Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

On the Effect of Robot Errors on Human Teaching Dynamics

Jindan Huang, Isaac Sheidlower, Reuben M. Aronson, Elaine Schaertl Short

TL;DR

Investigation of the presence and severity of robot errors affect three dimensions of human teaching dynamics: feedback granularity, feedback richness, and teaching time, in both forced-choice and open-ended teaching contexts shows that people tend to spend more time teaching robots with errors, provide more detailed feedback over specific segments of a robot’s trajectory.

Abstract

Human-in-the-loop learning is gaining popularity, particularly in the field of robotics, because it leverages human knowledge about real-world tasks to facilitate agent learning. When people instruct robots, they naturally adapt their teaching behavior in response to changes in robot performance. While current research predominantly focuses on integrating human teaching dynamics from an algorithmic perspective, understanding these dynamics from a human-centered standpoint is an under-explored, yet fundamental problem. Addressing this issue will enhance both robot learning and user experience. Therefore, this paper explores one potential factor contributing to the dynamic nature of human teaching: robot errors. We conducted a user study to investigate how the presence and severity of robot errors affect three dimensions of human teaching dynamics: feedback granularity, feedback richness, and teaching time, in both forced-choice and open-ended teaching contexts. The results show that people tend to spend more time teaching robots with errors, provide more detailed feedback over specific segments of a robot's trajectory, and that robot error can influence a teacher's choice of feedback modality. Our findings offer valuable insights for designing effective interfaces for interactive learning and optimizing algorithms to better understand human intentions.

On the Effect of Robot Errors on Human Teaching Dynamics

TL;DR

Investigation of the presence and severity of robot errors affect three dimensions of human teaching dynamics: feedback granularity, feedback richness, and teaching time, in both forced-choice and open-ended teaching contexts shows that people tend to spend more time teaching robots with errors, provide more detailed feedback over specific segments of a robot’s trajectory.

Abstract

Human-in-the-loop learning is gaining popularity, particularly in the field of robotics, because it leverages human knowledge about real-world tasks to facilitate agent learning. When people instruct robots, they naturally adapt their teaching behavior in response to changes in robot performance. While current research predominantly focuses on integrating human teaching dynamics from an algorithmic perspective, understanding these dynamics from a human-centered standpoint is an under-explored, yet fundamental problem. Addressing this issue will enhance both robot learning and user experience. Therefore, this paper explores one potential factor contributing to the dynamic nature of human teaching: robot errors. We conducted a user study to investigate how the presence and severity of robot errors affect three dimensions of human teaching dynamics: feedback granularity, feedback richness, and teaching time, in both forced-choice and open-ended teaching contexts. The results show that people tend to spend more time teaching robots with errors, provide more detailed feedback over specific segments of a robot's trajectory, and that robot error can influence a teacher's choice of feedback modality. Our findings offer valuable insights for designing effective interfaces for interactive learning and optimizing algorithms to better understand human intentions.
Paper Structure (25 sections, 10 figures)

This paper contains 25 sections, 10 figures.

Table of Contents

  1. Introduction
  2. Background
  3. Methodology
  4. Experiment Setup
  5. Environment
  6. Robot Trajectories & Error Severity.
  7. Interactive teaching interface.
  8. Measures
  9. Feedback granularity
  10. Feedback richness
  11. Teaching time
  12. Experiment Procedure
  13. Results
  14. Influence on Feedback Granularity
  15. Results of forced-choice teaching
  16. ...and 10 more sections

Figures (10)

  • Figure 1: In both forced-choice and open-ended teaching contexts, people adapt their teaching behavior based on the presence or absence of robot errors.
  • Figure 2: Salad preparation study environment. The robot arm is tasked with picking up the correct salad ingredient and pouring that ingredient into the bowl.
  • Figure 3: Interactive teaching interface. The participant watches a robot practicing video, specifies a feedback window and provides feedback to the robot learner.
  • Figure 4: Types of feedback used in our forced-choice teaching and open-ended teaching contexts
  • Figure 5: Feedback granularity data from forced-choice teaching, grouped by the presence of robot errors.
  • ...and 5 more figures