Story
This was my first real technical project, started when I was about twelve. A maths teacher encouraged me to enter the BT Young Scientist Exhibition after I sent him a small electronics experiment. While looking for a larger idea, I met an engineer online who explained thrust-vector control and challenged me to build a model-scale version.
I said yes before I understood how much I did not know.
The project became a fast education in mechanics, electronics, control, CAD, firmware, and the gap between a system that works on a bench and one proven in its real environment. The rocket never flew. The motors I obtained were unreliable, and the project had grown far beyond what I could safely and realistically validate at the time.
By the normal definition, that is a failed project. I still count it as one of the most important things I have built.
I presented it on RTÉ’s The Late Late Show and received a minor award at the exhibition. The free pizza was also excellent.
System overview
The airframe held a two-axis gimbal, custom mechanical parts, a flight computer, sensors, and servo control. I designed the non-standard parts in CAD, produced a custom PCB, and wrote the firmware for sensor readout, orientation estimation, and real-time control.
The control loop behaved well in simulation and bench tests. It was never flight-validated, so I do not treat the result as proven.
Durable lesson
The project changed my sense of what I could learn, but it also made one rule permanent: a difficult system deserves staged evidence. Ambition is useful; skipping validation is not.