Student presentations in HRS lecture
As usual, at the end of the semester, the students of the classes demonstrated their project work. In the lecture "Humanoid Robotic Systems" they worked with the Nao robot. We congratulate the student groups, who succeeded this semester:
Team: The Fellowship of NAO
The objective of this project is to have NAO execute a set of challenges. First, the robot should ask the user what he wants him to do. The user should have the choice between “climb stairs”, “point the ball” and “kick the ball” behaviors. The robot should then navigate to its target and execute the expected behavior.
Team: Let’s Play Basket Ball NAO
The objective of this project is to implement a basic basketball game on NAO using a ping-pong ball. First, the robot should ask the user what he wants him to do. The user should have the choice between “shoot” and “dunk ” behaviors. In any case, the robot should first ask the user to give him a ball. Once the robot has the ball, it should look for the hoop autonomously (using an Aruco marker), then walk until the shooting (or dunking) distance and execute the desired behavior in a robust manner (i.e. selecting the proper motion in a dedicated library depending on the context).
The robot starts in a random initial pose and should reach a target location labeled by an Aruco marker. Obstacles (also identified with Aruco markers) are placed in its way. The arrangement of obstacles is considered as dynamic. The robot starts by defining a straight path from its current position to the target marker. Each obstacle has an area of influence. If an obstacle area of influence intersects with the robot collision area (approximated by a cylinder) along this path, the robot should react accordingly and “negotiate” the obstacle depending on its type. If the obstacle is small enough, the robot should step over it (only lateral spanning is possible with NAO). If the obstacle is too wide to be spanned but still small enough to be walked on, then the robot should step on it (up and down). Finally, if the obstacle it too bulky, then the robot should get around it. At the end of each maneuver, the robot should reacquire its target and recheck for obstacles. If the robot falls, it must be able to get up automatically and resume its operation.
Team: Teleoperating Nao
The objective of this project is to achieve a teleoperation task on NAO using a video interface. The user has feedback of what the robot sees on a fixed screen. For simplicity purpose, the robot is expected to respond to the user speech commands (e.g. start/stop and open/close hands). The user should wear a set of Aruco markers on its torso and arms. These markers will be tracked by a separate webcam. After a short calibration step, the robot should be able to replicate the user arm and chest movements. The torso marker should be used as a reference to track the user position in the (x,y) plane and then generate walking motions accordingly. There should be a smooth switch between walking and arm motion generation.