Postdoctoral Research Fellow @ KU Center for Autonomous Robotic Systems (KUCARS)
Khalifa University, Abu Dhabi, UAE π¦πͺ
π€ Robot Learning Β Β·Β π§ Task & Motion Planning Β Β·Β π Marine Robotics Β Β·Β ποΈ Robotic Vision
I work to bring robots out of the lab and make them capable of performing manipulation and navigation tasks robustly and intelligently β robots that don't hesitate to interact with the real world. My research bridges classical control and motion planning with modern learning-based methods.
- π Ph.D. in Automatic Control, Robotics & Computer Vision β Institute of Industrial and Control Engineering (IOC), Universitat PolitΓ¨cnica de Catalunya (UPC), Barcelona, Spain
- π Thesis: Physics-based Motion Planning for Grasping and Manipulation β knowledge-oriented physics-based planning among movable obstacles, grasping in clutter, and planning with temporal goals
- π°οΈ Former Project Manager for the MBZIRC Maritime Grand Challenge (heterogeneous USV + UAV + manipulator systems in GNSS-denied conditions)
- π’ Previously R&D Engineer at ITK System Engineering (BOSCH Group) β medical robotics, AR, and software testing
- π Visiting scholar at the Kavraki Lab (Rice University) and the Institute of Robotics and Mechatronics (DLR, German Aerospace Center)
- Vision-Language-Action (VLA) models for robot manipulation
- LLM / ontology-driven task and motion planning β refining prompts with knowledge-based reasoning for context-aware symbolic plans
- Autonomous marine navigation β environmental monitoring, maritime security, and port inspection
- Robotic perception β visual servoing, tracking, and underwater vision
| Repository | Description |
|---|---|
| VLAs | Vision-Language-Action models for robot learning and manipulation |
| llm-tamp | LLM-driven task and motion planning |
| llm-vlm-fusion-port-inspection | LLM + VLM fusion for autonomous port inspection |
| vision-tracking | Vision-based object tracking for robotic systems |
| visual-servoing | Visual servoing for manipulation and control |
| ros-docker | Containerized ROS development environments |
Real-robot experience: ABB YuMi Β· KUKA-LWR Β· PAL TIAGo Β· UR5 / UR10 Β· Franka FR3 Β· Bravo5 underwater manipulator
"I don't have any comfort zone β I can quickly adapt to new tools for tackling challenging problems."

