[TB3] TurtleBot3 AutoRace 2017 teaser #2

Hi again

As mentioned in previous topic (https://discourse.ros.org/t/tb3-turtlebot3-autorace-2017-teaser-1/2626), 15 teams brang their customized TurtleBots into the field of RBIZ Challenge 2017 TurtleBot3 AutoRace, RobotWorld (2017.09.13~09.16), South Korea, and showed their exciting skills and dreams of Autonomous Driving technology.



In the field, three out of 15 teams had succeeded to overcome all 4 missions in time. The robots have never used just a simple line tracing algorithm that can be easily seen in Black & White IR sensoring Line Tracer Competition, but used various recognition technology region from common vision recognition techniques to Advanced Machine Learning methods, such as Tensor Flow, CAFFE, and so on, which can match their database with many real artifacts, i.e. stop sign, parking sign, flags, etc.


The participants made Autonomy algorithms based on given LIDAR but also put a series of sensors which are not only USB WebCam but also Infra Red sensor, Ultrasonic sensor additionally on their TurtleBots to optimize their hardware to the algorithms they built.


The team ‘Real Rice Thief (aka rial-bop-doduk)’ wons 1st prize is a group from Hanyang University, a famous in Vehicle Autonomy fields in Korea, next to DongA University (2nd place), KwangUn University (3rd place). In 4th and 5th place, two teams from OROCA (Open RObotics Community Association), a biggest Robotics Community in Korea, took prizes by their brilliant ideas and full of enthusiasm.


This year, the TurtleBot3 AutoRace Challenge is just beginning, and the next competition is already under scheduling. Watch the races that those teams presented in front of over 1,000 audiences. See you guys on next year with your TurtleBots (maybe at ROSCon2018?).

(All devices for judging are also running by ROS, too!!)

To see previous clip, go https://discourse.ros.org/t/tb3-turtlebot3-autorace-2017-teaser-1/2626
Then, see you in next clip!

Oh, we will be in ROSCon2017 Vancouver, so come to us to talk more about.
Thank you!



This is amazing, i am very much interested to see how it has been done. I am also trying to do something like this using Camera but without LiDar