GLAS: Global-to-Local Safe Autonomy Synthesis for Multi-Robot Motion Planning With End-to-End Learning

We present GLAS: G lobal-to- L ocal A utonomy S ynthesis, 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...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE robotics and automation letters Jg. 5; H. 3; S. 4249 - 4256
Hauptverfasser: Riviere, Benjamin, Honig, Wolfgang, Yue, Yisong, Chung, Soon-Jo
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Piscataway IEEE 01.07.2020
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Schlagworte:
Collision avoidance
Distributed robot systems
imitation learning
Learning
Measurement
Microcontrollers
Motion planning
Multiple robots
Optimal control
path planning for multiple mobile robots or agents
Planning
Policies
Robot dynamics
Robots
Safety
State (computer science)
Trajectory
ISSN:2377-3766, 2377-3766
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:We present GLAS: G lobal-to- L ocal A utonomy S ynthesis, 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.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:2377-3766
2377-3766
DOI:10.1109/LRA.2020.2994035