Transcript defacto

The DEFACTO System:
Training Incident Commanders
Nathan Schurr
Janusz Marecki, Milind Tambe, Nikhil
Kasinadhuni, and J. P. Lewis
University of Southern California
Paul Scerri
Carnegie Mellon University
Outline





Motivation and Domain
DEFACTO
Team Level Adjustable Autonomy
Experiments with DEFACTO
Conclusions
Motivation: Help Incident
Commanders



Incident Commander
First Response
Disaster Rescue Scenario





Urban Environment
Large Scale
Crime Scene
Incident commander must control situation,
monitor situation, and allocate resources
Goal: Initially a Training Simulation

Later: Decision Support/Replacement
LAFD Exercise: Simulations by
People Playing Roles
Aims of DEFACTO

LAFD Exercise Challenges




Key Exercise Components




Personnel Heavy
Smaller Scale
Low Fidelity Environment
Communication
Allocation
Agent-teams replace people playing roles
Demonstrating Effective Flexible Agent
Coordination of Teams via Omnipresence
Outline





Motivation and Domain
DEFACTO
Team Level Adjustable Autonomy
Experiments with DEFACTO
Conclusions
DEFACTO Architecture
Disaster Rescue Simulation:
USC Map, Different underlying simulators
Statistics
Challenges in Extending to
Human-Agent Teams

Teamwork



Communication
Role Allocation
Agent team to incorporate human

Adjustable Autonomy (Scerri et al JAIR
2002)

Interface
DEFACTO

Teamwork Proxies

Machinetta



Flexible Interaction



Continued development with CMU
Used in many other domains – UAVs, sensor nets etc.
Team Level Adjustable Autonomy Strategies
Dynamic Strategy Selection
Omni-Viewer



2D – Standard with Simulator
3D – Developed by us
Interaction
DEFACTO Architecture
Proxy Architecture

Abstracted Theories of Teamwork (Scerri et al AAMAS 03)

Communication: communication with other proxies
Coordination: reasoning about team plans and communication
State: the working memory of the proxy
Adjustable Autonomy: reasoning about whether to act
autonomously or pass control to the team member
RAP Interface: communication with the team member




Other
Proxies
Communication
RAP Interface
State
Coordination
Adjustable Autonomy
RAP
Teamwork Proxies

Higher level TOP
Reuse across domain
Flexible Teamwork (Tambe JAIR 97)

Communication




Allocation


Joint Intentions (Cohen & Levesque 1991)
Role allocation algorithms (Xu et al AAMAS 2005)
Machinetta



Platform Independent
Modular Structure
Downloadable – Free, Publicly available
Outline





Motivation and Domain
DEFACTO
Team Level Adjustable Autonomy
Experiments with DEFACTO
Conclusions
Adjustable Autonomy(AA)
Strategies for Teams

Agents dynamically adjust own level of
autonomy



Agents act autonomously, but also...
Give up autonomy, transferring control to
humans
When to transfer decision-making control



Whenever human has superior expertise
Yet, too many interrupts also problematic
Previous: Individual agent-human interaction
AA: Novel Challenges in
Teams

Transfer of control strategies for AA in teams

Planned sequence of transfers of control

AT - Team level A strategy
H - Human strategy for all tasks
AH - Individual A followed by H
ATH - Team level A strategy followed by H

Goal: Improve Team Performance



DEFACTO Architecture
Omni-Viewer
DEFACTO Movie
Outline





Motivation and Domain
DEFACTO
Team Level Adjustable Autonomy
Experiments with DEFACTO
Conclusions
Experiments


Initial evaluation of system and of strategies
Details



3 Subjects
Allocation Viewer
Same Map for each scenario





Building size and location
Initial position of fires
4, 6, and 10 agents
A, H, AH, ATH Strategies
Averaged over 3 runs
Empirical Studies with Users
Subject B
Subject A
Buildings Saved
250
200
150
100
50
0
3
5
7
9
300
250
200
150
100
50
0
3
11
5
A
H
AH
7
A
ATH
H
Subject C
300
250
200
150
100
50
0
3
5
9
Number of Fire Engines
Number of Fire Engines
Buildings Saved
Buildings Saved
300
7
9
Number of Fire Engines
A
H
AH
ATH
11
AH
ATH
11
Conclusions from Results



No strategy dominates through all cases
Humans may sometimes degrade agent team
results
Slope of strategy A > Slope of H

Humans are not as good at exploiting additional
agents resources

If EQH is low, then as we grow to larger
numbers of agents, A will dominate AH, ATH
and H
Dip at 6…

LAFD – “Not surprising.”

Summary

DEFACTO
 Teamwork
 Team Level Adjustable Autonomy Strategies
 Interface
Experimented with strategies for adjustable autonomy

Future Directions





Experiments with LAFD
Study strategy behavior
Train the “system”
Training today, real response in the future.
Future
Thank You



Email: [email protected]
Web Site: http://teamcore.usc.edu
Machinetta


http://teamcore.usc.edu/doc/Machinetta/
Thanks


CREATE Center
Fred Pighin and Pratik Patil
Related Work: Disaster Response
Simulations

LA County Fire Department Simulators


DEFACTO focuses on “incident
commander”
“Environment” simulators:
E.g., Terrasim, EPICS
 Not provide on agent behaviors


“Agent-based” simulators
E.g., Battlefield simulators
 Adjustable autonomy

Strategy Models
Models of the Strategies
Models of the Strategies
Outline

Motivation




DEFACTO






Simulator
Teamwork Proxies
3D Visualization
Team Level Adjustable Autonomy


Objectives
CREATE Research Center
Current State of the Art
Models
Predictions
Experiments with DEFACTO
Conclusions
DEFACTO: Key Research Areas

Enable effective interactions of agents with humans



Scale-up to 100s of agents with fire engines,
ambulances, police



Research: Adjustable autonomy
Previous work: Often single agent-single human
interactions
Research: Scale-up in team coordination
Previous work: Limited numbers of agents coordinating in
teams
Visualization

Robust 3D visualization
Adjustable Autonomy:
Novel Challenges in Teams

Previous transfer-of-control fails in teams:


Ignore costs to team (just concerned about
individual)
One shot transfers of control, too rigid




Transfer control to a human (H) or agent (A)
If human fails to make a decision, miscoordination!!
Forcing agent to decide can cause a poor decision
Expensive lesson learned in the “Electric-Elves”
project


Major errors by software assistants
Hence need more flexible transfer of control
Predictions
EQh: Expected quality of human decision
AGH: How many agents human can control
A Strategy has constant slope



Low B, High EQh
Low B, Low EQh
80
Strategy value
Strategy value
80
60
40
20
60
40
20
0
0
2
3
4
5
6
7
8
9 10
Number of agents
A
H
ATH
2
3
4 5 6 7 8 9 10
Number of agents
A
H
ATH
CREATE Research Center


Center for Risk and Economic Analysis of
Terrorism Events
MANPAD Scenario




Large Scale Disaster
Limited Resources
First Response
Help incident commander control situation


Large Scale
Crime Scene

Simulator

Robocup Rescue


10 different Simulators
Multiple Agent Types
Team Level AA Model


How to select the strategy among many?
Key idea: Calculate expected utility of
different strategies


Traditional Expected Utility


Mathematical model of strategies
 EQ: Quality of an entity’s decision
 P: Probability of response of that entity
 W: Cost of miscoordination
Probability of response * decision quality
Integrate over time
Agents Per Fire
Subject B
4
3.5
3.5
Agents/Fire
4
3
2.5
3
2.5
2
3
4
5
6
7
8
9
Number of Agents
AH
ATH
10
11
2
3
4
5
6
7
8
Number of Agents
AH
Subject C
4
Agents/Fire
Agents/Fire
Subject A
3.5
3
2.5
2
3
4
5
6
7
8
9
Number of Agents
AH
ATH
10
11
9
ATH
10
11
LA City Fire Dept Exercise:
Fire Progression

Fire starts on
1st floor

Spreads to Attic
LAFD Exercise: Simulations by
People Playing Roles

LAFD officials simulate
fire progression and the
resource availability

Battalion Chief allocates
available resources to
tasks
Proxy Architecture



Abstracted Theories of Teamwork
(Machinetta)
Platform Independent
Modular Structure
Other
Proxies
Communication
RAP Interface
State
Coordination
Adjustable Autonomy
RAP
DEFACTO Movie
Objectives: Agent-based Simulation
Tools for Disaster Response


Improve training and decision making
Present


Future


Teach and evaluate LAFD response tactics
Agent/Robot disaster response
Key research questions in:


Multiagent coordination, Adjustable Autonomy
Visualization of multiagent systems