ElSayed ElSheikh

I worked on creating a digital twin of an intelligent conveyor warehouse in NVIDIA Isaac Sim integrated with ROS2 to test and validate automation workflows in a realisitc environment. I led the development of a computer vision pipeline using (YOLO + SORT), trained on 30,000 synthetic images to achieve accurate real-time detection and tracking of packages/trays on the conveyor. I had also worked on implementing ROS2 based reactive controller for omni-conveyor units and designed a PDDL-based planner for system-wide orchestration.

I worked on developing simulation setup for Software-In-The-Loop testing. Using NVIDIA Isaac Sim, I built rich simulation scenes covering both urban and suburban environments with realistic assets, including Omniverse PhysX Vehicles like sedan and SUV cars. I also played a key role on improving the simulation framework for aerial vehicles built on top of Isaac Sim as an extension, allowing us to mimic multirotor dynamics with support for PX4 and Ardupilot. This framework is flexible and ready to handle future aerial vehicle types. I brought in solid software development practices using trunk based development Git workflow and clean architecture using ROS2, raising the bar for our autonomous vehicle testing. Docker was used for clean, reproducible work on a GCP VM, with Trello keeping the team organized and aligned throughout the project.