Real-time Optimal Navigation Planning Using Learned Motion Costs

Navigation on challenging terrain topographies requires the understanding of robots' locomotion capabilities to produce optimal solutions. We present an integrated framework for real-time autonomous navigation of mobile robots based on elevation maps. The framework performs rapid global path pl...

Full description

Saved in:
Bibliographic Details
Main Authors: Yang, Bowen, Wellhausen, Lorenz, Miki, Takahiro, Liu, Ming, Hutter, Marco
Format: Conference Proceeding
Language:English
Subjects:
Computational modeling
Costs
Navigation
Real-time systems
Sampling methods
Surfaces
Training data
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page 9289
container_issue
container_start_page 9283
container_title
container_volume
creator Yang, Bowen
Wellhausen, Lorenz
Miki, Takahiro
Liu, Ming
Hutter, Marco
description Navigation on challenging terrain topographies requires the understanding of robots' locomotion capabilities to produce optimal solutions. We present an integrated framework for real-time autonomous navigation of mobile robots based on elevation maps. The framework performs rapid global path planning and optimization that is aware of the locomotion capabilities of the robot. A GPU-aided, sampling-based path planner combined with a gradient-based path optimizer provides optimal paths by using a neural network-based locomotion cost predictor which is trained in simulation. We show that our approach is capable of planning and optimizing paths three orders of magnitude faster than RRT* on GPU-enabled hardware, enabling real-time deployment on mobile platforms. We successfully evaluate the framework on the ANYmal C quadrupedal robot in both simulations and real-world environments for path planning tasks on multiple complex terrains.
doi_str_mv 10.1109/ICRA48506.2021.9561861
format conference_proceeding
fullrecord <record><control><sourceid>ieee_CHZPO</sourceid><recordid>TN_cdi_ieee_primary_9561861</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>9561861</ieee_id><sourcerecordid>9561861</sourcerecordid><originalsourceid>FETCH-LOGICAL-i251t-e225a5b01f96fc8d702e274f480081227d0e9c1a18dbbb9ee5ea0f20520b67f43</originalsourceid><addsrcrecordid>eNotj91Kw0AQhVdBsK0-gSB5gcSZSfbvzhK0FqKVYsG7smlmy0pMSjYIvr1Ve3O-iwMf5whxi5Ahgr1blut5YSSojIAws1KhUXgmpqjJoAWt4VxMSGqdgtHvl2Ia4wcA5LlSE3G_ZtemY_jkZHU4wrXJi_sKezeGvkteW9d1odsnm_ibFbuh4yZ57v_aso9jvBIX3rWRr0-cic3jw1v5lFarxbKcV2kgiWPKRNLJGtBb5Xem0UBMuvCFATBIpBtgu0OHpqnr2jJLduAJJEGttC_ymbj59wZm3h6G49The3t6m_8AFU5Jhg</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Real-time Optimal Navigation Planning Using Learned Motion Costs</title><source>IEEE Xplore All Conference Series</source><creator>Yang, Bowen ; Wellhausen, Lorenz ; Miki, Takahiro ; Liu, Ming ; Hutter, Marco</creator><creatorcontrib>Yang, Bowen ; Wellhausen, Lorenz ; Miki, Takahiro ; Liu, Ming ; Hutter, Marco</creatorcontrib><description>Navigation on challenging terrain topographies requires the understanding of robots' locomotion capabilities to produce optimal solutions. We present an integrated framework for real-time autonomous navigation of mobile robots based on elevation maps. The framework performs rapid global path planning and optimization that is aware of the locomotion capabilities of the robot. A GPU-aided, sampling-based path planner combined with a gradient-based path optimizer provides optimal paths by using a neural network-based locomotion cost predictor which is trained in simulation. We show that our approach is capable of planning and optimizing paths three orders of magnitude faster than RRT* on GPU-enabled hardware, enabling real-time deployment on mobile platforms. We successfully evaluate the framework on the ANYmal C quadrupedal robot in both simulations and real-world environments for path planning tasks on multiple complex terrains.</description><identifier>EISSN: 2577-087X</identifier><identifier>EISBN: 1728190770</identifier><identifier>EISBN: 9781728190778</identifier><identifier>DOI: 10.1109/ICRA48506.2021.9561861</identifier><language>eng</language><publisher>IEEE</publisher><subject>Computational modeling ; Costs ; Navigation ; Real-time systems ; Sampling methods ; Surfaces ; Training data</subject><ispartof>2021 IEEE International Conference on Robotics and Automation (ICRA), 2021, p.9283-9289</ispartof><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/9561861$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,780,784,789,790,23930,23931,25140,27925,54555,54932</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/9561861$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Yang, Bowen</creatorcontrib><creatorcontrib>Wellhausen, Lorenz</creatorcontrib><creatorcontrib>Miki, Takahiro</creatorcontrib><creatorcontrib>Liu, Ming</creatorcontrib><creatorcontrib>Hutter, Marco</creatorcontrib><title>Real-time Optimal Navigation Planning Using Learned Motion Costs</title><title>2021 IEEE International Conference on Robotics and Automation (ICRA)</title><addtitle>ICRA</addtitle><description>Navigation on challenging terrain topographies requires the understanding of robots' locomotion capabilities to produce optimal solutions. We present an integrated framework for real-time autonomous navigation of mobile robots based on elevation maps. The framework performs rapid global path planning and optimization that is aware of the locomotion capabilities of the robot. A GPU-aided, sampling-based path planner combined with a gradient-based path optimizer provides optimal paths by using a neural network-based locomotion cost predictor which is trained in simulation. We show that our approach is capable of planning and optimizing paths three orders of magnitude faster than RRT* on GPU-enabled hardware, enabling real-time deployment on mobile platforms. We successfully evaluate the framework on the ANYmal C quadrupedal robot in both simulations and real-world environments for path planning tasks on multiple complex terrains.</description><subject>Computational modeling</subject><subject>Costs</subject><subject>Navigation</subject><subject>Real-time systems</subject><subject>Sampling methods</subject><subject>Surfaces</subject><subject>Training data</subject><issn>2577-087X</issn><isbn>1728190770</isbn><isbn>9781728190778</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2021</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><recordid>eNotj91Kw0AQhVdBsK0-gSB5gcSZSfbvzhK0FqKVYsG7smlmy0pMSjYIvr1Ve3O-iwMf5whxi5Ahgr1blut5YSSojIAws1KhUXgmpqjJoAWt4VxMSGqdgtHvl2Ia4wcA5LlSE3G_ZtemY_jkZHU4wrXJi_sKezeGvkteW9d1odsnm_ibFbuh4yZ57v_aso9jvBIX3rWRr0-cic3jw1v5lFarxbKcV2kgiWPKRNLJGtBb5Xem0UBMuvCFATBIpBtgu0OHpqnr2jJLduAJJEGttC_ymbj59wZm3h6G49The3t6m_8AFU5Jhg</recordid><startdate>20210530</startdate><enddate>20210530</enddate><creator>Yang, Bowen</creator><creator>Wellhausen, Lorenz</creator><creator>Miki, Takahiro</creator><creator>Liu, Ming</creator><creator>Hutter, Marco</creator><general>IEEE</general><scope>6IE</scope><scope>6IH</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIO</scope></search><sort><creationdate>20210530</creationdate><title>Real-time Optimal Navigation Planning Using Learned Motion Costs</title><author>Yang, Bowen ; Wellhausen, Lorenz ; Miki, Takahiro ; Liu, Ming ; Hutter, Marco</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i251t-e225a5b01f96fc8d702e274f480081227d0e9c1a18dbbb9ee5ea0f20520b67f43</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2021</creationdate><topic>Computational modeling</topic><topic>Costs</topic><topic>Navigation</topic><topic>Real-time systems</topic><topic>Sampling methods</topic><topic>Surfaces</topic><topic>Training data</topic><toplevel>online_resources</toplevel><creatorcontrib>Yang, Bowen</creatorcontrib><creatorcontrib>Wellhausen, Lorenz</creatorcontrib><creatorcontrib>Miki, Takahiro</creatorcontrib><creatorcontrib>Liu, Ming</creatorcontrib><creatorcontrib>Hutter, Marco</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan (POP) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE/IET Electronic Library (IEL)</collection><collection>IEEE Proceedings Order Plans (POP) 1998-present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Yang, Bowen</au><au>Wellhausen, Lorenz</au><au>Miki, Takahiro</au><au>Liu, Ming</au><au>Hutter, Marco</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Real-time Optimal Navigation Planning Using Learned Motion Costs</atitle><btitle>2021 IEEE International Conference on Robotics and Automation (ICRA)</btitle><stitle>ICRA</stitle><date>2021-05-30</date><risdate>2021</risdate><spage>9283</spage><epage>9289</epage><pages>9283-9289</pages><eissn>2577-087X</eissn><eisbn>1728190770</eisbn><eisbn>9781728190778</eisbn><abstract>Navigation on challenging terrain topographies requires the understanding of robots' locomotion capabilities to produce optimal solutions. We present an integrated framework for real-time autonomous navigation of mobile robots based on elevation maps. The framework performs rapid global path planning and optimization that is aware of the locomotion capabilities of the robot. A GPU-aided, sampling-based path planner combined with a gradient-based path optimizer provides optimal paths by using a neural network-based locomotion cost predictor which is trained in simulation. We show that our approach is capable of planning and optimizing paths three orders of magnitude faster than RRT* on GPU-enabled hardware, enabling real-time deployment on mobile platforms. We successfully evaluate the framework on the ANYmal C quadrupedal robot in both simulations and real-world environments for path planning tasks on multiple complex terrains.</abstract><pub>IEEE</pub><doi>10.1109/ICRA48506.2021.9561861</doi><tpages>7</tpages><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier EISSN: 2577-087X
ispartof 2021 IEEE International Conference on Robotics and Automation (ICRA), 2021, p.9283-9289
issn 2577-087X
language eng
recordid cdi_ieee_primary_9561861
source IEEE Xplore All Conference Series
subjects Computational modeling
Costs
Navigation
Real-time systems
Sampling methods
Surfaces
Training data
title Real-time Optimal Navigation Planning Using Learned Motion Costs
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-27T09%3A17%3A58IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_CHZPO&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=Real-time%20Optimal%20Navigation%20Planning%20Using%20Learned%20Motion%20Costs&rft.btitle=2021%20IEEE%20International%20Conference%20on%20Robotics%20and%20Automation%20(ICRA)&rft.au=Yang,%20Bowen&rft.date=2021-05-30&rft.spage=9283&rft.epage=9289&rft.pages=9283-9289&rft.eissn=2577-087X&rft_id=info:doi/10.1109/ICRA48506.2021.9561861&rft.eisbn=1728190770&rft.eisbn_list=9781728190778&rft_dat=%3Cieee_CHZPO%3E9561861%3C/ieee_CHZPO%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-i251t-e225a5b01f96fc8d702e274f480081227d0e9c1a18dbbb9ee5ea0f20520b67f43%3C/grp_id%3E%3Coa%3E%3C/oa%3E%3Curl%3E%3C/url%3E&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=9561861&rfr_iscdi=true