]> pere.pagekite.me Git - homepage.git/blob - mypapers/robocup2000/robocup2000.tex
Generated.
[homepage.git] / mypapers / robocup2000 / robocup2000.tex
1 \documentclass[a4, landscape, semrot]{seminar}
2
3 \usepackage{fancybox}
4 \usepackage{graphicx}
5 \usepackage{picinpar}
6
7 \slideframe{Oval}
8
9 \def\urlend#1{#1\endgroup}
10 \def\url{\begingroup \tt
11 \catcode`\_=13 % Don't know why this works.
12 \catcode`\~=11 \catcode`\#=11 \catcode`\^=11
13 \catcode`\$=11 \catcode`\&=11 \catcode`\%=11
14 \urlend}% \url for plain \TeX.
15
16 \newcommand{\heading}[1]{%
17 \begin{center}
18 \large\bf #1%
19 \end{center}
20 \vspace{1ex minus 1ex}}
21
22 \newpagestyle{MH}%
23 {UWA/CIIPS Seminar, Perth, 2000-10-04\hfill\thepage}%
24 {RoboCup 2000: CIIPS Glory goes Melbourne \hfill Petter Reinholdtsen}
25 \pagestyle{MH}
26
27 \rotateheaderstrue
28
29
30 \begin{document}
31
32 \begin{slide}
33
34 \begin{center}
35 \includegraphics[height=2cm]{soccer-logo-med}
36 \end{center}
37
38 \heading{RoboCup 2000: CIIPS Glory goes Melbourne}
39
40 \begin{center}
41 Petter Reinholdtsen $<$pere@hungry.com$>$
42 \end{center}
43
44 \begin{abstract}
45 This is the story of a team of robots, with two Norwegian and one
46 Australian coach, participating in the Robot Soccer World Cup.
47
48 A short multimedia summary of a trip to Melbourne.
49 \end{abstract}
50
51 \newslide
52
53
54 \heading{Content}
55
56 \begin{itemize}
57 \item The team
58 \item Our experience
59 \item The results
60 \item The local vision derby
61 \item Some highlights from the workshop
62 \item Some videos
63 \end{itemize}
64
65 \newslide
66
67 \heading{The team}
68
69 5 Soccer Eyebots.
70
71 Stephen Humble, volunteer. Programming field player, handling
72 mechanics and public relations.
73
74 Petter Reinholdtsen, volunteer and team leader. Programming field
75 player and operating system.
76
77 Mari Wang, volunteer. Programming goal keeper.
78
79 Jack Daniels, biped Eyebot. Public relations.
80
81 \newslide
82
83 \heading{Experience}
84
85 \begin{itemize}
86
87 \item Compasses worked perfectly, with less then 2 degrees error. (as
88 opposed to our lab). We integrated it into our player on site.
89
90 \item Overhead tracking system did not work at all. The lens was to
91 wide, and we did not have zoom. No one had been working on this
92 part since march, so everything where untested when we left for
93 Melbourne.
94
95 \item 24x7 access to the competition area a real must. We had a lot
96 of programming left before we where ready to compete.
97
98 \item No Internet-access was a pain. Forgot some headerfiles required
99 to compiel the overhead tracker, and had to get one of the
100 organizers to drive me to RMIT to download.
101
102 \end{itemize}
103 \newslide
104 \heading{Experience (cont...)}
105 \begin{itemize}
106
107 \item Some resets during game, the batteries holders are not good
108 enough.
109
110 \item Self localization using distance sensors failed due to badly
111 calibrated sensors and too long calculation delay (~10 second).
112
113 \item Always hard to get access to the 418 and 433 MHz radio
114 frequency. We had to make it easier to turn on and off radio.
115 We should try moving to different frequencies.
116
117 \item Omni-directional wheels is a big advantage.
118
119 \item The dribbler allowed real ball control.
120
121 \item Easy to get stuck on the angled walls, with no wheel touching
122 the ground.
123
124 \end{itemize}
125 \newslide
126 \heading{Experience (cont...)}
127 \begin{itemize}
128
129 \item The all-wood soccer field was very nice. Knee-high table
130 allowed to reach across without climbing. All-wood gave no compass
131 disturbances.
132
133 \item 4 Stooges had ~12 frames per second using the same hardware as
134 us. They could turn off interrupts and run the camera on full speed
135 as they do not use (or have) any other sensors.
136
137 \item Linux where used almost on more then half the computers in the
138 competition area. Many of the middle size league teams used Linux
139 on their robots as well.
140
141 \item The information flow where flowing somewhere else. Find the
142 league organizer and stay close to him.
143
144 \end{itemize}
145
146 \newslide
147
148 \heading{The results}
149
150 \begin{table}
151 \centering
152 \begin{tabular}{|l|l|c|}
153 \hline
154 CIIPS Glory & TPOTS & 0 - 3 \\
155 CIIPS Glory & Rogi Team & 0 - 10 \\
156 CIIPS Glory & FU Fighters & 0 - 7 \\
157 \hline
158 Viperoos & TUD & 2 - 0 \\
159 \hline
160 \end{tabular}
161 \end{table}
162
163 It could have been a lot worse.
164
165 TPOTS stayed up all night before their game with us to adjust their
166 software to stay further away from the opponents (us) to avoid
167 damaging our robots. The adjustments did not turn out too well.
168
169 FU Fighters turned off their kickers and slowed down their robots a
170 lot. So much the ball stopped on the goal line several times without
171 entering the goal.
172
173 \newslide
174
175 \heading{Local vision derby}
176
177 The local vision teams had a small derby Wednesday evening to compare
178 our performance. It was not too impressive. :-)
179
180 \begin{table}
181 \centering
182 \begin{tabular}{|l|l|c|}
183 \hline
184 CIIPS Glory & 4 Stooges & 0 - 0 \\
185 Viperroos & 4 Stooges & 0 - 0 \\
186 CIIPS Glory & Viperoos & 0 - 0 \\
187 \hline
188 \end{tabular}
189 \end{table}
190
191 We scored a goal in the penalty shootout against Viperoos.
192
193 A lot of people where interested in the local vision and local CPU
194 systems. A lot of teams wanted to start with ``real robotics'', as
195 one from FU Fighters phrased it.
196
197 I spoke to a Japanese professor with a local vision team with
198 omni-directional vision and off board computers. There where not
199 competing in RoboCup, but might come next year.
200
201 \newslide
202
203 \heading{Highlights from the workshop}
204
205 \begin{itemize}
206
207 \item Omni-directional vision using isometric mirror (F.M. Marchese,
208 D.G.Sorrenti).
209
210 Linear distance on the soccer field, full height view and detailed
211 view close to the robot.
212
213 \item Robust Real Time Color Tracking (M. Simon, S. Behnke, R. Rojas).
214
215 FU Fighters plan to release the source for their overhead tracking
216 system.
217
218 \end{itemize}
219
220 \newslide
221
222 \heading{Biped robots}
223
224 There where several biped robots on display. Some of them could even
225 walk.
226
227 \begin{itemize}
228
229 \item Jack Daniels from University of Western Australia
230
231 \item ? from De Laboratory Robotics de Paris, France
232
233 \item PINO from Kitano Symbiotic Systems Laboratory, Japan
234
235 \item MkII from Aoyama Universityis, Japan
236
237 \end{itemize}
238
239 \newslide
240
241 \heading{Proposed rule changes}
242
243 At the end of the last day, the small size league sat down to discuss
244 our experiences with the tournament. Some important rule changes
245 where discussed, and they will probably be implemented next year.
246
247 \begin{itemize}
248
249 \item The soccer field will probably be wider next year
250
251 \item The colored balls will be replaced by colored paper patches
252
253 \end{itemize}
254
255 \newslide
256
257 \heading{Videos}
258
259 \begin{itemize}
260
261 \item Robot helmet-cam by FU Fighters
262
263 \item Daily highlights by RoboCup 2000 webcast team
264 \begin{itemize}
265 \item Arrival Saturday 26th August
266 \item F180; Pre-competition Sunday 27th August
267 \item F2000; Naist vs Golem Tuesday 29th August
268 \end{itemize}
269
270 \item (Some random recordings by Petter Reinholdtsen)
271
272 \end{itemize}
273
274 \newslide
275
276 \heading{Thank you}
277
278 \begin{center}
279 \includegraphics[width=5cm]{soccer-logo-med}
280
281 \url{http://www.robocup.org/} and \url{http://www.robocup2000.org/}
282
283 \url{http://ciips.ee.uwa.edu.au/~pere/soccer/}
284
285 Slides are available from Petter Reinholdtsen $<$pere@hungry.com$>$,
286 \url{http://ciips.ee.uwa.edu.au/~pere/mypapers/robocup2000/}
287
288
289
290 \end{center}
291 \end{slide}
292 \end{document}