by John Anderson, Mark Evans
http://www.cs.umanitoba.ca/~andersj/gensim.ps
Add To MetaCart
Abstract:
Intelligent agents designed to perform in the real world should by definition be tested and evaluated in the real world. However, this is impossible in many situations: a lack of resources may rule out construction of a complete robotic environment, for example, or the desired domain may be physically inaccessible for testing. In such situations, the use of a simulation system to provide an environment in which to test and examine the intelligent system is necessitated. In the past, such systems have acquired a poor reputation within the AI community, mainly due to the sometimes grandiose claims of systems that are tested solely under simulated conditions. In this paper we explore the conditions under which simulation is justified, examine the inadequacies of currently available systems for the testing and examination of intelligent agents, and describe Gensim, a new system designed to address these inadequacies. Rather than providing a single, parameterized domain, Gensim provides a collection of facilities allowing users to design complete environments for examining and testing intelligent agents. The system also provides a clean interface, allowing widely differing types of agents to be studied. While some bias is unavoidable, these facilities are designed to be as widely applicable as possible.
Citations
|
366
|
Qualitative simulation
– Kuipers
- 1986
|
|
249
|
A temporal logic for reasoning about processes and plans
– McDermott
- 1982
|
|
165
|
What are plans for
– Agre, Chapman
- 1990
|
|
143
|
Introducing the Tileworld: Experimentally Evaluating Agent Architectures
– Pollack, Ringuette
- 1990
|
|
115
|
Trial by Fire: Understanding the Design Requirements for Agents in Complex Environments
– Cohen, Greenberg, et al.
- 1989
|
|
111
|
The dynamic structure of everyday life
– Agre
- 1988
|
|
96
|
Commitment and Effectiveness of Situated Agents
– Kinney, Georgeff
- 1991
|
|
78
|
Cognition and reality
– Neisser
- 1976
|
|
70
|
Sensation and Perception
– Goldstein
- 1999
|
|
54
|
Artificial intelligence meets natural stupidity
– McDermott
- 1976
|
|
41
|
MICE: A flexible testbed for intelligent coordination experiements
– Durfee, Montgomery
- 1989
|
|
41
|
Intelligence without robots (a reply to brooks
– Etzioni
- 1993
|
|
31
|
Cultural support for improvisation
– Agre, Horswill
- 1992
|
|
24
|
Learning to Coordinate Behaviours
– Maes, Brooks
- 1990
|
|
18
|
From Micro-Worlds to Knowledge Representation: AI at an Impasse
– Dreyfus
- 1981
|
|
16
|
Softbots as testbeds for machine learning
– Etzioni, Segal
- 1992
|
|
15
|
Integrating planning and acting in a case-based framework
– Hammond, Converse, et al.
- 1990
|
|
11
|
Achieving Flexible Autonomy in Multi-Agent Systems using Constraints
– Evans, Anderson, et al.
- 1992
|
|
11
|
Responding to Environmental Change
– Howe, Cohen
- 1990
|
|
8
|
Constraint-Directed Improvisation for Everyday Activities
– Anderson
- 1995
|
|
8
|
Mice users guide
– Montgomery, Lee, et al.
- 1992
|
|
5
|
Intelligent Agent Modelling for Natural Resource Management
– Anderson, Evans
- 1994
|
|
5
|
Quantification of Autonomy in Multi-Agent Systems
– Barker, Evans, et al.
- 1992
|
|
5
|
and Niklas Bertani. Ars Magna: The abstract robot simulator manual, version 1.0
– Engelson
- 1992
|
|
3
|
Supporting Flexible Autonomy in a Simulation Environment for Intelligent Agent Designs
– Anderson, Evans
- 1993
|
|
3
|
A Representation of Parallel Activity
– Lansky
- 1986
|
|
2
|
Values and Environmental Modelling", in
– Friedland
- 1977
|
|
2
|
Perception of Movement", in
– Graham
- 1965
|
|
2
|
Evaluating Planning through Simulation: An Example Using Phoenix
– Howe
|
|
2
|
A Framework for Representing Knowledge", in Haugeland
– Minsky
- 1981
|
