Robotik 1 - Kapitel 10 - Roboterprogrammierung und Programmierung durch Vormachen

In this talk

1. 1:28

   Why robot programming needs to be flexible

   The lecture connects flexible programming to varied products, small production runs, and service work with changing tasks. Robots should be programmable by the people who use them, including workers without robotics expertise.

2. 6:13

   Direct and text-based programming

   Direct methods include teaching points on the robot, recording a path by physically guiding its arm, controlling a slave robot through a master, and using sensors to define trajectories. Textual programming gives more control over complex programs, but requires users to learn robot-specific languages such as ABB RAPID or KUKA KRL.

3. 41:16

   Graphical programming with reusable statecharts

   Graphical interfaces let users assemble robot actions without writing code, while statecharts represent tasks as connected capabilities such as locating, grasping, and handing over an object. The AMAX framework adds data flow, reusable components, and success signals that can support replanning when an action fails.

4. 1:11:49

   Learning tasks from human demonstrations

   The research goal is for robots to interpret human demonstrations, build task models, and transfer those models to robots with different bodies and movement capabilities. The lecturer distinguishes this from simply replaying a recorded path: learning from demonstration should support generalization, potentially from only a few examples or one demonstration.