English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Human-robot handover is a fundamental yet challenging task in human-robotinteraction and collaboration. Recently, remarkable progressions have been madein human-to-robot handovers of unknown objects by using learning-based graspgenerators. However, how to responsively generate smooth motions to take anobject from a human is still an open question. Specifically, planning motionsthat take human comfort into account is not a part of the human-robot handoverprocess in most prior works. In this paper, we propose to generate smoothmotions via an efficient model-predictive control (MPC) framework thatintegrates perception and complex domain-specific constraints into theoptimization problem. We introduce a learning-based grasp reachability model toselect candidate grasps which maximize the robot's manipulability, giving itmore freedom to satisfy these constraints. Finally, we integrate a neural netforce/torque classifier that detects contact events from noisy data. Weconducted human-to-robot handover experiments on a diverse set of objects withseveral users (N=4) and performed a systematic evaluation of each module. Thestudy shows that the users preferred our MPC approach over the baseline systemby a large margin. More results and videos are available athttps://sites.google.com/nvidia.com/mpc-for-handover.

Model Predictive Control for Fluid Human-to-Robot Handovers