Jonathan van Zyl
Controls and AI Engineer
MSc Aerospace Engineering Student at TU Delft, graduating in Summer 2025, having held positions at Forze Hydrogen Racing and Infosys, along with part-time positions at student teams: Lunar Zebro and DARE (Moon Rover and Rockets).
I'm a passionate engineer with expertise in aerospace, controls, and artificial intelligence. My goal is to create innovative solutions that solve real-world problems.
My Work
Check out some of the projects I've worked on and the technologies I've used.
3D CNN for semantic segmentation of mouse embroygeneous.
3D convolutional neural networks-based segmentation to acquire quantitative criteria of the nucleus during mouse embryogenesis
Read the Article by clicking here!
Image Source: Y. Tokuoka et al., '3D convolutional neural networks-based segmentation to acquire quantitative criteria of the nucleus during mouse embryogenesis', npj Syst Biol Appl, vol. 6, no. 1, p. 32, Oct. 2020, doi: 10.1038/s41540-020-00152-8.
Cadence
A take on 'To-Good-To-Go' meets 'Uber Eats' for preventing food spoilage in bulk selling firms.
Entry for BizHack 2024 Hackathon, won 'Most Collaborative Team' award.
Learn front-end development in 4 hours to make these mockup UIs:supplier UI andbuyer UI.
Computer vision and navigation module for drone
Worked in the paparrazi environment to design a computer vision system for drone navigation.
Tested in the TU Delft MAV Lab.
Forze Hydrogen Racing
Working on a future Le Mann 24 hour hydrogen race car:
- Anode hydrogen flow control system.
- Fuel cell software-in-the-loop simulator.
- Ejector test-bench
BSc Thesis
Design of a non-linear model predictive controller for a eVTOL under urban turbluence.
Presented at the 2022 AIAA Student Conference in Adelaide, Australia.
My Skills and Experience
I'm a skilled engineer with a passion for building innovative control systems using novel technologies. My areas of interest are: Deep Learning, Reinforcement Learning, Control Theory, Operations Research, Evolutionary Algorithms, Swarming Intelligence and Intelligent Controls. Open to working on new projects and collaborations in different fields.
Controls Engineering
Proficient in Matlab/Simulink, Python, C and C++. Experienced in designing and implementing control systems both PID and MPC, for a hydrogen race car and autonomous vehicle in real-life. Also, VOR, pitch-rate and GNC satellite using PID, and an urban air mobility vehicle using MPC; in simulation for University projects. Developed FSMs, tuned controllers in real-life and implemented evolutionary algorithms (GA and PSO) to optimize controllers.
Intelligent control systems: Data-driven modelling (LNN, GPR) and Iterative Learning Control.Artificial Intelligence
Proficient in Python, Pytorch and OpenCV.
Data driven modelling: LNN for double pendulum simulation & NARX for system identification of a hydrogen flow simulation.
Deep Learning: 3D CNN for semantic segmentation of mouse embryogenesis, and a CNN for image classification.
Reinforcement Learning: transfer learning to play MsPacman from Pacman.
Machine Learning: including XOR (Python & C++) and MINST. CNN, RNN, LSTM, GRU, GAN and Transformers.'Aerospace' Engineering
Operations Research: solved vehicle routing problems using Mixed-Integer Linear Programming.
Computer vision: for a drone's navigation system.
Engineering: hold a BSc at the 3rd best Aerospace Engineering University in World and 1st in Europe. Now pursuing my MSc at the Aeropsace Engineering faculty of TU Delft on the Simulation and Controls track.
My Background
I have a strong academic foundation that has prepared me for a successful career.
A-levels: A* Maths, A Economics, A Physics
Courses from MSc Space Engineering, with supplementary courses in Business and Spanish.
Scoring 8.5/10 in my final year project.
Part-time positions at student teams:
- Solid Propulsion Engineer at DARE.
- Mechanics Engineer at Lunar Zebro.
TU Delft, Netherlands.
Bangalore, India.
Control and Simulation
Get in Touch
I'd love to hear from you! Feel free to reach out with any questions or opportunities.