Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Kinematically Constrained Human-like Bimanual Robot-to-Human Handovers

Yasemin Göksu, Antonio De Almeida Correia, Vignesh Prasad, Alap Kshirsagar, Dorothea Koert, Jan Peters, Georgia Chalvatzaki

TL;DR

This paper uses a Hidden Semi-Markov Model to reactively generate suitable response trajectories for a robot based on the observed human partner's motion to ensure seamless and natural robot-to-human object handovers.

Abstract

Bimanual handovers are crucial for transferring large, deformable or delicate objects. This paper proposes a framework for generating kinematically constrained human-like bimanual robot motions to ensure seamless and natural robot-to-human object handovers. We use a Hidden Semi-Markov Model (HSMM) to reactively generate suitable response trajectories for a robot based on the observed human partner's motion. The trajectories are adapted with task space constraints to ensure accurate handovers. Results from a pilot study show that our approach is perceived as more human--like compared to a baseline Inverse Kinematics approach.

Kinematically Constrained Human-like Bimanual Robot-to-Human Handovers

TL;DR

This paper uses a Hidden Semi-Markov Model to reactively generate suitable response trajectories for a robot based on the observed human partner's motion to ensure seamless and natural robot-to-human object handovers.

Abstract

Bimanual handovers are crucial for transferring large, deformable or delicate objects. This paper proposes a framework for generating kinematically constrained human-like bimanual robot motions to ensure seamless and natural robot-to-human object handovers. We use a Hidden Semi-Markov Model (HSMM) to reactively generate suitable response trajectories for a robot based on the observed human partner's motion. The trajectories are adapted with task space constraints to ensure accurate handovers. Results from a pilot study show that our approach is perceived as more human--like compared to a baseline Inverse Kinematics approach.
Paper Structure (11 sections, 16 equations, 1 figure, 1 table)

This paper contains 11 sections, 16 equations, 1 figure, 1 table.

Table of Contents

  1. Introduction
  2. Learning Kinematically Constrained Bimanual Handovers
  3. Learning Bimanual Robot Handovers
  4. Conditional Generation of Robot Trajectories
  5. Optimizing Predicted Trajectory
  6. Inverse Kinematics
  7. Evaluation
  8. Training Dataset and Experiment Setup
  9. Study Design
  10. Experiment Results
  11. Discussion

Figures (1)

  • Figure 1: This diagram gives an overview of our proposed pipeline. First, we train the HSMM model using recordings of human-to-human handovers. The trained HSMM model is used to predict robot hand trajectories during a robot-to-human handover. For each time step, we predict the giver's hand positions based on the receiver's hands' trajectory. We then optimize the predicted giver hand positions to maintain a constant grip-width. Finally, we generate the appropriate robot arm motion to reach the predicted position by applying inverse kinematics.