Project for Sensor and Control for Mechatronics System subject, using the Fetch platform to grasp objects
The Fetch system will start from an arbitrary initial position and use its RGB-D camera to navigate around and identify the table. Then, the platform's integrated mobile base will take control to reach a location where it can comfortably manipulate the target objects on the table. Since then, it will involve further visualisation and grasping actions as a collaboration between the camera and the robotic arm.
Steps breakdown:
- Identify a near-table tasking position from an arbitrary initial position using the camera
- Control the mobile base to reach the tasking position
- Identify the objects on the table using the camera
- Control the arm to pick and place the objects in sequence
Required Python 3.7 above, the Robotics Toolbox for Python (RTB-P), and some dependencies:
$ pip install rvc3python
$ pip install machinevision-toolbox-python
$ pip install ir-support
Execute main_file.py
Video for what is expected to happen: https://www.youtube.com/watch?v=hcLFaT_tkqg