Samet Kocak

Development of an autonomous vehicle (1:10 scale) for a multi-vehicle control R&D project. Embedded software in C on STM32 (ultrasonic sensors, obstacle detection, PWM). Implementation of embedded/high-level communication. Development of a SLAM localization system on Raspberry Pi 4 (LiDAR, IMU, ROS2) with data fusion. Experimental validation of perception and localization. Result: Autonomous perception and localization successfully validated. Key skills: C, STM32, ROS2, Raspberry Pi 4, SLAM, LiDAR, IMU, ultrasonic sensors, embedded systems, data fusion, mobile robotics, WSL

Development of an autonomous landing system for drones (Copter, VTOL): analysis of firmwares (ArduPlane, ArduCopter), Gazebo simulation (SITL), MAVLink communication (UDP/TCP). Python scripts for GNSS, UWB, LIDAR data fusion, Lua control logic, test automation. Result: solution validated in simulation and currently being integrated on real platforms. Key skills: ArduPilot, MAVLink, Gazebo, SITL, Scripts, Lua, Python, C, C++, Bash, Data fusion, Sensors, Kalman filters, WSL

Development and optimization of embedded Deep Learning on STM32 under Zephyr RTOS with TensorFlow Lite, Keras, Scikit-Learn. Deployment in C/C++, optimization by quantization and pruning, automated performance analysis (Python, Bash). Result: prediction models achieving 95% accuracy, validated. Key skills: STM32, Zephyr RTOS, TensorFlow Lite, Keras, Scikit-Learn, Embedded Deep Learning, Python, C, C++

Development of a Python experiment integrating an eye-tracking system (EyeLink 1000 Plus) for annotating machine translation errors on ACCOLÉ. Processing and visualization of eye data with NumPy, Pandas, and PsychoPy. Result: experiment validated with 15 linguists. Key skills: Python, PsychoPy, Pandas, EyeLink 1000 Plus