Reward-field Guided Motion Planner for Navigation with Limited Sensing Range

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F24%3A00382148" target="_blank" >RIV/68407700:21230/24:00382148 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/10.1109/IROS58592.2024.10801843" target="_blank" >https://doi.org/10.1109/IROS58592.2024.10801843</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/IROS58592.2024.10801843" target="_blank" >10.1109/IROS58592.2024.10801843</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Reward-field Guided Motion Planner for Navigation with Limited Sensing Range

Popis výsledku v původním jazyce

In this paper, we focus on improving planning efficiency for ground vehicles in navigation and exploration tasks where the environment is unknown or partially known, leading to frequent updates of the navigational goal as new sensory information is acquired. Asymptotically optimal motion planners like RRT* or FMT* can be used to plan the sequence of actions the robot can follow to achieve its current goal. Frequent replanning of the whole action sequence becomes computationally demanding when actions are not executed precisely because of limited information about the foreground terrain. The decoupled approach can decrease the computational burden with separated path planning and path following; however, it might lead to suboptimal solutions. Therefore, we propose a novel approach based on generating a reusable reward function that guides a fast sampling-based motion planner. The proposed method provides improved results in navigation scenarios compared to the former approaches, and it led to about 7% faster autonomous exploration than the decoupled approach. The present results support the suitability of the proposed method in navigation tasks with continuously updated navigation goals.

Název v anglickém jazyce

Reward-field Guided Motion Planner for Navigation with Limited Sensing Range

Popis výsledku anglicky

In this paper, we focus on improving planning efficiency for ground vehicles in navigation and exploration tasks where the environment is unknown or partially known, leading to frequent updates of the navigational goal as new sensory information is acquired. Asymptotically optimal motion planners like RRT* or FMT* can be used to plan the sequence of actions the robot can follow to achieve its current goal. Frequent replanning of the whole action sequence becomes computationally demanding when actions are not executed precisely because of limited information about the foreground terrain. The decoupled approach can decrease the computational burden with separated path planning and path following; however, it might lead to suboptimal solutions. Therefore, we propose a novel approach based on generating a reusable reward function that guides a fast sampling-based motion planner. The proposed method provides improved results in navigation scenarios compared to the former approaches, and it led to about 7% faster autonomous exploration than the decoupled approach. The present results support the suitability of the proposed method in navigation tasks with continuously updated navigation goals.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2024

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2024)

ISBN

979-8-3503-7770-5

ISSN

2153-0858

e-ISSN

2153-0866

Počet stran výsledku

8

Strana od-do

3167-3174

Název nakladatele

IEEE

Místo vydání

Piscataway

Místo konání akce

Abu Dhabi

Datum konání akce

14. 10. 2024

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

001411890000338