Unified Project Topics 2009.1
Jacques Robin
O ntologies
R easoning
C omponents
A gents
S imulations
Application 1: Fully Customizable Games
Innovative software process for fast prototyping fully customizable games where the user changes the field, the rules etc.
Allows use of focus groups too define most exciting rule sets for games based on imaginary, science-fiction sports
Allows the Web2.0 evolutions of game variations and community based emergence of the most popular sets of rules
Application 2: Workbench for Controlled Experiments to Comparatively
Evaluate Multi-Agent Systems:
1. MAS architecture
2. Communication, cooperation and competition protocols
3. Internal architecture of each agent
4. Automated reasoning task of each component of a given internal architecture
5. Knowledge representation language of each component of a given internal architecture
6. Agent team strategies
7. Software process
Full customization of environment characteristics (observability, determinism, dynamicity, continuity, etc.) for a given set of rules
Application 3: Tutorial System for Agents and Automated Reasoning
Fully customizable environment allowing students to build agents and test agents and agent teams for most possible combinations of environment characteristics
Link to Wiki pages with didactic material about societal and agent internal architectures, automated reasoning tasks and knowledge representation languages
Built-in societal and agent internal architectural patterns to be selected from a menu
Built-in component-based object-oriented rule-based knowledge representation language (COFL) and associated inference engine
(COFLUX)
Students introduced to gradually more sophisticated:
Environments, all different versions of a single, unifying domain (Rollerslam sport simulation)
Societal and agent internal architectures
Automated reasoning tasks and knowledge representation language
Soccer
Gaelic
Football
Australian
Football
Handball Rugby
Basketball
American
Football
Slamball
Ice
Hockey
Roller
Derby
Trampoline
Track Zone Division
24m in-track
In-Tramp 45m
64,25m
Bounce
Circle
Field
Center
Circle
78m
Out-Tramp
Goal out-track
Ramp
16m
Edge
Edge Wall
(Ramp) Edge
Ramp
Goal
Basket
Field
Inner Trampoline
Goal Line
Edge Shot Square
Outer Trampoline
Offensive
Edge Shot Circle
Defensive
Edge Shot Circle
7m
10m
7m
2.5m
4m
5m
3m
6m
1.25m
1.5m
3m
7m
Low Goal:
Baskets:
Touched close hand: 1 pt
Clear close hand, touched close foot, touched far hand, touched close hand volley: 2 pts
Clear close foot, clear far hand, clear close hand volley: 4 pts
Try: 7 pts
Clear far foot, clear close foot volley: 8 pts
Clear far foot volley: 16 pts
Front tramp hand hoop, simple dunk: 4 pts
Front ramp hand hoop, front alley dunk: 6 pts
Back tramp hand hoop, far dunk: 8 pts
Back ramp hand hoop, back alley dunk, front far hand hoop, front tramp foot hoop: 12 pts
Front ramp foot hoop: 18 pts
Back far hand hoop, back tramp foot hoop: 24 pts
Front track hand hoop: 36 pts
Back track hand hoop, front track foot hoop: 72 pts
Back track foot hoop: 144 pts
High Goal:
Close hand high volley: 1 pt
Close hand high goal: 2 pts
Close foot high volley: 3 pts
Close foot high goal: 4 pts
Far foot high volley: 6 pts
Far high goal: 16 pts
Laps:
1 passing lap: 16 pts
N consecutive passing laps: (N+1)*16 pts
1 rushing lap: 32 pts
N consecutive rushing laps: (N+1)*32 pts
LT
11
BT
1
LB
3
GK
0
RB
2
FB
6
CB
4
FB
5
HB
8
RV
13
QB
9
HF
10
RV
12
RT
7
LF
16
FF
18
CF
15
FF
17
RF
14
FT
19
IC
23
IF
24
IB
20
IT
21
IR
22
action(s,legs,shoot(2)) action(k,legs,move(right))
3
2
1
Y
X
1 k s, b
2 3 action(s,legs,shoot(3)) action(k,legs,move(right))
3
2
1
Y
X
3
2
1
Y
X
1
1 s
2 b s
2 k
3 k, b
3 actions(k,legs,move(right))
3
2
1
Y
X gameOver, goal b k s
1 2 3 action(k,hands,grab(yes))
3
2
1
Y
X
1 s
2 k, b gameOver, nogoal
3
1. Agent-Oriented KobrA2 Modeling Framework (Eric et al.)
a) Specify Object-Oriented Fluent Calculus in UML/OCL (OOFC, Eric),
10/06
b) KobrA2 PIM of Internal Agent Architecture (Eric, Ramon, Marcellus),
17/06
c) KobrA2 Multi-Agent Simulation PIM Framework (KMAPF, Eric), 24/06
d) Java (?) Implementation of MAS PIM Framework (Eric, Peterson), 08/07
2. Agent-Oriented Implementation Platform (Cleyton)
a) Port CHORD to CHROME, 10/06
b) Implement OOFC in CHORD, 17/06
c) Implement Agent PIM Framework in CHORD (OOFC-based belief revision
+ CHORD based action strategy), 08/07
3. 3D Animation KobrA2 Modeling Framework (Peterson)
a) KobrA2 PIM of Java 3D e Direct 3D (others free platforms better candidates?) 17/06
4. PenaltySim PIM as Instance of KMAPF (Eric, Peterson) 01/07
5. PenaltySim GamePhysics Agent PIM (Eric) 01/07
6. PenaltySim Player Agents PIM (Cleyton) 01/07
7. PenaltySimPlayer 2D Animation Agent PIM (Peterson) 01/07
8. PenaltySim GamePhysics Agent Implementation (Eric) 15/07
9. PenaltySim Player Agents Implementation (Cleyton) 15/07
10. PenaltySimPlayer 2D Animation Agent Implementation (Peterson)
15/07
11. Rollerslam PIM as Instance of KMAPF (Eric, Peterson) 05/08
12. Rollerslam GamePhysics and Referee Agents PIM (Eric) 05/08
13. Rollerslam Player Agents PIM (Cleyton) 05/08
14. Rollerslam 2D Animation Agent PIM (Peterson) 05/08
15. Rollerslam GamePhysics Agent Implementation (Eric)
16. Rollerslam Player Agents Implementation (Eric)
17. Rollerslam 2D Animation Agent Implementation (Peterson)