Robotics and AI engineer building marine robots and AI for infrastructure.
My current work sits in two places: on and under the water, where I build hydrofoil craft and underwater vehicles, and in applied AI, where I'm developing a foundation model for infrastructure and energy construction.
I have contributed to open-source robotics software since 2014, including Gazebo, MoveIt and BlenderProc, and wrote my master's thesis at the German Aerospace Center (DLR).
The goal is a vehicle that can travel long distances, navigate to depth and examine underwater infrastructure without a ship standing by. The work brings together marine engineering and robotics: vehicle design, underwater navigation, and the autonomy needed to operate far from an operator.
Status
Research
An electric hydrofoil board that stabilises itself in flight.
Onboard flight control adjusts the hydrofoil to keep the board stable, rather than leaving balance entirely to the rider. The work covers hydrofoil hardware, embedded flight-stabilisation control, and state estimation from GPS, IMU, depth and velocity sensing. Three prototypes have been built and tested.
Status
Prototyping and field testing
A foundation and world model for infrastructure and energy construction.
It combines large language models, retrieval-augmented generation, graph RAG and convolutional vision models into a single system in which the components work together, including computer vision applied to construction sites and projects.
Status
In development
Contributing since 2014
Consultant, intern and Google Summer of Code participant
Worked on open-source robot simulation tooling.
Master's thesis (TUM, 2021)
“Indoor Scene Synthesis: Generation of Interior Designs with Transformer Models”. A transformer-based generative model that produces complete 3D indoor scenes — object categories, positions and orientations — conditioned on the shape and size of a room.
M.Sc. Robotics, Cognition, Intelligence
Technical University of Munich (TUM)
B.Sc. Computer Engineering
Cairo University