const alex = {
role: "Robotics MSc Student @ TU Delft",
location: "Delft, NL 🇳🇱 | Antwerp, BE 🇧🇪",
education: {
masters: "MSc Robotics @ TU Delft (2025–Present)",
bachelors: "BSc Mechanical Engineering @ TU Delft (2021–2025)",
minor: "Offshore Wind Energy @ Aerospace Engineering Faculty"
},
interests: ["Autonomy", "Computer Vision", "Machine Learning", "Planning & Decision Making"],
languages: ["Dutch 🇳🇱", "English 🇬🇧", "French 🇫🇷"],
hobbies: ["Skiing ⛷️", "Tennis 🎾", "Piano 🎹", ]
};| Project | What | Stack | |
|---|---|---|---|
| 🚁 | Autonomous MAV Flight (paparazzi) | Full-Pipeline Integration. Developed gate detection from monocular camera feeds and a sensor-fusion-based action controller. Deployed in a real-world test environment. | Python · C++ · Computer Vision · Control Theory · SIM-TO-REAL |
| 💻 | Autonomous Quadcopter Navigation | System Architect. Designed a dynamic simulation environment and integrated the full software stack. Implemented path planning (RRT*) and a centralized controller for autonomous flight. | Python · PyBullet · RRT* · MPPI |
| 🎨 | Evolutionary Van Gogh | Genetic Algorithm Optimization. Hypothesized, designed, and tested various Genetic Algorithms to solve a single-objective evolutionary computing task to reproduce a painting. | Python · Genetic Algorithms · Evolutionary Computing · Optimization |
| 🤖 | Machine Learning for Navigation | Behavioral Cloning. Trained a Random Forest model on robotic image sequences (video) to predict optimal navigation steps for object tracking. Contributed to the feature extractor and model architecture. | Python · Random Forest · CNN · Feature Extraction · Computer Vision |
| 🧍 | Sensor Fusion & Mapping | Multi-Modal Perception System. Implemented homogeneous coordinate transformations to fuse camera, lidar, and radar. Executed ground segmentation, YOLO?LiDAR-based pedestrian localization, and temporal point cloud mapping from a recorded data sequence. | Python · YOLO · Sensor Fusion |
| 🚙 | Ground Robot Perception & Control | Perception-Driven Obstacle Avoidance. Engineered a holonomic robot for path-following between cones using Lidar and PCL, with camera-based emergency stops for human detection in a Gazebo simulation. | ROS2 · Gazebo · Python · PCL · Lidar |
Note: The last 3 of these projects were developed within private local environments. While the source code is not public, I am happy to discuss.