TechX Challenge Singapore 2008

Autonomous operating an elevator (~ 8MB) [xvid]
First the robot generates a 3D model of the environment by rolling the Hokyo LRF (orange). Second, the robot plans a collision free trajectory. Third, the robot moves the arm into location close to the panel. Fifth, the robot presses the button by visual servoing. All known in advance was the image of the elevator panel and the number to press.

Elrob 2008

Autonomous camp surveillance (~ 18MB) [flash] [mpg] [wmv]
The video shows a robot autonomously entering an unknown area (camp) for searching for intruders. The robot travels to surveillance points provided by the human in advance. At each point it extends its manipulator in order to get a better view of the surroundings. Then, a surveillance software processing the video stream from all three cameras of the robot, is executed. If a human has been detected, the software automatically stores the GPS location and an image of the detection on the hard disk. Finally, the robot returns to its starting position for delivering its observations.

TechX Challenge Singapore Qualification

Autonomous navigation (~ 8MB) [wmv] [mpeg2] [xvid]
During the navigating challenge the robot had to autonomously overcome a obstacle field of 30 meters.

Autonomous stair climbing side view (~ 6MB) [wmv] [mpeg] [xvid]
Autonomous stair climbing back view (~ 9MB) [wmv] [mpeg] [xvid]
During the stair challenge the robot had to autonomously climb-up a stair.

Autonomous elevator operation during the competition (~ 27MB) [wmv] [mpeg2] [xvid]

Autonomous elevator operation with a moc-up elevator (~ 32MB) [wmv] [mpeg2] [xvid]
During the elevator challenge the robot had to autonomously press the elevator panel inside the elevator. (1) The DTSA official starts the robot with the estop. (2) The robot goes into elevator flipper position. (3) The robot approaches the back wall of the elevator while adjusting its orientation. (4) The manipulator starts to measure the elevator interior with the laser range finder mounted on the gripper. (5) The planner finds a safe position and trajectory to move the arm towards. (6) The gripper takes another 3D scan for determining the distance of the panel. (7) The vision module identifies the button to press. (8) The gripper approaches the panel by visual servoing.