PC: unsplashed

Integrating vision into a collaborative robotic arm

In many applications robotic arms are used to transfer parts from one location to another with an end-effector (gripper) at the tip of the robotic arm. The geometry of the part is well known so the challenge is not to detect the part type but its pose. In addition, the actual pose of the part is uncertain only to a small extent since it is placed in a pallet that limits its deviation from a nominal pose (i.e., this is not the much studied bin picking problem).

The project entails selection of appropriate vision sensors for range and precision, integrating the vision module at the tip of the robotic arm, developing firmware for calibration and image processing for pose detection, and interfacing with ROS on a Real Time Linux based controller of the robotic arm.

This project at Systemantics aims to design and develop a working prototype of an integrated vision module.

Qualifications

Bachelor’s or Master’s degree student from the relevant field

Key skills

Familiarity and experience with any 3 of the following

– Vision sensors and modules

– Image processing

– ROS2 (Robot Operating System)

– Open CV

– C++/ Python

– Linux based development environment.