Control systems
Modelling, identification and controller design, from PID and state feedback to predictive and nonlinear control, then deployed on a real target.
Simulink · EKF · MPC · STM32
Automation and industrial computing engineer in the making. I work on control systems, robots and industrial IoT, and I design the whole system behind them: the board, the logic, the firmware and the software.
Tune the PID. The loop is simulated live in your browser.
Tomorrow's machines will sense, decide and act on their own. I want to build them whole: from the copper trace to the control law, from the robot's joint to the cloud that watches over it.
Three fields I work in, and one habit that ties them together: designing the hardware and the software as a single system.
Modelling, identification and controller design, from PID and state feedback to predictive and nonlinear control, then deployed on a real target.
Simulink · EKF · MPC · STM32
Mobile robots and manipulators: perception, tracking and motion under ROS 2, proven in Gazebo before the field.
ROS 2 · Gazebo · YOLOv8 · Docker
Sensor networks and supervision chains that survive power cuts and lost links, from the node to the dashboard.
MQTT · InfluxDB · Grafana · ESP32
The whole machine, not one layer: PCB and power stage, FPGA logic, firmware, and the software above it.
KiCad · VHDL · C / C++ · Python
Hands-on by default. I prefer work that leaves the simulator and runs on a board, a bench or a robot.
I'm a third-year electrical engineering student at ENIT (École Nationale d'Ingénieurs de Tunis), in the Automation & Industrial Computing option: control systems, robotics and industrial IoT.
I'm equally at home designing the system around the controller: routing a board, writing VHDL for an FPGA, scheduling tasks on an RTOS, and building the software and dashboards on top. That range lets me take a machine from a block diagram to a working prototype.
At ACTIA Engineering Services I built a battery digital twin fusing an Extended Kalman Filter with XGBoost to estimate state of charge, deployed and validated on ARM64.
Six-month research internship starting March 2027.
Thevenin ECM identification, EKF in Simulink with C code generation for ARM64, LSTM / BiLSTM / XGBoost comparison, EKF + XGBoost fusion and a real-time monitoring dashboard on Raspberry Pi 5.
First industrial immersion in power generation and distribution operations.
Ranked in the top 100 of about 800 candidates at the national entrance exam.
NativeB2B2Each trace is drawn from the kind of signal the project deals with. Open a card for its datasheet.