Simulating human-in-the-loop optimization of exoskeleton assistance to compare optimization algorithm performance

Assistive robotic devices like exoskeletons offer the promise of improving mobility for millions of people. However, developing devices that improve an objective mobility metric is challenging. Human-in-the-loop optimization is a systematic approach for personalizing robotic assistance to maximize a...

Full description

Saved in:
Bibliographic Details
Published in:bioRxiv
Main Authors: Kutulakos, Zoe, Slade, Patrick
Format: Paper
Language:English
Published: Cold Spring Harbor Laboratory 09.04.2024
Edition:1.1
Subjects:
ISSN:2692-8205
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract Assistive robotic devices like exoskeletons offer the promise of improving mobility for millions of people. However, developing devices that improve an objective mobility metric is challenging. Human-in-the-loop optimization is a systematic approach for personalizing robotic assistance to maximize a mobility metric that has improved device performance for different metrics and applications. Successfully performing human-in-the-loop optimization requires the experimenter to make many decisions, like selecting the appropriate optimization algorithm, hyperparameters, and convergence criteria. Typically, selecting these experimental settings involves pilot experimentation. We propose an approach that uses a probabilistic surrogate model, mapping assistance parameters to corresponding experimental evaluations of the objective mobility metric, to simulate human-in-the-loop optimization and inform these decisions. In this paper, we form a surrogate model of the metabolic landscape of walking with exoskeleton assistance using an existing experimental dataset. We simulate human-in-the-loop optimization by using a synthetic metabolic landscape model to evaluate the metabolic cost of walking with different assistance parameters, instead of performing an experimental measurement. We perform three simulated scenarios optimizing assistance for an expert subject, a novice subject adapting to the device, and an expert subject with up to 20 assistance parameters. The code and analyses from this work are open-source to promote use by other researchers. Simulation enables direct comparison of optimization settings to inform experimental human-in-the-loop optimization and potentially reduce the resources and time required to develop effective assistive devices.
AbstractList Assistive robotic devices like exoskeletons offer the promise of improving mobility for millions of people. However, developing devices that improve an objective mobility metric is challenging. Human-in-the-loop optimization is a systematic approach for personalizing robotic assistance to maximize a mobility metric that has improved device performance for different metrics and applications. Successfully performing human-in-the-loop optimization requires the experimenter to make many decisions, like selecting the appropriate optimization algorithm, hyperparameters, and convergence criteria. Typically, selecting these experimental settings involves pilot experimentation. We propose an approach that uses a probabilistic surrogate model, mapping assistance parameters to corresponding experimental evaluations of the objective mobility metric, to simulate human-in-the-loop optimization and inform these decisions. In this paper, we form a surrogate model of the metabolic landscape of walking with exoskeleton assistance using an existing experimental dataset. We simulate human-in-the-loop optimization by using a synthetic metabolic landscape model to evaluate the metabolic cost of walking with different assistance parameters, instead of performing an experimental measurement. We perform three simulated scenarios optimizing assistance for an expert subject, a novice subject adapting to the device, and an expert subject with up to 20 assistance parameters. The code and analyses from this work are open-source to promote use by other researchers. Simulation enables direct comparison of optimization settings to inform experimental human-in-the-loop optimization and potentially reduce the resources and time required to develop effective assistive devices.
Author Kutulakos, Zoe
Slade, Patrick
Author_xml – sequence: 1
  givenname: Zoe
  orcidid: 0009-0008-2091-050X
  surname: Kutulakos
  fullname: Kutulakos, Zoe
  organization: School of Engineering and Applied Sciences, Harvard University
– sequence: 2
  givenname: Patrick
  orcidid: 0000-0001-9302-3911
  surname: Slade
  fullname: Slade, Patrick
  email: slade@seas.harvard.edu
  organization: School of Engineering and Applied Sciences, Harvard University
BookMark eNpVUMtKAzEADKJgrf0Abzl6Sc1js5scpfiCggd7X7KbpI1ukjVJpfr1bqkXYWAY5nGYK3AeYjAA3BC8JASTO4pptcQT-JKLRgp6Bma0lhQJivklWOT8jjGmsiasqWbg8835_aCKC1u423sVkAuo7AwaYhxhHIvz7meyY4DRQnOI-cMMpkxS5exyUaE3sETYRz-qZP431LCNyZWdh6NJNiZ_TF-DC6uGbBZ_PAebx4fN6hmtX59eVvdr1DU1RVJoqYmhTBMpBee9qnrLGsuFkV2tldaacSyYlVVnFLGSa66UJF1thGZGsjm4Pc12LqaD-2rH5LxK3-3xoBZP4O3pIPYLV2RhRA
ContentType Paper
Copyright 2024, Posted by Cold Spring Harbor Laboratory
Copyright_xml – notice: 2024, Posted by Cold Spring Harbor Laboratory
DBID FX.
DOI 10.1101/2024.04.05.587982
DatabaseName bioRxiv
DatabaseTitleList
DeliveryMethod fulltext_linktorsrc
Discipline Biology
EISSN 2692-8205
Edition 1.1
ExternalDocumentID 2024.04.05.587982v1
GroupedDBID 8FE
8FH
AFKRA
ALMA_UNASSIGNED_HOLDINGS
BBNVY
BENPR
BHPHI
FX.
HCIFZ
LK8
M7P
NQS
PIMPY
PROAC
RHI
ID FETCH-LOGICAL-b762-98d9d1e23d199855ca4cf37f58e9b6daddd35083f94bea1f95d5aa91b6e8d3e93
IngestDate Tue Jan 07 18:50:45 EST 2025
IsDoiOpenAccess true
IsOpenAccess true
IsPeerReviewed false
IsScholarly false
Language English
License This pre-print is available under a Creative Commons License (Attribution-NonCommercial 4.0 International), CC BY-NC 4.0, as described at http://creativecommons.org/licenses/by-nc/4.0
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-b762-98d9d1e23d199855ca4cf37f58e9b6daddd35083f94bea1f95d5aa91b6e8d3e93
Notes Competing Interest Statement: The authors have declared no competing interest.
ORCID 0000-0001-9302-3911
0009-0008-2091-050X
OpenAccessLink https://www.biorxiv.org/content/10.1101/2024.04.05.587982
PageCount 9
ParticipantIDs biorxiv_primary_2024_04_05_587982
PublicationCentury 2000
PublicationDate 20240409
PublicationDateYYYYMMDD 2024-04-09
PublicationDate_xml – month: 4
  year: 2024
  text: 20240409
  day: 9
PublicationDecade 2020
PublicationTitle bioRxiv
PublicationYear 2024
Publisher Cold Spring Harbor Laboratory
Publisher_xml – name: Cold Spring Harbor Laboratory
References 2024.04.05.587982v1.26
Kim, Liu, Kim, Lee, Meguid, Walsh, Kuindersma (2024.04.05.587982v1.18) 2019
Xu, Liu, Chen, Yu, Yan, Yang, Zhou, Yang (2024.04.05.587982v1.14) 2023; 11
Slade, Kochenderfer, Delp, Collins (2024.04.05.587982v1.12) 2022; 610
Lee, Kim, Panizzolo, Zhou, Baker, Galiana, Malcolm, Walsh (2024.04.05.587982v1.10) 2017; 2
Kim, Ding, Malcolm, Speeckaert, Siviy, Walsh, Kuindersma (2024.04.05.587982v1.17) 2017; 12
Okoro (2024.04.05.587982v1.1) 2018; 67
Nogueira (2024.04.05.587982v1.25) 2014
Catkin, Patoglu (2024.04.05.587982v1.16) 2023
Fang, Orekhov, Lerner (2024.04.05.587982v1.2) 2021; 69
Hybart, Villancio-Wolter, Ferris (2024.04.05.587982v1.3) 2023; 11
Ding, Kim, Kuindersma, Walsh (2024.04.05.587982v1.9) 2018; 3
Poggensee, Collins (2024.04.05.587982v1.23) 2021; 6
Sawicki, Beck, Kang, Young (2024.04.05.587982v1.4) 2020; 17
Bryan, Franks, Song, Voloshina, Reyes, O’Donovan, Gregorczyk, Collins (2024.04.05.587982v1.11) 2021; 18
Seth, Hicks, Uchida, Habib, Dembia, Dunne, Ong, DeMers, Rajagopal, Millard (2024.04.05.587982v1.5) 2018; 14
Hamaya, Matsubara, Noda, Teramae, Morimoto (2024.04.05.587982v1.22) 2017
Zhang, Fiers, Witte, Jackson, Poggensee, Atkeson, Collins (2024.04.05.587982v1.7) 2017; 356
Caputo, Collins (2024.04.05.587982v1.15) 2014; 136
Felt, Selinger, Donelan, Remy (2024.04.05.587982v1.19) 2015; 10
Kantharaju, Jeong, Ramadurai, Jacobson, Jeong, Kim (2024.04.05.587982v1.8) 2022; 30
Kochenderfer, Wheeler (2024.04.05.587982v1.24) 2019
Kim, Quinlivan, Deprey, Revi, Eckert-Erdheim, Murphy, Orzel, Walsh (2024.04.05.587982v1.13) 2022; 12
Garcia-Rosas, Tan, Oetomo, Manzie, Choong (2024.04.05.587982v1.20) 2019; 51
Franks, Bryan, Reyes, O’Donovan, Gregorczyk, Collins (2024.04.05.587982v1.6) 2022; 30
Wen, Si, Brandt, Gao, Huang (2024.04.05.587982v1.21) 2019; 50
References_xml – year: 2014
  ident: 2024.04.05.587982v1.25
  publication-title: Bayesian Optimization: Open source constrained global optimization tool for Python
– volume: 17
  start-page: 1
  year: 2020
  end-page: 9
  ident: 2024.04.05.587982v1.4
  article-title: The exoskeleton expansion: improving walking and running economy
  publication-title: Journal of neuroengineering and rehabilitation
– volume: 18
  start-page: 1
  year: 2021
  end-page: 12
  ident: 2024.04.05.587982v1.11
  article-title: Optimized hip-knee-ankle exoskeleton assistance reduces the metabolic cost of walking with worn loads
  publication-title: Journal of neuroengineering and rehabilitation
– volume: 67
  year: 2018
  ident: 2024.04.05.587982v1.1
  article-title: Prevalence of disabilities and health care access by disability status and type among adults—united states, 2016
  publication-title: MMWR. Morbidity and mortality weekly report
– volume: 51
  start-page: 1070
  issue: 2
  year: 2019
  end-page: 1084
  ident: 2024.04.05.587982v1.20
  article-title: Personalized online adaptation of kinematic synergies for human-prosthesis interfaces
  publication-title: IEEE transactions on cybernetics
– volume: 12
  start-page: 11004
  issue: 1
  year: 2022
  ident: 2024.04.05.587982v1.13
  article-title: Reducing the energy cost of walking with low assistance levels through optimized hip flexion assistance from a soft exosuit
  publication-title: Scientific reports
– start-page: 9173
  year: 2019
  end-page: 9179
  ident: 2024.04.05.587982v1.18
  publication-title: In 2019 international conference on robotics and automation (icra)
– volume: 50
  start-page: 2346
  issue: 6
  year: 2019
  end-page: 2356
  ident: 2024.04.05.587982v1.21
  article-title: Online reinforcement learning control for the personalization of a robotic knee prosthesis
  publication-title: IEEE transactions on cybernetics
– start-page: 5907
  year: 2017
  end-page: 5912
  ident: 2024.04.05.587982v1.22
  publication-title: In 2017 IEEE International Conference on Robotics and Automation (ICRA)
– volume: 3
  start-page: eaar5438
  issue: 15
  year: 2018
  ident: 2024.04.05.587982v1.9
  article-title: Human-in-the-loop optimization of hip assistance with a soft exosuit during walking
  publication-title: Science robotics
– volume: 2
  start-page: eaan6708
  issue: 6
  year: 2017
  ident: 2024.04.05.587982v1.10
  article-title: Reducing the metabolic cost of running with a tethered soft exosuit
  publication-title: Science Robotics
– volume: 14
  start-page: e1006223
  issue: 7
  year: 2018
  ident: 2024.04.05.587982v1.5
  article-title: Opensim: Simulating musculoskeletal dynamics and neuromuscular control to study human and animal movement
  publication-title: PLoS computational biology
– volume: 11
  start-page: 1006326
  year: 2023
  ident: 2024.04.05.587982v1.14
  article-title: Reducing the muscle activity of walking using a portable hip exoskeleton based on human-in-the-loop optimization
  publication-title: Frontiers in Bioengineering and Biotechnology
– volume: 69
  start-page: 2143
  issue: 7
  year: 2021
  end-page: 2152
  ident: 2024.04.05.587982v1.2
  article-title: Improving the energy cost of incline walking and stair ascent with ankle exoskeleton assistance in cerebral palsy
  publication-title: IEEE Transactions on Biomedical Engineering
– year: 2023
  ident: 2024.04.05.587982v1.16
  article-title: Preference-based human-in-the-loop optimization for perceived realism of haptic rendering
  publication-title: IEEE Transactions on Haptics
– volume: 136
  start-page: 035002
  issue: 3
  year: 2014
  ident: 2024.04.05.587982v1.15
  article-title: A universal ankle–foot prosthesis emulator for human locomotion experiments
  publication-title: Journal of biomechanical engineering
– volume: 6
  start-page: eabf1078
  issue: 58
  year: 2021
  ident: 2024.04.05.587982v1.23
  article-title: How adaptation, training, and customization contribute to benefits from exoskeleton assistance
  publication-title: Science Robotics
– volume: 356
  start-page: 1280
  issue: 6344
  year: 2017
  end-page: 1284
  ident: 2024.04.05.587982v1.7
  article-title: Human-in-the-loop optimization of exoskeleton assistance during walking
  publication-title: Science
– volume: 11
  start-page: e15775
  year: 2023
  ident: 2024.04.05.587982v1.3
  article-title: Metabolic cost of walking with electromechanical ankle exoskeletons under proportional myoelectric control on a treadmill and outdoors
  publication-title: PeerJ
– volume: 610
  start-page: 277
  issue: 7931
  year: 2022
  end-page: 282
  ident: 2024.04.05.587982v1.12
  article-title: Personalizing exoskeleton assistance while walking in the real world
  publication-title: Nature
– volume: 10
  start-page: e0135342
  issue: 8
  year: 2015
  ident: 2024.04.05.587982v1.19
  article-title: “body-in-the-loop”: Optimizing device parameters using measures of instantaneous energetic cost
  publication-title: PloS one
– ident: 2024.04.05.587982v1.26
– volume: 30
  start-page: 2494
  year: 2022
  end-page: 2505
  ident: 2024.04.05.587982v1.6
  article-title: The effects of incline level on optimized lower-limb exoskeleton assistance: A case series
  publication-title: IEEE Transactions on Neural Systems and Rehabilitation Engineering
– volume: 30
  start-page: 1786
  year: 2022
  end-page: 1795
  ident: 2024.04.05.587982v1.8
  article-title: Reducing squat physical effort using personalized assistance from an ankle exoskeleton
  publication-title: IEEE Transactions on Neural Systems and Rehabilitation Engineering
– volume: 12
  issue: 9
  year: 2017
  ident: 2024.04.05.587982v1.17
  article-title: Human-in-the-loop bayesian optimization of wearable device parameters
  publication-title: PloS one
– year: 2019
  ident: 2024.04.05.587982v1.24
  publication-title: Algorithms for optimization
SSID ssj0002961374
Score 1.7188787
SecondaryResourceType preprint
Snippet Assistive robotic devices like exoskeletons offer the promise of improving mobility for millions of people. However, developing devices that improve an...
SourceID biorxiv
SourceType Open Access Repository
SubjectTerms Bioengineering
Title Simulating human-in-the-loop optimization of exoskeleton assistance to compare optimization algorithm performance
URI https://www.biorxiv.org/content/10.1101/2024.04.05.587982
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV3PT9swFLYQBYnbGCB-DORJ3JCBOHFiX4dAHFBVQYUQl8pJbIjaxqXNqu6_33NsmrDuwA6ToiiKYid6n_38xfZ7H0Kngc650okkGbAFElEZEyEo9CuZUattwWW9ev54l3S7_OlJ9LxK36yWE0jKki8WYvJfoYZ7ALYNnf0HuJeVwg24BtDhDLDD-VPAPxTjWpKrfHECfKQoCbA8MjJmcmbAQ4x96KXliWphZkMYeWxyDaDRlkzang6E1G9O_1hCjl7MtKhexzbd8XvAQZvfpoW5XxTzZoGogm8Zuq18z2bZiB5GPt2vUwgYtuceqNuy0ni4KzMCUlxPQNqYI2i0Z3eu6Rq_h1nVnozGAtwureOr_-Kza60AW32ddJadM54Ip0j0RyrslWfm8N_boQkT4M86P667vfvl_BoVQFSSyC9kwzsuVkrDLw9YZQpWaVGK_hfU6Ukw4jZaU-VXtOk0QX_toLcGP7yCH26jgY3GLfxwgx-uDPb4fSyxxA-38NtF_Zvr_tUt8boYJIWhiwieizxQNMxtfCRjmYwyHSaacSXSOIcBKw9tkn8tolTJQAuWMylFkMaK56ES4R5aL02p9hG24qEySDUUZlEccAljTkx1rEPBY8mjA_TdW2gwcclPBtaKg0s42MBZ8fATzxyhrab5fEPr1fSnOkYb2bwqZtMTD91vcgNTVQ
linkProvider ProQuest
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Simulating+human-in-the-loop+optimization+of+exoskeleton+assistance+to+compare+optimization+algorithm+performance&rft.jtitle=bioRxiv&rft.au=Kutulakos%2C+Zoe&rft.au=Slade%2C+Patrick&rft.date=2024-04-09&rft.pub=Cold+Spring+Harbor+Laboratory&rft.eissn=2692-8205&rft_id=info:doi/10.1101%2F2024.04.05.587982&rft.externalDocID=2024.04.05.587982v1