Human-Robot Interaction CMSC 691 Spring 2016 2 u What is an - - PDF document

human robot interaction
SMART_READER_LITE
LIVE PREVIEW

Human-Robot Interaction CMSC 691 Spring 2016 2 u What is an - - PDF document

2/9/16 Human-Robot Interaction CMSC 691 Spring 2016 2 u What is an interaction with a robot? u What is a robot? u What counts as interaction? u Robot(ic)s, for our purposes, is where computation meets the physical world Principles of u


slide-1
SLIDE 1

2/9/16 1

CMSC 691 Spring 2016

  • Dr. Cynthia Matuszek

Principles of Human-Robot Interaction

an introduction to robots in our daily lives

2

u What is an interaction with a robot?

u What is a robot? u What counts as interaction?

u Robot(ic)s, for our purposes, is where

computation meets the physical world

u Sensing: seeing, hearing, range-finding… u Actuation: moving, manipulating objects u Physical interactions with humans

u Speaking, hearing, gaze contact, holding hands...

Human-Robot Interaction

3

Competitions and Areas

u Search & Rescue

u Useful near-term

u Hors d’oeuvres, anyone?

u Involves crowd navigation, direct social interaction

u Assistive Robotics

u Social, physical, and cognitive support u Vulnerable populations

u Space and Extreme Condition Robotics

u Extreme operating conditions u Partial, delayed, or no communication 4

Robocup Search and Rescue

u fewkahbfk

5

Search and Rescue

6

Assistive Robotics

HAL exoskeleton: medicine, construction, disaster response DLR Institute – arm controlled by electrical brain signals

http://www.dlr.de/rmc/rm/en/desktopdefault.aspx/tabid-7315/12280_read-29190/

slide-2
SLIDE 2

2/9/16 2

7

Assistive Wheelchairs

Chiba Institute of Technology

8

Assistive Wheelchairs

Robust Robotics Group, MIT

9

Assistive Robotics

RIBA-II Transfer Robot

10

Assistive Robotics

Casper Cognitive/Social Assistant

11

Assistive Robotics

Intelligent System Corporation PARO social robot

12

Social Robotics

Leonardo, Personal Robot Group, MIT

slide-3
SLIDE 3

2/9/16 3

13

u Non-interactive work

u Still doing tasks for humans

u Teleoperating u Supervising

u Correct errors, give low-level instructions

u Providing goals

u From “Pick up the block” to “Cook dinner”

u Non-supervisory interactions

u With people u In environments (e.g., cleaning)

Types of Interaction

Low High Interaction Directness

14

u Engineering:

u Building, maintaining, (re)programming

u Teleoperating u Supervising

u Correct errors, give low-level instructions

u Providing goals

u From “Pick up the block” to “Cook dinner”

u Non-supervisory interactions

u With people u In environments (e.g., cleaning)

Types of Interaction

Low High Autonomy There are a number of formal autonomy scales!

15

System Design

u How autonomous should this robot(s) be? u What kind of group is the robot in?

u Other robots? Humans? u How is it structured? Who’s in charge?

u Information management

u Who knows what? Is it transferred?

u How adaptive or predefined is the task? u What is

is the e task sk or

  • r task

sk doma

  • main

in?

16

Autonomy

u How autonomous should this robot(s) be?

u For what kind of interaction? u What is the target problem(s)? u Are their human collaborators?

u Partial; full; social; correctable; haptic; feedback; … u What kind of autonomy?

u Goal-based u Behavior-based u Probabilistic

In practice, these are incredibly high-level simpli6ications of all of arti6icial intelligence – but useful!

17

Information Exchange

u Who knows what? Who needs to know what? u When information transferred:

u When exactly? u From whom to whom? u How slow? How costly?

u How information is transferred:

u Using what media? u In what format? (Scripted, one-way, free?)

u Do you need a cognitive model

  • f the agents?

Many human- robot interactions can also be human- human, robot- robot, or a more complex

  • mix. This is

why I often say agent.

18

Information Exchange (II)

u The medium

u Seeing u Hearing u Touch

“Put the pot here and hand me the sugar, please” OKAY

slide-4
SLIDE 4

2/9/16 4

19

Group Makeup

u Humans and robots

u 1-1, many-many, 1-many, many-1, 1-0, 0-many, … u Related to autonomy, task

u Who has authority? Who has responsibility?

u When does the human not have pure authority? u Timing? Task-based?

u What is optimal?

Autonomy Authority

20

Adaptation

u What can be learned?

u Task domain; task actions

u Changes over time

u Communication style and kind u Authority and communication flow

u What can change?

u The task? The authority? Agent capabilities? u Over what span of time? u In response to what information?

u What agent(s) should do the learning?

21

u Search & Rescue u Hors d’oeuvres, anyone? u Assistive Robotics

u Transfer u Social u Wheelchair

u Space/Extreme

Conditions

u Surgical

Practical Discussion

u Autonomy u Group Structure u Information exchange u Adaptation and

Learning