Robotics & AI · Perception · Control · State Estimation
Portfolio · LinkedIn · Projects
I am an Arts et Métiers engineering graduate specializing in mechatronics, robotics and AI. I explore how robots perceive their environment, estimate their state and make decisions — from control theory and computer vision to reinforcement learning.
Incoming Ontology Engineer – Data & AI at SLB in January 2027, under a contract with Odyssey Sponsorship.
Robot trajectory following, target reaching and obstacle avoidance, combining reinforcement learning, ROS 2, vision and state estimation. The simple trajectory-following simulation reports 97% completion over 100 random trajectories. The real-robot implementation remains partial: end-to-end sim-to-real transfer is not yet validated.
A simulated robot navigates a corridor, follows a line and pushes a ball using LiDAR, camera and odometry within a unified ROS 2 state machine. Built with David Tovmassian. The repository documents an external course dependency needed to reproduce the Gazebo environment.
Co-developed an embedded control system for a motorized target robot, integrating an ESP32, LiDAR feedback, motor control, a wireless interface and custom electronics. Delivered with Tedjeddine Sebiane, in collaboration with JEECE and AMJE Arts et Métiers Junior Études.
A control engineering study of an unstable cart-pendulum system, with RST polynomial control, pole placement, disturbance rejection and MATLAB/Simulink analysis.
Three geometric vision projects covering optical-flow tracking and pose estimation, 3D point-cloud registration, and Hough-based pore segmentation. The repository explains the external data required for some examples.
Robot learning and control
- UR5e RL Pick-and-Place — grasp-and-place training with SAC and HER in ROS 2 and Gazebo.
- Twin-Rotor Helicopter Control — experimental identification, integral state feedback and LQR.
Engineering research and design
- VertiGaN Transistor Research — exploratory GaN device modeling and analysis; results require further validation.
- Compact Axial Turbine Design — a team project spanning hydraulic design, CAD, CFD and prototyping.
- Spatial Verification of a CMM — trajectory design and metrology, with a MATLAB correction pipeline.
Software applications
- ProcureLens AI — a desktop and command-line toolkit for auditable product-price extraction.
- The Game 2048 — an educational Python, Tkinter and SQLite implementation.
Robotics: ROS 2, robot navigation, motion control
AI and perception: reinforcement learning, computer vision, sensor fusion
Engineering: state estimation, system modeling, MATLAB/Simulink
Programming: Python, C++, MATLAB
Explore the projects and demonstrations on my portfolio.