Muhammad Umar, an IFRoS Master’s graduate (2021-2023), is now working as a Robotics Software Engineer at Paltech GmbH in Germany. He is part of a team developing an MVP for an autonomous weed extraction robot, applying skills he gained during the IFRoS program. You can read his first-hand account below.
“I am currently a Robotics Software Engineer at Paltech GmbH in Germany, where I am developing the MVP for a weed extractor robot. My role involves implementing advanced navigation, localization, and perception solutions using ROS2.
The robot is equipped with a variety of sensors, including an IMU, RTK-GPS, a front camera, an internal camera, and a RealSense depth camera. For localization, we use odometry, RTK-GPS, and IMU data to achieve positioning accuracy within 10 cm. The path planning utilises regulated pure pursuit combined with Reeds-Shepp’s algorithms to ensure comprehensive field coverage.
The front camera, running a deep learning model, identifies weeds, allowing the robot to precisely position itself over the target plant. The GRBL controller then activates the motors to maneuver the internal camera above the plant, where another deep learning model identifies keypoints to locate the plant’s center. The mechanical tool inside the robot is designed to extract the plant using a drilling mechanism, and the RealSense depth camera ensures obstacle avoidance throughout the process. By integrating navigation, localization, and perception modules, we deliver an effective weed extraction solution for farmers.
During the IFRoS program, I worked on different projects related to each of the modules that I am currently developing. My Master’s studies have significantly enhanced my skills and knowledge, proving invaluable in my professional journey.”
Want to experience IFRoS opportunities like this? IFRoS program