An affordance and distance minimization based method for computing object orientations for robot human handovers

Wesley P. Chan, Matthew K.X.J. Pan, Elizabeth A. Croft, Masayuki Inaba

Research output: Contribution to journalArticleResearchpeer-review

Abstract

The ability to hand over objects to humans is an important skill for service robots. However, determining the proper object pose for handover is a challenging task. Our approach, based on observations of a set of natural human handovers, addresses three related challenges in teaching robots how to hand over objects: (1) how to compute mathematically an appropriate ‘standard’ or ‘mean’ handover orientation, (2) how to ascertain whether an observed set is of good or poor quality, and (3) using (1) and (2), how to compute an appropriate handover orientation from a set, in a manner that is robust to the quality of the set. We first compare three methods for computing mean orientations and show that our proposed distance minimization based method yields the best results. Next, we show that using the concept of affordance axes, we can evaluate the quality of a set of observed orientations. Finally, using affordance axes together with random sample consensus, we devise a method for computing an appropriate handover orientation from a set of observed natural handover orientations. User study data verified that our methods are successful in identifying both good and poor quality sets of handover orientations and in computing appropriate handover orientations from observed natural handover orientations. These results enable robots to automatically learn proper handover orientations for various objects.

Original languageEnglish
Number of pages20
JournalInternational Journal of Social Robotics
DOIs
Publication statusAccepted/In press - 15 Apr 2019

Keywords

  • Assistive robots
  • Human–robot handover
  • Human–robot interaction
  • Object handover

Cite this

@article{b06dc836731b41509d5f6bd4acbb6676,
title = "An affordance and distance minimization based method for computing object orientations for robot human handovers",
abstract = "The ability to hand over objects to humans is an important skill for service robots. However, determining the proper object pose for handover is a challenging task. Our approach, based on observations of a set of natural human handovers, addresses three related challenges in teaching robots how to hand over objects: (1) how to compute mathematically an appropriate ‘standard’ or ‘mean’ handover orientation, (2) how to ascertain whether an observed set is of good or poor quality, and (3) using (1) and (2), how to compute an appropriate handover orientation from a set, in a manner that is robust to the quality of the set. We first compare three methods for computing mean orientations and show that our proposed distance minimization based method yields the best results. Next, we show that using the concept of affordance axes, we can evaluate the quality of a set of observed orientations. Finally, using affordance axes together with random sample consensus, we devise a method for computing an appropriate handover orientation from a set of observed natural handover orientations. User study data verified that our methods are successful in identifying both good and poor quality sets of handover orientations and in computing appropriate handover orientations from observed natural handover orientations. These results enable robots to automatically learn proper handover orientations for various objects.",
keywords = "Assistive robots, Human–robot handover, Human–robot interaction, Object handover",
author = "Chan, {Wesley P.} and Pan, {Matthew K.X.J.} and Croft, {Elizabeth A.} and Masayuki Inaba",
year = "2019",
month = "4",
day = "15",
doi = "10.1007/s12369-019-00546-7",
language = "English",
journal = "International Journal of Social Robotics",
issn = "1875-4791",
publisher = "Springer",

}

An affordance and distance minimization based method for computing object orientations for robot human handovers. / Chan, Wesley P.; Pan, Matthew K.X.J.; Croft, Elizabeth A.; Inaba, Masayuki.

In: International Journal of Social Robotics, 15.04.2019.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - An affordance and distance minimization based method for computing object orientations for robot human handovers

AU - Chan, Wesley P.

AU - Pan, Matthew K.X.J.

AU - Croft, Elizabeth A.

AU - Inaba, Masayuki

PY - 2019/4/15

Y1 - 2019/4/15

N2 - The ability to hand over objects to humans is an important skill for service robots. However, determining the proper object pose for handover is a challenging task. Our approach, based on observations of a set of natural human handovers, addresses three related challenges in teaching robots how to hand over objects: (1) how to compute mathematically an appropriate ‘standard’ or ‘mean’ handover orientation, (2) how to ascertain whether an observed set is of good or poor quality, and (3) using (1) and (2), how to compute an appropriate handover orientation from a set, in a manner that is robust to the quality of the set. We first compare three methods for computing mean orientations and show that our proposed distance minimization based method yields the best results. Next, we show that using the concept of affordance axes, we can evaluate the quality of a set of observed orientations. Finally, using affordance axes together with random sample consensus, we devise a method for computing an appropriate handover orientation from a set of observed natural handover orientations. User study data verified that our methods are successful in identifying both good and poor quality sets of handover orientations and in computing appropriate handover orientations from observed natural handover orientations. These results enable robots to automatically learn proper handover orientations for various objects.

AB - The ability to hand over objects to humans is an important skill for service robots. However, determining the proper object pose for handover is a challenging task. Our approach, based on observations of a set of natural human handovers, addresses three related challenges in teaching robots how to hand over objects: (1) how to compute mathematically an appropriate ‘standard’ or ‘mean’ handover orientation, (2) how to ascertain whether an observed set is of good or poor quality, and (3) using (1) and (2), how to compute an appropriate handover orientation from a set, in a manner that is robust to the quality of the set. We first compare three methods for computing mean orientations and show that our proposed distance minimization based method yields the best results. Next, we show that using the concept of affordance axes, we can evaluate the quality of a set of observed orientations. Finally, using affordance axes together with random sample consensus, we devise a method for computing an appropriate handover orientation from a set of observed natural handover orientations. User study data verified that our methods are successful in identifying both good and poor quality sets of handover orientations and in computing appropriate handover orientations from observed natural handover orientations. These results enable robots to automatically learn proper handover orientations for various objects.

KW - Assistive robots

KW - Human–robot handover

KW - Human–robot interaction

KW - Object handover

UR - http://www.scopus.com/inward/record.url?scp=85064660848&partnerID=8YFLogxK

U2 - 10.1007/s12369-019-00546-7

DO - 10.1007/s12369-019-00546-7

M3 - Article

JO - International Journal of Social Robotics

JF - International Journal of Social Robotics

SN - 1875-4791

ER -