Loading…

Pioneering SE(2)-Equivariant Trajectory Planning for Automated Driving

Planning the trajectory of the controlled ego vehicle is a key challenge in automated driving. As for human drivers, predicting the motions of surrounding vehicles is important to plan the own actions. Recent motion prediction methods utilize equivariant neural networks to exploit geometric symmetri...

Full description

Saved in:
Bibliographic Details
Main Authors: Hagedorn, Steffen, Milich, Marcel, Condurache, Alexandru P.
Format: Conference Proceeding
Language:English
Subjects:
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
cited_by
cites
container_end_page 2102
container_issue
container_start_page 2097
container_title
container_volume
creator Hagedorn, Steffen
Milich, Marcel
Condurache, Alexandru P.
description Planning the trajectory of the controlled ego vehicle is a key challenge in automated driving. As for human drivers, predicting the motions of surrounding vehicles is important to plan the own actions. Recent motion prediction methods utilize equivariant neural networks to exploit geometric symmetries in the scene. However, no existing method combines motion prediction and trajectory planning in a joint step while guaranteeing equivariance under roto-translations of the input space. We address this gap by proposing a lightweight equivariant planning model that generates multi-modal joint predictions for all vehicles and selects one mode as the ego plan. The equivariant network design improves sample efficiency, guarantees output stability, and reduces model parameters. We further propose equivariant route attraction to guide the ego vehicle along a high-level route provided by an off-the-shelf GPS navigation system. This module creates a momentum from embedded vehicle positions toward the route in latent space while keeping the equivariance property. Route attraction enables goal-oriented behavior without forcing the vehicle to stick to the exact route. We conduct experiments on the challenging nuScenes dataset to investigate the capability of our planner. The results show that the planned trajectory is stable under roto-translations of the input scene which demonstrates the equivariance of our model. Despite using only a small split of the dataset for training, our method improves L2 distance at 3 s by 20.6 % and surpasses the state of the art.
doi_str_mv 10.1109/IV55156.2024.10588408
format conference_proceeding
fullrecord <record><control><sourceid>ieee_CHZPO</sourceid><recordid>TN_cdi_ieee_primary_10588408</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>10588408</ieee_id><sourcerecordid>10588408</sourcerecordid><originalsourceid>FETCH-LOGICAL-i106t-eadc9589058e9b9a9a67019eef83287ac95847b8fa13923e34dd9e804ac23033</originalsourceid><addsrcrecordid>eNo1j09Lw0AUxFdBsNZ-A4Uc9ZD0vf2T7B5LTbVQaMHgtbwmL7KlTXSTFvrtbVFPAzM_hhkhHhESRHDj-YcxaNJEgtQJgrFWg70SI5c5qwwobS3itRjIVMs4k6hvxV3XbQGMkRIHYrbybcMcfPMZvedP8jnOvw_-SMFT00dFoC2XfRtO0WpHTXOh6jZEk0Pf7qnnKnoJ_nh278VNTbuOR386FMUsL6Zv8WL5Op9OFrFHSPuYqSqdse68k93GkaM0A3TMtVXSZnQJdbaxNaFyUrHSVeXYgqZSKlBqKB5-az0zr7-C31M4rf9vqx9J5kud</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Pioneering SE(2)-Equivariant Trajectory Planning for Automated Driving</title><source>IEEE Xplore All Conference Series</source><creator>Hagedorn, Steffen ; Milich, Marcel ; Condurache, Alexandru P.</creator><creatorcontrib>Hagedorn, Steffen ; Milich, Marcel ; Condurache, Alexandru P.</creatorcontrib><description>Planning the trajectory of the controlled ego vehicle is a key challenge in automated driving. As for human drivers, predicting the motions of surrounding vehicles is important to plan the own actions. Recent motion prediction methods utilize equivariant neural networks to exploit geometric symmetries in the scene. However, no existing method combines motion prediction and trajectory planning in a joint step while guaranteeing equivariance under roto-translations of the input space. We address this gap by proposing a lightweight equivariant planning model that generates multi-modal joint predictions for all vehicles and selects one mode as the ego plan. The equivariant network design improves sample efficiency, guarantees output stability, and reduces model parameters. We further propose equivariant route attraction to guide the ego vehicle along a high-level route provided by an off-the-shelf GPS navigation system. This module creates a momentum from embedded vehicle positions toward the route in latent space while keeping the equivariance property. Route attraction enables goal-oriented behavior without forcing the vehicle to stick to the exact route. We conduct experiments on the challenging nuScenes dataset to investigate the capability of our planner. The results show that the planned trajectory is stable under roto-translations of the input scene which demonstrates the equivariance of our model. Despite using only a small split of the dataset for training, our method improves L2 distance at 3 s by 20.6 % and surpasses the state of the art.</description><identifier>EISSN: 2642-7214</identifier><identifier>EISBN: 9798350348811</identifier><identifier>DOI: 10.1109/IV55156.2024.10588408</identifier><language>eng</language><publisher>IEEE</publisher><subject>Navigation ; Neural networks ; Predictive models ; Space vehicles ; Stability analysis ; Training ; Trajectory planning</subject><ispartof>2024 IEEE Intelligent Vehicles Symposium (IV), 2024, p.2097-2102</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/10588408$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,780,784,789,790,27925,54555,54932</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/10588408$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Hagedorn, Steffen</creatorcontrib><creatorcontrib>Milich, Marcel</creatorcontrib><creatorcontrib>Condurache, Alexandru P.</creatorcontrib><title>Pioneering SE(2)-Equivariant Trajectory Planning for Automated Driving</title><title>2024 IEEE Intelligent Vehicles Symposium (IV)</title><addtitle>IV</addtitle><description>Planning the trajectory of the controlled ego vehicle is a key challenge in automated driving. As for human drivers, predicting the motions of surrounding vehicles is important to plan the own actions. Recent motion prediction methods utilize equivariant neural networks to exploit geometric symmetries in the scene. However, no existing method combines motion prediction and trajectory planning in a joint step while guaranteeing equivariance under roto-translations of the input space. We address this gap by proposing a lightweight equivariant planning model that generates multi-modal joint predictions for all vehicles and selects one mode as the ego plan. The equivariant network design improves sample efficiency, guarantees output stability, and reduces model parameters. We further propose equivariant route attraction to guide the ego vehicle along a high-level route provided by an off-the-shelf GPS navigation system. This module creates a momentum from embedded vehicle positions toward the route in latent space while keeping the equivariance property. Route attraction enables goal-oriented behavior without forcing the vehicle to stick to the exact route. We conduct experiments on the challenging nuScenes dataset to investigate the capability of our planner. The results show that the planned trajectory is stable under roto-translations of the input scene which demonstrates the equivariance of our model. Despite using only a small split of the dataset for training, our method improves L2 distance at 3 s by 20.6 % and surpasses the state of the art.</description><subject>Navigation</subject><subject>Neural networks</subject><subject>Predictive models</subject><subject>Space vehicles</subject><subject>Stability analysis</subject><subject>Training</subject><subject>Trajectory planning</subject><issn>2642-7214</issn><isbn>9798350348811</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2024</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><recordid>eNo1j09Lw0AUxFdBsNZ-A4Uc9ZD0vf2T7B5LTbVQaMHgtbwmL7KlTXSTFvrtbVFPAzM_hhkhHhESRHDj-YcxaNJEgtQJgrFWg70SI5c5qwwobS3itRjIVMs4k6hvxV3XbQGMkRIHYrbybcMcfPMZvedP8jnOvw_-SMFT00dFoC2XfRtO0WpHTXOh6jZEk0Pf7qnnKnoJ_nh278VNTbuOR386FMUsL6Zv8WL5Op9OFrFHSPuYqSqdse68k93GkaM0A3TMtVXSZnQJdbaxNaFyUrHSVeXYgqZSKlBqKB5-az0zr7-C31M4rf9vqx9J5kud</recordid><startdate>20240602</startdate><enddate>20240602</enddate><creator>Hagedorn, Steffen</creator><creator>Milich, Marcel</creator><creator>Condurache, Alexandru P.</creator><general>IEEE</general><scope>6IE</scope><scope>6IL</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIL</scope></search><sort><creationdate>20240602</creationdate><title>Pioneering SE(2)-Equivariant Trajectory Planning for Automated Driving</title><author>Hagedorn, Steffen ; Milich, Marcel ; Condurache, Alexandru P.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i106t-eadc9589058e9b9a9a67019eef83287ac95847b8fa13923e34dd9e804ac23033</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Navigation</topic><topic>Neural networks</topic><topic>Predictive models</topic><topic>Space vehicles</topic><topic>Stability analysis</topic><topic>Training</topic><topic>Trajectory planning</topic><toplevel>online_resources</toplevel><creatorcontrib>Hagedorn, Steffen</creatorcontrib><creatorcontrib>Milich, Marcel</creatorcontrib><creatorcontrib>Condurache, Alexandru P.</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan All Online (POP All Online) 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 All) 1998-Present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Hagedorn, Steffen</au><au>Milich, Marcel</au><au>Condurache, Alexandru P.</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Pioneering SE(2)-Equivariant Trajectory Planning for Automated Driving</atitle><btitle>2024 IEEE Intelligent Vehicles Symposium (IV)</btitle><stitle>IV</stitle><date>2024-06-02</date><risdate>2024</risdate><spage>2097</spage><epage>2102</epage><pages>2097-2102</pages><eissn>2642-7214</eissn><eisbn>9798350348811</eisbn><abstract>Planning the trajectory of the controlled ego vehicle is a key challenge in automated driving. As for human drivers, predicting the motions of surrounding vehicles is important to plan the own actions. Recent motion prediction methods utilize equivariant neural networks to exploit geometric symmetries in the scene. However, no existing method combines motion prediction and trajectory planning in a joint step while guaranteeing equivariance under roto-translations of the input space. We address this gap by proposing a lightweight equivariant planning model that generates multi-modal joint predictions for all vehicles and selects one mode as the ego plan. The equivariant network design improves sample efficiency, guarantees output stability, and reduces model parameters. We further propose equivariant route attraction to guide the ego vehicle along a high-level route provided by an off-the-shelf GPS navigation system. This module creates a momentum from embedded vehicle positions toward the route in latent space while keeping the equivariance property. Route attraction enables goal-oriented behavior without forcing the vehicle to stick to the exact route. We conduct experiments on the challenging nuScenes dataset to investigate the capability of our planner. The results show that the planned trajectory is stable under roto-translations of the input scene which demonstrates the equivariance of our model. Despite using only a small split of the dataset for training, our method improves L2 distance at 3 s by 20.6 % and surpasses the state of the art.</abstract><pub>IEEE</pub><doi>10.1109/IV55156.2024.10588408</doi><tpages>6</tpages></addata></record>
fulltext fulltext_linktorsrc
identifier EISSN: 2642-7214
ispartof 2024 IEEE Intelligent Vehicles Symposium (IV), 2024, p.2097-2102
issn 2642-7214
language eng
recordid cdi_ieee_primary_10588408
source IEEE Xplore All Conference Series
subjects Navigation
Neural networks
Predictive models
Space vehicles
Stability analysis
Training
Trajectory planning
title Pioneering SE(2)-Equivariant Trajectory Planning for Automated Driving
url http://sfxeu10.hosted.exlibrisgroup.com/loughborough?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-07T15%3A47%3A56IST&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=Pioneering%20SE(2)-Equivariant%20Trajectory%20Planning%20for%20Automated%20Driving&rft.btitle=2024%20IEEE%20Intelligent%20Vehicles%20Symposium%20(IV)&rft.au=Hagedorn,%20Steffen&rft.date=2024-06-02&rft.spage=2097&rft.epage=2102&rft.pages=2097-2102&rft.eissn=2642-7214&rft_id=info:doi/10.1109/IV55156.2024.10588408&rft.eisbn=9798350348811&rft_dat=%3Cieee_CHZPO%3E10588408%3C/ieee_CHZPO%3E%3Cgrp_id%3Ecdi_FETCH-LOGICAL-i106t-eadc9589058e9b9a9a67019eef83287ac95847b8fa13923e34dd9e804ac23033%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=10588408&rfr_iscdi=true