Pandu Raditya Rohman

Melbourne, Australia

• Assisted in simulating robotic systems using PyBullet for testing and validation.
• Developed robot control workflows using ROS 2 and GPT APIs to command the Unitree Go1 quadruped robot.
• Currently assisting in the development of the lab's Digital Twin with the use of LiDAR and depth camera sensors and a high level ROS2 control architecture of a bipedal robot.

Wuxi, Jiangsu, China

• Designed and developed a ROS2-based person-following navigation system, integrating perception, tracking, and motion planning for real-time operation.
• Integrated human tracking outputs into the navigation stack, enabling motion prediction, re-acquisition after occlusions, and smooth trajectory following.
• Implemented, integrated, and tuned global and local planners such as Smac2D and Timed Elastic Band (TEB) within ROS2 Nav2 to safely follow a moving human target in dynamic environments.
• Contributed to the ROS1 MoveBase to ROS2 Nav2 migration and assisted with on-robot field testing in real-world environments.

Melbourne, Australia

• Managed AI-driven medical imaging projects focused on Tuberculosis detection and lesion segmentation.
• Applied EfficientNet, ConvNet, U-Net, U-Net++, and Swin-UNet for classification, segmentation, and model deployment.

Melbourne, Australia

• Researched, implemented, and benchmarked medical image segmentation models (UNet, MSEG-HarDNet) for brain stroke segmentation.
• Researched and prepared datasets and dataloaders for Tuberculosis detection and segmentation task.

Australia

• Supported and maintained the association's existing web platform, and contributed to designing and developing a new website.

• Trained for and competed in multiple chemistry competitions as part of the school's chemistry olympiad team.