As I mentioned earlier, I always had an interest to space. Sadly, I had no possibility (yet?) to apply myself in aerospace industry. But despite this, I studied astrodynamics and orbital mechanics myself, following the books of best experts in the world.
Practical application
Any knowledge, would it be coding or math, foreign language or sport activity, forget easily in case if you not apply it. Complex and very specific knowledge, like astrodynamics or orbital mechanics, requires even more attention, because they staying intact only if you’re applying it — I mean practically!
But how I can do astrodynamics excercises in practice, in case if I have no my own spacecraft?
When I have been a volunteer astrodynamic of Edelveiss (satellite of Geoscan Space), it was great. But — Edelveiss decayed at 2024, so what I had to do without it?
Simulation
So, the only way I had was simulation. It’s not so bad idea at all — most of serious space companies using simulations of spacecrafts in their projects. But this simulations in most cases project-oriented, and built on very sophisticated and freakingly expensive commercial software (from MatLab and 42 to AGI STK). For me it was not a big deal to study it (after GMAT, Poliastro, etc.) But:
- I am limited with money.
- This simulation software working on narrow problems (separately aerodynamics, separately orbital mechanics, etc.)
That’s why I turn my sight to space simulators.
Orbiter 2016
As I said, in aerospace simulators are OK, and, for example, in aviation they widely used for pilot training. But space is totally different thing. There is no serious space simulators exists, except one — Orbiter 2016.
And it is so rare, that now, at October 2025, its official web site — http://orbit.medphys.ucl.ac.uk — appeared to be no longer available!
But despite this, it is extremely good serious space simulator — i.e., where orbital dynamics realized in as realistic way as it is possible. (Yes, it has its own troubles, but they are fixable — it’s pretty easy to modify it). European Space Agency used it as training simulator for their crew!
LUA, C++, autopilots
To modify the thing, you can use embedded LUA environment, and C++, allowing you to create add-ons using documentation for simulator. So, I focused on development of autopilot systems, allows me to control winged spacecraft precisely in all stages of flight — from launch on back of heavy launch vehicle to landing after long journey on the orbit.
I choose LUA as basic language for implementation autopilots, put around myself math and GNC textbooks, and dive into work!
That was freakingly hard, because… Just because space is freakingly hard! But despite this, I implemented this autopilots and made a full-scale real time simulation of spaceflight, including intercept and de-orbiting of super-heavy scientific satellite ESA Envisat.
Mission planning and simulation
Studies on GNC, orbital mechanics, navigation in space and other matters required lot of time. Development took almost 4 months, and that was a result of lot of previous studies (required 5+ years in total).
Using my own Python astrodynamical library based on algorithms from David Vallado «Fundamentals of Astrodynamics and Applications», I planned a mission to Envisat, and at May 12, 2025, in 12.5 hours of simulated flight, controlled by autopilots of my development, together with my friends Anna and Rebecca, and ChatGPT as ground control, realized all of planned goals — against all odds!
That was hard, but we did it.
Results
I greatly improved my coding and astrodynamics skills gathering this through unique experience. This simulation was highly assessed by top experts in Geoscan Space, ESA and US space industry.
Total line count of LUA code is 11473.
More details about all the project you can here in final report about Operation Silent Giant.