2012 Simulated Car Racing Championship @ CIG-2012

The 2012 Simulated Car Racing
Championship @ CIG-2012
Daniele Loiacono, Luigi Cardamone, and Pier Luca Lanzi

2012 Simulated Car Racing Championship @ CIG-2012

2012 Simulated Car Racing Championship
9 races during 3 conferences

Develop a driver for TORCS
(hand-coded, learned, evolved, …)

Drivers will be awarded based on their score
in each conference competition

At the end, the team with highest overall score
wins the championship


Roadmap of the 2012 Championship

1.  EVO*-2012, Malaga (Spain)
April 11-13, 2012

2.  ACM GECCO-2012, Philadelphia (US)
July 7-11, 2012

3.  IEEE CIG-2012, Granada (Spain)
September 11-14, 2012


Motivations

q  Proposing a relevant game-based competition
" more representative of commercial games AI
" more similar to a real-world problem

q  Proposing a funny and exciting competition
" you can see and play with the entries of this competition
" human players can interact with AI
" a lot of programmed AI available for comparison

q  Proposing a challenging competition
" computationally expensive
" real-time
" on-line learning
" noisy sensors


The Open Racing Car Simulator
& the Competition Software

TORCS TORCS
PATCH

BOT BOT BOT SBOT SBOT SBOT

q  The competition server
UDP UDP UDP
q  Separates the bots from TORCS
q  Build a well-defined sensor model
q  Works in real-time
BOT BOT BOT


Sensors and actuators

q  Rangefinders for edges on the track and opponents (with noise)
q  Speed, RPM, fuel, damage, angle with track, distance race, position
on track, etc.

q  Six effectors: steering wheel [-1,+1], gas pedal [0, +1], brake
pedal [0,+1], gearbox {-1,0,1,2,3,4,5,6}, clutch [0,+1], focus
direction

What is the structure of a race?

Three stages: warm up, qualifiers, actual race

During warm-up, each driver can explore the
track and learn something useful

During qualifiers, each driver races alone against
the clock (the best 8 drivers move to the race)

During the race all the drivers race together


The competitors

q  Four entries in 2012:
" Ho Duc Thang, University of Nottingham
" BFB4, Sejong Univ, Korea
" Mr. Racer, TU Dortmund
" Smile Sheep, Taiwan National Chiao Tung University

q  Three reference entries (selected from the top entries of the
past championship) :
" Mariscal & Fernández, Málaga University
" Ready2Win, Slovak University of Technology – FIIT
" AUTOPIA, Madrid and Granada


Ho Duc Thang
University of Nottingham


Ho Duc Thang

q  Hybrid fuzzy controller with a module to learn the model
track as the car drives, a module to perform some simple
plannings based on the modelled segments of the track and
a fuzzy controller which actually drives the car.

q  All the modelling, planning and driving are done during the
race, the warm-up stage is not exploited

q  No opponent handling


BFB4

Kyung-Joong Kim
Jun-ho Seo
Tae-seong Kim
SejongCar Racing Championship @ CIG-2012
2012 Simulated Univ, Korea

BFB4

q  Scripted approach
" Simple approach for steering and accelerating

q  Jump Detection
" When the Z-Sensor is more than 0.4
" While the car is jumping and landing(few times, 10 tick
count), the steer of car is fixed to forward

q  Opponent avoidance:
" Just brake if the opponent is approached on a straight
" Modify the normal steering when the opponent is
approached in a bend


Mr Racer

Jan Quadflieg, Tim Delbruegger and Mike Preuss
TU Dortmund


Mr. Racer 2012

q  Joined effort of Jan Quadflieg, Tim Delbruegger and Mike
Preuss
q  Variables learned offline with the CMA-ES, now including
optimized settings for recovery , the clutch and…
q  NEW: Parameters of the opponent handling module have
been optimized
q  Optimization of the opponent handling will hopefuly result in
a CIG competition paper


Noise Handling

1.  Low pass filter on sensor
values
2.  2 Regression polynoms are
fitted to each side
3.  Resample the polygons
4.  Calculate our measure (see
CIG paper 2010)
5.  Low pass filter on the last
measurements


Online Learning during Warmup

q  We learn a model of the racetrack
q  One of two parameter sets is chosen
q  Target speeds for all corners are then fine tuned
q  All information is saved at the end of the warmup
q  And later used during qualifying and the race


Opponent Handling

q  Noise filtering with a lowpass filter
q  Observer module interprets filtered sensor readings
q  This leads to a recommended speed and an overtaking line
(created on the fly!)
q  Planning module incorporates recommended target speed
and overtaking line into the plan


Smile Sheep

吳晞浩(Si-Hao Wu)
李揚(Yang Lee)
李懷哲(Wiser Lee)
Advisor: 王才沛(Tsaipei Wang)
Taiwan National Chiao Tung University

Smile Sheep

q  Steering heuristics
" drive toward the track sensor with biggest distance
" small corection toward middle of the track to avoid
borders
q  Speed control heuristics
" Full accelerating when central sensor is clear for >99m
" Slow down when car is too fast
q  Recovery policy heuristics
" identify stuck condition with heuristics
" drive backward and get the car parallel to the track


Mariscal & Fernández

q  General approach: expert systems improved by multi-
objectives evolutionary computation techniques
" Maximize distance.
" Minimize damage

q  Steering based on a
simple rule follow the
track sensor with
biggest distance.

q  Overtaking based on
simple rules and expert
systems


Ready2Win

q  Modular architecture:
" Driving module:
•  The speed is computed using the maximum braking distance
" Overtaking module
" Recovery module
" ABS module

q  Track learning during the warm-up:
" First lap, driven slow, to identify turns (start, end, entry
position, curvature) and learn the track model
" Other laps, speed adaptation:
•  Increasing according to lateral speed
•  Reducing when the car goes off-road


AUTOPIA

Industrial Computer Science Department.
Centro de Automática y Robótica
Consejo Superior de Investigaciones Científicas
Madrid, Spain
Contact:E. Onieva (enrique.onieva@car.upm-csic.es)

AUTOPIA

q  Fuzzy Architecture based on three basic modules for gear,
steering and speed control
" optimized with a genetic algorithm

q  Learning in the warm-up stage:
" Maintain a vector with as many real values as tracklength
in meters.
" Vector initialized to 1.0
" If the vehicle goes out of the track or suffers damage
then multiply vector positions from 250 meters before the
current position by 0.95.

q  During the race the vector is multiplied by F to make the
driver more cautious in function of the damage:
" F=1-0.02*round(damage/1000)


Scoring process: Warm-up & Qualifying

q  Scoring process involves three tracks:
" IllSchwang (desert track)
" Noceda (city track)
" Sehnor (hill track)

q  The tracks are not distributed with TORCS:
" Generated using the Interactive Track Generator for
TORCS and Speed Dreams available at:
•  http://trackgen.pierlucalanzi.net
" The competitors cannot know the tracks

q  Each controller raced for 100000 game ticks in the warm-up
stage and then its performance is computed in the qualifying
stage as the distance covered within 10000 game ticks


Illschwang

Qualifying: Ilschwang

SmileSheep
7765.42

BFB4
6998.2

Mariscal
&

5066.24

Fernandez

Ready2Win
10010.1

Autopia
10701.6

Ho
Duc
Thang
10808.2

Mr.
Racer
10891.4

0
2000
4000
6000
8000
10000
12000


Noceda

Qualifying: Noceda

BFB4
9658.8

Mariscal
&

11265.2

Fernandez

Ready2Win
9405.52

AUTOPIA
11648.9

Ho
Duc
Thang
11738.82

Mr.
Racer
12233.2

0
2000
4000
6000
8000
10000
12000
14000


Sehnor

Qualifying: Sehnor

SmileSheep
8683.94

BFB4
9043.72

Mariscal
&
9610.96

Fernandez

Ready2Win
9413.6

Ho
9646.59

Autopia
10280.7

Mr.
Racer
10483

0
2000
4000
6000
8000
10000
12000


Qualifying summary (scores with noise)

Driver Illschwang Noceda Sehnor Total
Mr. Racer 10 10 10 30
Ho Duc Thang 8 8 6 22
AUTOPIA 6 6 8 20
Ready2Win 5 3 4 12
Mariscal&Fernandez 2 5 5 12
BFB4 3 4 3 10
SmileSheep 4 2 2 8

CI Approaches


What about the qualifying
results?

Since 2009 AUTOPIA is not the
fastest controller for the first time!

Top controllers clearly better than
others


Three Tracks

For each track we run 6 races
with different starting grids

Each race is scored using the F1 point system
(10 to first, 8 to second, 6 to third, …)

Two points to the controller with lesser damage

Two points for the fastest lap of the race


Final Results

Driver Illschwang Noceda Sehnor Total
Autopia 10 12 10 32
Mr. Racer 10 6 10 26
Ready2Win 8 3 6 17
SmileSheep 5 5 5 15
Mariscal&Fernandez 4 6 4 14
BFB4 3 5 4 12
Ho 2 2 2 6

Mr. Racer is winner of SCR@CIG 2012


What
about
the
race
results?

Qualifying
and
ﬁnal
results
are
similar:

the
only
excepFon
is
driver
Ho

Some
drivers
were
not
able
to
complete
the

race

Further
analysis
will
be
performed
to
have

more
insight
on
these
results!

Final
Standings
Championship

CompeGtor Evo* Gecco CIG Total
AUTOPIA
39
35
32
96

Mr.
Racer
16.5
18
26
60.5

Ready2Win
19
20
17
56

Mariscal
&
Fernandez
20
18,5
14
52.5

BFB4
-‐
11
12
23

Ho
Duc
Thang
-‐
13
6
19

Smile
Sheep
-‐
-‐
15
15

Final
Standings
Championship

CompeGtor Evo* Gecco CIG Total
AUTOPIA
39
35
32
96

Mr.
Racer
16.5
18
26
60.5

Ready2Win
19
Racer

Mr.
20 17
56

Mariscal
&
Fernandez

WINNER
of

012
S18,5

20 2 14

imulated
Car
52.5

BFB4
Racing
Championship
12

-‐
11
23

Ho
Duc
Thang
-‐
13
6
19

Smile
Sheep
-‐
-‐
15
15

We
wait

for

you…
…next
year!

www.geccocompetitions.com

info@geccocompetitions.c
om
@geccocomp
/geccocompetitions

2012 Simulated Car Racing Championship @ CIG-2012

Recommended

Recommended

More Related Content

Viewers also liked

Viewers also liked (20)

Similar to 2012 Simulated Car Racing Championship @ CIG-2012

Similar to 2012 Simulated Car Racing Championship @ CIG-2012 (20)

More from Daniele Loiacono

More from Daniele Loiacono (20)

Recently uploaded

Recently uploaded (20)

2012 Simulated Car Racing Championship @ CIG-2012