GLAS: Global-to-Local Safe Autonomy Synthesis for Multi-Robot Motion Planning with End-to-End Learning
We present GLAS: Global-to-Local Autonomy Synthesis, a provably-safe, automated distributed policy generation for multi-robot motion planning. Our approach combines the advantage of centralized planning of avoiding local minima with the advantage of decentralized controllers of scalability and distr...
Main Authors: | , , , |
---|---|
Format: | Article in Journal/Newspaper |
Language: | unknown |
Published: |
arXiv
2020
|
Subjects: | |
Online Access: | https://dx.doi.org/10.48550/arxiv.2002.11807 https://arxiv.org/abs/2002.11807 |
id |
ftdatacite:10.48550/arxiv.2002.11807 |
---|---|
record_format |
openpolar |
spelling |
ftdatacite:10.48550/arxiv.2002.11807 2023-05-15T17:53:49+02:00 GLAS: Global-to-Local Safe Autonomy Synthesis for Multi-Robot Motion Planning with End-to-End Learning Rivière, Benjamin Hoenig, Wolfgang Yue, Yisong Chung, Soon-Jo 2020 https://dx.doi.org/10.48550/arxiv.2002.11807 https://arxiv.org/abs/2002.11807 unknown arXiv https://dx.doi.org/10.1109/lra.2020.2994035 arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ Robotics cs.RO FOS Computer and information sciences article-journal Article ScholarlyArticle Text 2020 ftdatacite https://doi.org/10.48550/arxiv.2002.11807 https://doi.org/10.1109/lra.2020.2994035 2022-03-10T16:03:47Z We present GLAS: Global-to-Local Autonomy Synthesis, a provably-safe, automated distributed policy generation for multi-robot motion planning. Our approach combines the advantage of centralized planning of avoiding local minima with the advantage of decentralized controllers of scalability and distributed computation. In particular, our synthesized policies only require relative state information of nearby neighbors and obstacles, and compute a provably-safe action. Our approach has three major components: i) we generate demonstration trajectories using a global planner and extract local observations from them, ii) we use deep imitation learning to learn a decentralized policy that can run efficiently online, and iii) we introduce a novel differentiable safety module to ensure collision-free operation, thereby allowing for end-to-end policy training. Our numerical experiments demonstrate that our policies have a 20% higher success rate than optimal reciprocal collision avoidance, ORCA, across a wide range of robot and obstacle densities. We demonstrate our method on an aerial swarm, executing the policy on low-end microcontrollers in real-time. : Accepted at IEEE RA-L, see DOI below Article in Journal/Newspaper Orca DataCite Metadata Store (German National Library of Science and Technology) |
institution |
Open Polar |
collection |
DataCite Metadata Store (German National Library of Science and Technology) |
op_collection_id |
ftdatacite |
language |
unknown |
topic |
Robotics cs.RO FOS Computer and information sciences |
spellingShingle |
Robotics cs.RO FOS Computer and information sciences Rivière, Benjamin Hoenig, Wolfgang Yue, Yisong Chung, Soon-Jo GLAS: Global-to-Local Safe Autonomy Synthesis for Multi-Robot Motion Planning with End-to-End Learning |
topic_facet |
Robotics cs.RO FOS Computer and information sciences |
description |
We present GLAS: Global-to-Local Autonomy Synthesis, a provably-safe, automated distributed policy generation for multi-robot motion planning. Our approach combines the advantage of centralized planning of avoiding local minima with the advantage of decentralized controllers of scalability and distributed computation. In particular, our synthesized policies only require relative state information of nearby neighbors and obstacles, and compute a provably-safe action. Our approach has three major components: i) we generate demonstration trajectories using a global planner and extract local observations from them, ii) we use deep imitation learning to learn a decentralized policy that can run efficiently online, and iii) we introduce a novel differentiable safety module to ensure collision-free operation, thereby allowing for end-to-end policy training. Our numerical experiments demonstrate that our policies have a 20% higher success rate than optimal reciprocal collision avoidance, ORCA, across a wide range of robot and obstacle densities. We demonstrate our method on an aerial swarm, executing the policy on low-end microcontrollers in real-time. : Accepted at IEEE RA-L, see DOI below |
format |
Article in Journal/Newspaper |
author |
Rivière, Benjamin Hoenig, Wolfgang Yue, Yisong Chung, Soon-Jo |
author_facet |
Rivière, Benjamin Hoenig, Wolfgang Yue, Yisong Chung, Soon-Jo |
author_sort |
Rivière, Benjamin |
title |
GLAS: Global-to-Local Safe Autonomy Synthesis for Multi-Robot Motion Planning with End-to-End Learning |
title_short |
GLAS: Global-to-Local Safe Autonomy Synthesis for Multi-Robot Motion Planning with End-to-End Learning |
title_full |
GLAS: Global-to-Local Safe Autonomy Synthesis for Multi-Robot Motion Planning with End-to-End Learning |
title_fullStr |
GLAS: Global-to-Local Safe Autonomy Synthesis for Multi-Robot Motion Planning with End-to-End Learning |
title_full_unstemmed |
GLAS: Global-to-Local Safe Autonomy Synthesis for Multi-Robot Motion Planning with End-to-End Learning |
title_sort |
glas: global-to-local safe autonomy synthesis for multi-robot motion planning with end-to-end learning |
publisher |
arXiv |
publishDate |
2020 |
url |
https://dx.doi.org/10.48550/arxiv.2002.11807 https://arxiv.org/abs/2002.11807 |
genre |
Orca |
genre_facet |
Orca |
op_relation |
https://dx.doi.org/10.1109/lra.2020.2994035 |
op_rights |
arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ |
op_doi |
https://doi.org/10.48550/arxiv.2002.11807 https://doi.org/10.1109/lra.2020.2994035 |
_version_ |
1766161518719664128 |