EUVIMEDEUVIMED entdecken
Crossref

Deep domain adaptation eliminates costly data required for task-agnostic wearable robotic control

Keaton L Scherpereel iD, Matthew C Gombolay iD, Max K Shepherd iD, Carlos A Carrasquillo iD, Omer T Inan iD, Aaron J Young iD

Science Robotics · 2025

DOI: 10.1126/scirobotics.ads8652 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Science Robotics, 2025. Verfasst von Keaton L Scherpereel und 5 weiteren Autor:innen. Beteiligte Einrichtung: George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA 30332, USA. Hauptthemen laut MEDLINE: Wearable Electronic Devices, Exoskeleton Device, Robotics, Deep Learning. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 53 Literaturangaben. Gefördert von: National Science Foundation Graduate Research Fellowship Program, National Science Foundation Foundational Research in Robotics.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Data-driven methods have transformed our ability to assess and respond to human movement with wearable robots, promising real-world rehabilitation and augmentation benefits. However, the proliferation of data-driven methods, with the associated demand for increased personalization and performance, requires vast quantities of high-quality, device-specific data. Procuring these data is often intractable because of resource and personnel costs. We propose a framework that overcomes data scarcity by leveraging simulated sensors from biomechanical models to form a stepping-stone domain through which easily accessible data can be translated into data-limited domains. We developed and optimized a deep domain adaptation network that replaces costly, device-specific, labeled data with open-source datasets and unlabeled exoskeleton data. Using our network, we trained a hip and knee joint moment estimator with performance comparable to a best-case model trained with a complete, device-specific dataset [incurring only an 11 to 20%, 0.019 to 0.028 newton-meters per kilogram (Nm/kg) increase in error for a semisupervised model and 20 to 44%, 0.033 to 0.062 Nm/kg for an unsupervised model]. Our network significantly outperformed counterpart networks without domain adaptation (which incurred errors of 36 to 45% semisupervised and 50 to 60% unsupervised). Deploying our models in the real-time control loop of a hip/knee exoskeleton ( N = 8) demonstrated estimator performance similar to offline results while augmenting user performance based on those estimated moments (9.5 to 14.6% metabolic cost reductions compared with no exoskeleton). Our framework enables researchers to train real-time deployable deep learning, task-agnostic models with limited or no access to labeled, device-specific data.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Wearable Electronic DevicesExoskeleton DeviceRoboticsDeep LearningHumansBiomechanical PhenomenaKnee JointHip JointAlgorithmsNeural Networks, Computer

Autor:innen und Institutionen

  • George W. Woodruff School of Mechanical Engineering, Georgia Institute of Technology, Atlanta, GA 30332, USA.
  • Institute for Robotics and Intelligent Machines, Georgia Institute of Technology, Atlanta, GA 30332, USA.
  • School of Interactive Computing, Georgia Institute of Technology, Atlanta, GA 30332, USA.
  • College of Engineering, Bouvé College of Health Sciences, and Institute for Experiential Robotics, Northeastern University, Boston, MA 02115, USA.
  • School of Electrical and Computer Engineering, Georgia Institute of Technology, Atlanta, GA 30332, USA.

ORCID-Profile: Keaton L Scherpereel, Matthew C Gombolay, Max K Shepherd, Carlos A Carrasquillo, Omer T Inan, Aaron J Young

Förderung

  • National Science Foundation Graduate Research Fellowship Program
  • National Science Foundation Foundational Research in Robotics

Zitieren

APA

Scherpereel, K. L., Gombolay, M. C., Shepherd, M. K., Carrasquillo, C. A., Inan, O. T., & Young, A. J. (2025). Deep domain adaptation eliminates costly data required for task-agnostic wearable robotic control. Science Robotics. https://doi.org/10.1126/scirobotics.ads8652

Vancouver

Scherpereel KL, Gombolay MC, Shepherd MK, Carrasquillo CA, Inan OT, Young AJ. Deep domain adaptation eliminates costly data required for task-agnostic wearable robotic control. Science Robotics. 2025. doi:10.1126/scirobotics.ads8652

BibTeX
@article{Scherpereel2025,
  author = {Scherpereel, Keaton L and Gombolay, Matthew C and Shepherd, Max K and Carrasquillo, Carlos A and Inan, Omer T and Young, Aaron J},
  title = {Deep domain adaptation eliminates costly data required for task-agnostic wearable robotic control},
  journal = {Science Robotics},
  year = {2025},
  doi = {10.1126/scirobotics.ads8652}
}
Kostenlos starten →