Robotour Marathon 2019
testing challenge II.
Robotour Marathon is a new discipline helping teams to prepare to hard Robotour contest held typically in September. The first marathon confirmed that the best 3 teams already participated and tested their robots in complex environments. Note, that there is a new monthly scoring scheme for year 2019. (Update 12/9: ARBot 1272m)
Rules
The rules for Robotour Marathon 2019 are similar to Robotour itself. The
robot has to demonstrate its ability to autonomously navigate on paved roads
without any human intervention.
It can be any road mapped in OpenStreetMap.
The main difference is that it is your challenge, performed when it
suits to you and where it suits to you. Well, you are also the judge,
which can be a bit awkward, but … pick a place, run your Robotour
code and share your video with the robot GPS track. The winner is robot
autonomously driving the longest distances over the year.
The Score Board(s)
The Longest Run (whole season)
| Place | Team | Distance | Date | Video | GPS Track |
|---|---|---|---|---|---|
| 1st | Istrobotics | 2599m | 2018-11-05 | QXu989FT2-g | #2836816 |
| 2nd | RoboTech Vision | 1361m | 2019-04-15 | 8v9Xis4DyxI | #2962510 |
| 3rd | ARBot | 1272m | 2019-09-12 | PpquE-zuhds | #3099537 |
| 4th | JECC Lois | 983m | 2019-08-11 | vES1Sr6H7ls | #3072241 |
| 5th | Cogito | 539m | 2018-09-30 | m5jnlGOBEOQ | #2813858 |
| 6th | Cogito - Emily | 427m | 2018-11-25 | CLmSgrhAPzE | #2849964 |
| 7th | JECC Fredl | 198m | 2019-05-31 | sM7GrZbH1Ug | #3001928 |
| 8th | Short Circuits Prague | 172m | 2018-10-31 | yaLrqkdnxCs | #2834268 |
| 9th | Eduro | 148m | 2019-04-07 | IF_OwL6mO9U | #2955889 |
| 10th | Robosoft | 84m | 2018-09-18 | w-jFvdBxycs | #2805962 |
The Monthly Challenge
September 2018
| Place | Team | Distance | Date | Video | GPS Track | Points |
|---|---|---|---|---|---|---|
| 1st | Cogito | 539m | 2018-09-30 | m5jnlGOBEOQ | #2813858 | 2 |
| 2nd | Robosoft | 84m | 2018-09-18 | w-jFvdBxycs | #2805962 | 1 |
October 2018
| Place | Team | Distance | Date | Video | GPS Track | Points |
|---|---|---|---|---|---|---|
| 1st | Short Circuits Prague | 172m | 2018-10-31 | yaLrqkdnxCs | #2834268 | 3 |
| 2nd | Cogito Emily | 117m | 2018-10-28 | 8JWoIwmr0RE | #2831929 | 2 |
| 3rd | Eduro | 64m | 2018-10-23 | 6yNynVL6FR4 | #2828667 | 1 |
November 2018
| Place | Team | Distance | Date | Video | GPS Track | Points |
|---|---|---|---|---|---|---|
| 1st | Istrobotics | 2599m | 2018-11-05 | QXu989FT2-g | #2836816 | 4 |
| 2nd | Cogito - Emily | 427m | 2018-11-25 | CLmSgrhAPzE | #2849964 | 3 |
| 3rd | Short Circuits Prague | 77m | 2018-11-27 | BRTaGau3c20 | #2851237 | 2 |
| 4th | Eduro | 15m | 2018-11-21 | sMaFBbhl9l8 | #2846981 | 1 |
December 2018
| Place | Team | Distance | Date | Video | GPS Track | Points |
|---|---|---|---|---|---|---|
| 1st | Short Circuits Prague | 24m | 2018-12-30 | jM1t4QRpTI0 | #2879107 | 1 |
March 2019
| Place | Team | Distance | Date | Video | GPS Track | Points |
|---|---|---|---|---|---|---|
| 1st | Istrobotics | 377m | 2019-03-19 | 3yO-3OSJJ2s | #2940585 | 2 | 2nd | Cogito - Emily | 257m | 2019-03-03 | WEHN1sXMDoE | #2928191 | 1 |
April 2019
| Place | Team | Distance | Date | Video | GPS Track | Points |
|---|---|---|---|---|---|---|
| 1st | RoboTech Vision | 1361m | 2019-04-15 | 8v9Xis4DyxI | #2962510 | 4 | 2nd | Istrobotics | 314m | 2019-04-14 | uPyekyr4USw | #2961766 | 3 | 3rd | Eduro | 148m | 2019-04-07 | IF_OwL6mO9U | #2955889 | 2 |
| 4th | JECC Fredl | 8m | 2019-04-30 | E2Ok5WUJ_pk | #2974173 | 1 |
May 2019
| Place | Team | Distance | Date | Video | GPS Track | Points |
|---|---|---|---|---|---|---|
| 1st | JECC Fredl | 198m | 2019-05-31 | sM7GrZbH1Ug | #3001928 | 1 |
June 2019
| Place | Team | Distance | Date | Video | GPS Track | Points |
|---|---|---|---|---|---|---|
| 1st | Istrobotics | 1163m | 2019-06-26 | IAOtIqYeh-8 | #3029168 | 1 |
July 2019
| Place | Team | Distance | Date | Video | GPS Track | Points |
|---|---|---|---|---|---|---|
| 1st | Istrobotics | 537m | 2019-07-31 | CJi8yZh9GWc | #3062036 | 1 |
August 2019
| Place | Team | Distance | Date | Video | GPS Track | Points |
|---|---|---|---|---|---|---|
| 1st | JECC Lois | 983m | 2019-08-11 | vES1Sr6H7ls | #3072241 | 1 |
September 2019
| Place | Team | Distance | Date | Video | GPS Track | Points |
|---|---|---|---|---|---|---|
| 1st | ARBot | 1272m | 2019-09-12 | PpquE-zuhds | #3099537 | 1 |
Questions and Answers
When the Marathon contest starts?
The contests starts now (September 2018) and it ends a day before the next
Robotour contest (September 2019).
Are multiple tries/submissions allowed?
Yes, definitely! The better one is kept in the score table (the old ones
are stored in backlog).
Is beer keg required in Marathon?
No, it is not. Nevertheless it is highly recommended to test with your payload
once a while.
Is there any time limit?
No, in this contest there is no time limit. The robot,
which manages to move further away, is better than the faster robot.
Video should be from robot or from mobile phone?
Both is accepted as robots do not have to be equipped with camera. You can
publish both and add the second URL in the comment. You can also speed up the
video.
Is there any naming convention for uploaded video?
It is highly recommended to use Robotour Marathon together with the name of
your team (for example Robosoft: Robotour Marathon 2019, attempt 0)
I submitted my video and GPS track but the form just gray out?!
There could be some issues with the form, so if you do not receive confirmation
within 24 hours, please send also mail to webmaster-at-robotika.cz. I am
sorry for any inconvenience caused.
Can I pause my robot in dangerous situations and then continue?
Yes, you can exceptionally pause your robot. The example of dangerous situation
is approaching horses, which could be scared by your robot, or a car on the
road where you do not expect it.
What is Monthly Scoring Scheme?
By the end of every month the order defines number of points the team gets, and
the next month starts with empty table. For example, if you have the longest
try in October 2019 and there were 5 other competitors you get 5+1 points. Even
for the last position in the given month you get 1 point.
Submissions/Backlog
9/18/2018 12:02 PM
- Robosoft: Robotour Marathon 2018/19, attempt 0, GPS log, 20180918_prvy_test_Robosoft
9/30/2018 7:41 PM
- Cogito: Robotour 2018, 1st round, GPS log, Robotour 2018, 1st round
10/23/2018 10:24 PM
- Eduro: Robotour Marathon 2019, attempt 0, GPS log, It was dark, cold and windy - what a crazy idea to take Eduro with SICK Robot Day 2018 setup, attach GPS and test it outside?! Speed 0.2m/s was incredibly slow, 0.8m/s reseted motors and stopped, so 0.6m/s was used. Laser is still mounted low to detect "storage box" 17cm high, so we rather stop it on the curb. Note, that log is available here: http://osgar.robotika.cz/eduro-181023_200112.log
10/28/2018 10:16 PM
- Cogito: Emily driving autonomously outdoors for the first time, GPS log, October, month of odometry.
10/31/2018 10:31 PM
11/5/2018 12:10 AM
Info from Istrobotics (via mail, in Slovak):
Prvy pokus:
Startovali sme pri parkovisku medzi N2 a N3. Ciel sme nastavili na S1. Pouzili
sme stare NiMH baterie a robot mal pomalu rychlost. Po cca 1km sme vysli z
cesty, kvoli napadanemu listiu. Zaznam je na:
https://www.openstreetmap.org/user/pablosksk/traces/2836815
Druhy pokus:
Kedze baterky boli takmer vymlatene, tak sme ich vymenili za LiPo, zvysili
rychlost a nastavili sme ciel na N1. Robot isiel pekne a my za nim rychlou
chodzou. Pod prvym mostom to boli riadne nervy, lebo robot isiel po pravej
strane a bali sme sa aby nespadol z toho velkeho zrazu Ked presiel bodom
N1kou, tak uz nemal mapu a ani ciel. Vtedy ide rovno a vyhyba sa trave a
prekazkam. Pod druhym mostom ho takyto styl navigacie zaviedol do rohu, z
ktoreho uz nemohol ist dalej a pokus sme ukoncili. Tu trasu sme potom dali ako
pokus do marathonu. Zaznam je na:
https://www.openstreetmap.org/user/pablosksk/traces/2836816
Treti pokus:
Sme len obratili robota a poslali k parkovisku na tych istych LiPo baterkach.
Zaznam je na: https://www.openstreetmap.org/user/pablosksk/traces/2836822
Ked zratam druhy a treti pokus, tak na tych LiPo baterkach by sme mohli
atakovat Udov minulorocny rekord. Ale interne sme si hovorili, ze by sme mohli
skusit 10km
11/21/2018 6:51 AM
This entry is rather to encourage other teams (as it is very hard to
beat Istrobotics now) and that they will surely do better than us. This time we
tested GPS and IMU and sure enough "something is wrong" whenever you
reach waypoint (taken from OSM map). The robot should just pause, but
the motors were reseted and robot got crazy. The log file is at
http://osgar.robotika.cz/gpsnav-181120_215347.log and code in
feature/magellan branch. Hopefully next Tuesday it will be better. p.s.
yes, it was cold and snowing shortly before the trial
11/25/2018 7:29 PM
Bearnav-style trajectory replay.
11/27/2018 11:24 PM
12/30/2018 11:16 PM
Moje historicky první nahrávka s ROS. Robot jede rovně zatím bez
jakéhokoliv řízení. Cílem bylo zatím ověřit základní spouštění,
nahravání logu a exportování GPX a videa z logu.
3/3/2019 9:28 PM
"
New stereo camera, new computer, new electromagnetic interference giving
GPS hard time :-/
"
3/19/2019 12:41 AM
Ano uz sme prebudili nasho robota zo zimneho spanku Testovali sme aj
trasu na Istrobot, ale robota nam tahalo do prava. Mal
problemy sa vyskrabat na chodnik. Tak sme ho radsej stopli:
4/7/2019 10:16 PM
This is Eduro version 0 for DARPA SubT Challenge in Idaho Springs, CO,
USA tested outside. The robot is supposed to navigate along left wall
for 4 minutes and return to the Base Station.
4/14/2019 2:54 PM
4/15/2019 1:34 PM
4/30/2019 9:13 PM
5/31/2019 1:28 AM
6/25/2019 5:38 AM
p.s. notice Tatra truck at 68s and the
fact that robot reached the goal and returned
7/31/2019 8:31 PM
8/11/2019 10:06 PM
This is the first Robotour Marathon attempt for this season from already
well known robot Lois. The robot is using a conventional image path
recognition algorithm realized in Opencv this year, so no neural
network. The reason therefore is achieving a higher framerate and lower
power consumption. The navigation is based on Ublox-8 gps, osm data and
the BNO055 imu, but maybe there will be a very reliable KVH C100 compass
unit at the Robotour 2019 contest. There were some critical situations
with the ditch that could be handled with the data from a SICK lidar
very close ( i almost had to push the stop button at 2.30 min... ). The
attempt ended because of empty batteries on a steep hill :-/
9/12/2019 8:15 AM
If you would like to somehow support this contest or you have some
comments/question, please use our standard contact form.