ATRV-Jr. Interface Human-Robot Interaction: A Straw Man approach to Interface Iteration
Rachel Mulcrone
Mills College
Goal
Improve the INEEL Interface
GoalIn other words . . .
• A usable and strong interface specialized for Urban Search and Rescue of competition caliber folding in unique and novel functions and features that improve the experience for the end user and ultimately, improve performance.
Overview
BackgroundCompetitor InterfacesInterface designFuture Directions
Background Search and Rescue Human-Robot Interaction Human-Computer Interaction
Background:Search and RescueNational Institute of Standards and
Technology (NIST)– 2003 IJCAI AAAI
• 1st INEEL• 2nd Swarthmore
Background:Search and RescueSome known HRI issues in USAR
– Bumping (Over 40% rear hits)– Camera pan awareness– Situation awareness– Pitch and roll
Background: HRI
TaxonomyHuman-robot interaction currently takes many forms. Dangerous tasks, such as urban search and rescue [Casper 2002, Casper and Murphy 2002] and hazardous material clean up [Bruemmer, Marble and Dudenhoffer 2002], require a human operator to be removed from the physical location of the robot. Robots that assist the elderly [Haigh and Yanco 2002] and the handicapped [Mittal et al. 1998] share the same physical space with their users, often transporting them through the world. Others, such as Sony’s Aibo, provide entertainment and companionship for people
• Yanco and Drury (2002)
Background: HCI
Evaluation methods– Testing 3-5 Subjects
– Nielson
HCI ModelsMental Models
• User’s understanding of how a system works.
System Models
• Way the system works from programmers perspective.
Design Model
• The way the designer represents the system to the user.
Norman
Interface Design
Survey of selected USAR competitors
UMass Lowell
Rochester - IJCAI/AAAI 2003
Map
Rochester - IJCAI/AAAI 2003
Surroundings: terrain and obstacles
Distracting colors: map relative to video
Readability: black on white vs white on black
Swarthmore - IJCAI/AAAI 2003
Swarthmore - IJCAI/AAAI 2003
Window managementDevelopment tools
New Orleans - IJCAI/AAAI 2003
New Orleans - IJCAI/AAAI 2003
Window managementSmall video
Mitre - AAAI 2004 (screenshot from Robocup 2003)
Mitre - AAAI 2004 Multi-step controls for simple tasks
INEEL - IJCAI/AAAI 2003
INEEL - IJCAI/AAAI 2003 TMIControlUser Experience (final runs)
% Map and Video %
% Map and Video %
New Orleans Video: 9.11%Map 20.79%
Swarthmore Video: 14.57%Map 13.98%
Rochester Video: 25.3%Map 45.46%
Mitre Video: 8.72%Map 57.43%
% Map and Video %
% Map and Video %
INEEL Video: 10.18%Map 8.26%
% Map and Video %
% Map and Video %
UMass Lowell Video: 26.96%Map 18.89%
%Map and Video Summary%
Video Map
MITRE 9% 57%ROCH. 25% 45%NEW O. 9% 21%SWATH. 15% 14%INEEL 10% 8%UMASS 27% 19%
Interface Design: UMass Lowell
QuickTime™ and aTIFF (LZW) decompressor
are needed to see this picture.
Interface Design: UMass Lowell
Interface Design: UMass LowellCamera tilt and pan overlayRear cameraControlsResponsive interface
Interface Design: UMass LowellTo test:
– Camera positioning– Rear front toggle (animation)– Map– Sonars
Resources
This Presentation and my DMP website:– www.mercury.cs.uml.edu/~rmulcron
Human-Computer Interaction– Jakob Neilson’s Use It site
• www.useit.com
– Donald Norman’s jnd.org • www.jnd.org
NIST USAR site– http://www.isd.mel.nist.gov/projects/USAR/
Thank You
Holly, Brenden, Bobby, Marbella, Andrew, Mike, and
Phil for a great summer.