CS 309: Autonomous Intelligent Robotics FRI I Lecture 09: - - PowerPoint PPT Presentation

cs 309 autonomous intelligent robotics fri i lecture 09
SMART_READER_LITE
LIVE PREVIEW

CS 309: Autonomous Intelligent Robotics FRI I Lecture 09: - - PowerPoint PPT Presentation

CS 309: Autonomous Intelligent Robotics FRI I Lecture 09: Introduction to HRI Instructor: Justin Hart http://justinhart.net/teaching/2019_spring_cs309/ Human-Robot Interaction Briefly, the study of how humans and robots interact with each


slide-1
SLIDE 1

CS 309: Autonomous Intelligent Robotics FRI I Lecture 09: Introduction to HRI Instructor: Justin Hart

http://justinhart.net/teaching/2019_spring_cs309/

slide-2
SLIDE 2

Human-Robot Interaction

  • Briefly, the study of how humans and robots

interact with each other

– There has been some form of interacting with

machinery since the very first robots, including simple electro-mechanical devices that arguably may count from the turn of the 20th century (Shultz & Goodrich, 2007)

slide-3
SLIDE 3

Human-Robot Interaction

  • Research encompasses

many themes

– Newer hot topics:

  • Measuring human emotion

& engagement

  • Human-robot autonomous

collaboration

  • Physical human-robot

interaction

  • HRI as design
slide-4
SLIDE 4

Human-Robot Interaction

  • Research encompasses many themes

– Long-standing hot topics:

  • Generating and interpreting gaze & joint attention
  • Language and HRI
  • Gesturing
  • Back-channel communication (things other than gestures

and language)

  • Proxemics
  • Vocal prosody
slide-5
SLIDE 5

An explosion happened in the 90s

  • Show Kismet video
slide-6
SLIDE 6

Interesting points from Kismet video

  • One of the more interesting angles on HRI is

that the human interpretation of the stimulus is the most important factor

– Robots as evocative devices – The use of psychological methods to describe and

interpret results

slide-7
SLIDE 7

Experimental HRI & HRI as Psych

  • User studies as a primary driver in HRI

– The assigned paper is an example

  • Likert scale measures

– The robot was lifelike 1 - - - 5

  • Cognitive difference scale measures

– The robot was <not at all lifelike> - - - <very lifelike>

slide-8
SLIDE 8

Experimental HRI & HRI as Psych

  • Intra-participant vs Inter-participant designs

– Will the study be harmed by users seeing all

possible conditions?

– Can you get enough data? – Does the order in which a participant sees stimuli

affect the outcome? Can you account for this?

slide-9
SLIDE 9

Experimental HRI & HRI as Psych

  • Establishing meaningful results

– Statistical testing

  • t-tests
  • ANOVA (and variants)

– Post-hoc tests

  • Correlation tests
  • p-values
  • Statistical significance
slide-10
SLIDE 10

Androids

  • Often tele-operated or

partially tele-operated

– Geminoid

  • Newer ones are

autonomous

– Erica

  • Erica video
slide-11
SLIDE 11

Humanoids

  • Androids are designed

to look very human

– Appearance

emphasized

  • Humanoids have a

roughly human form

– Functionality

emphasized

slide-12
SLIDE 12

Humanoids

  • Sometimes, similar

experiments appear on both

  • Erica

– Pre-programmed

conversation engine

  • Robovie

– Learning dialogue from

humans

slide-13
SLIDE 13

The Uncanny Valley

slide-14
SLIDE 14

Physical HRI (pHRI)

  • Handovers

– Video

  • Shared carrying
  • Exoskeletons
  • Collaborative assembly and manufacturing

– Video

slide-15
SLIDE 15

Today's paper

  • No Fair! An Interaction with a Cheating Robot

– Short documentary – Control

  • Robot Plays Rock Paper Scissors

– Verbal

  • Robot verbally announces itself the winner

– Action

  • Robot changes its gesture before announcing itself the

winner

slide-16
SLIDE 16

Hallway

1_1 1_1 2_1 2_1 3_1 3_1

slide-17
SLIDE 17

Hallway

1_1 1_1 2_1 2_1 3_1 3_1

slide-18
SLIDE 18

1_2 1_3 2_2 2_3 3_2 3_3 1_1 2_1 3_1

slide-19
SLIDE 19

1_1 1_1 1_2 1_2 1_3 1_3 2_1 2_1 2_3 2_3 3_1 3_1 3_2 3_2 3_3 3_3