Pushing the Limits of Reactive Planning: Learning to Escape Local Minima

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Meijer, Isar, Pantic, Michael, Oleynikova, Helen, Siegwart, Roland
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913435322679296
author Meijer, Isar
Pantic, Michael
Oleynikova, Helen
Siegwart, Roland
author_facet Meijer, Isar
Pantic, Michael
Oleynikova, Helen
Siegwart, Roland
contents When does a robot planner need a map? Reactive methods that use only the robot's current sensor data and local information are fast and flexible, but prone to getting stuck in local minima. Is there a middle-ground between fully reactive methods and map-based path planners? In this paper, we investigate feed forward and recurrent networks to augment a purely reactive sensor-based planner, which should give the robot geometric intuition about how to escape local minima. We train on a large number of extremely cluttered worlds auto-generated from primitive shapes, and show that our system zero-shot transfers to real 3D man-made environments, and can handle up to 30% sensor noise without degeneration of performance. We also offer a discussion of what role network memory plays in our final system, and what insights can be drawn about the nature of reactive vs. map-based navigation.
format Preprint
id arxiv_https___arxiv_org_abs_2407_13530
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Pushing the Limits of Reactive Planning: Learning to Escape Local Minima
Meijer, Isar
Pantic, Michael
Oleynikova, Helen
Siegwart, Roland
Robotics
When does a robot planner need a map? Reactive methods that use only the robot's current sensor data and local information are fast and flexible, but prone to getting stuck in local minima. Is there a middle-ground between fully reactive methods and map-based path planners? In this paper, we investigate feed forward and recurrent networks to augment a purely reactive sensor-based planner, which should give the robot geometric intuition about how to escape local minima. We train on a large number of extremely cluttered worlds auto-generated from primitive shapes, and show that our system zero-shot transfers to real 3D man-made environments, and can handle up to 30% sensor noise without degeneration of performance. We also offer a discussion of what role network memory plays in our final system, and what insights can be drawn about the nature of reactive vs. map-based navigation.
title Pushing the Limits of Reactive Planning: Learning to Escape Local Minima
topic Robotics
url https://arxiv.org/abs/2407.13530