, Christophe.
Proceedings - ICRA 2023: IEEE International Conference on Robotics and Automation. IEEE, 2023. p. 9183-9189 (Proceedings - IEEE International Conference on Robotics and Automation; Vol. 2023-May).

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - AvoidBench

T2 - 2023 IEEE International Conference on Robotics and Automation, ICRA 2023

AU - Yu, Hang

AU - De Croon, Guido C.H.E.

AU - De Wagter, Christophe

N1 - Green Open Access added to TU Delft Institutional Repository ‘You share, we take care!’ – Taverne project https://www.openaccess.nl/en/you-share-we-take-care Otherwise as indicated in the copyright section: the publisher is the copyright holder of this work and the author uses the Dutch legislation to make this work public.

PY - 2023

Y1 - 2023

N2 - Obstacle avoidance is an essential topic in the field of autonomous drone research. When choosing an avoidance algorithm, many different options are available, each with their advantages and disadvantages. As there is currently no consensus on testing methods, it is quite challenging to compare the performance between algorithms. In this paper, we propose AvoidBench, a benchmarking suite which can evaluate the performance of vision-based obstacle avoidance algorithms by subjecting them to a series of tasks. Thanks to the high fidelity of multi-rotors dynamics from RotorS and virtual scenes of Unity3D, AvoidBench can realize realistic simulated flight experiments. Compared to current drone simulators, we propose and implement both performance and environment metrics to reveal the suitability of obstacle avoidance algorithms for environments of different complexity. To illustrate AvoidBench's usage, we compare three algorithms: Ego-planner, MBPlanner, and Agile-autonomy. The trends observed are validated with real-world obstacle avoidance experiments. Code is available at: https://github.com/tudelft/AvoidBench

AB - Obstacle avoidance is an essential topic in the field of autonomous drone research. When choosing an avoidance algorithm, many different options are available, each with their advantages and disadvantages. As there is currently no consensus on testing methods, it is quite challenging to compare the performance between algorithms. In this paper, we propose AvoidBench, a benchmarking suite which can evaluate the performance of vision-based obstacle avoidance algorithms by subjecting them to a series of tasks. Thanks to the high fidelity of multi-rotors dynamics from RotorS and virtual scenes of Unity3D, AvoidBench can realize realistic simulated flight experiments. Compared to current drone simulators, we propose and implement both performance and environment metrics to reveal the suitability of obstacle avoidance algorithms for environments of different complexity. To illustrate AvoidBench's usage, we compare three algorithms: Ego-planner, MBPlanner, and Agile-autonomy. The trends observed are validated with real-world obstacle avoidance experiments. Code is available at: https://github.com/tudelft/AvoidBench

UR - http://www.scopus.com/inward/record.url?scp=85168689518&partnerID=8YFLogxK

U2 - 10.1109/ICRA48891.2023.10161097

DO - 10.1109/ICRA48891.2023.10161097

M3 - Conference contribution

AN - SCOPUS:85168689518

T3 - Proceedings - IEEE International Conference on Robotics and Automation

SP - 9183

EP - 9189

BT - Proceedings - ICRA 2023

PB - IEEE

Y2 - 29 May 2023 through 2 June 2023

ER -

AvoidBench: A high-fidelity vision-based obstacle avoidance benchmarking suite for multi-rotors

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this