Hello! I’ve spent 1 month in Eindhoven meaning I’m long overdue for an update on my time here.
The most recent project in my Udacity nanodegree gave me a 6 jointed robot arm and a desired path for the robot's gripper. My job was to perform the inverse kinematic analysis to calculate the joint positions that will move the gripper along the requested path. After learning the forward kinematics and doing a full geometric analysis for the inverse kinematics, my robot arm was able to follow the paths it received with negligible error.
This project focused on building an image perception pipeline for an RBGD camera. A table with assorted objects was presented in Gazebo simulation space, and after cleaning and segmenting the data I built an SVM perception model to identify each object. My perception pipeline accurately identified all items in each of the 3 provided scenes and packaged the information for a robot to pick up and place each object in a predefined bin.
