Thrust vectored scale rockets flown with quadcopter flight controllers

Oct 15, 2026 Last reply: 1 day ago 4 Replies
Jump to replies (4)

Most scale rocket builds in this hobby fly straight up on passive fins and call it done. T-Zero Systems took a different path: the ascent abort vehicle and the SLS model are flown using quadcopter flight controllers, configured in iNav as if they were airplanes. That alone puts this work in a different category from the usual tube-and-fins approach, and the debut trailer gathers a full year of it, covering fabrication, ground testing and launches of a Falcon 9, the Space Launch System and an Orion launch abort vehicle.



The way it comes together is worth studying. There is an Adafruit Metro Mini board paired with a BMP388 sensor, 3D-printed mounting brackets and avionics gimbals, and a handheld launch controller with clearly labeled POWER, RESET and ARM switches. The airframe side gets servo-actuated grid fins and motorized strongback arm retractions, with terminal blocks and status LEDs rounding out the electronics. Ascent and descent simulations run in Excel, and on-screen control loop code in C/C++ defines variables such as XInput and YOutput. The custom launch mechanism, with its dark vertical plate, three circular cutouts and brass L-brackets, looks properly shop-built.



What stands out to me is the honesty of the program: static motor firings bolted to the ground, a spoken pre-launch status check from a launch director, an onboard camera view during ascent, a pad abort separation, and an on-pad SLS failure with flames and smoke left in. The maker admits there have been more than a few failures, and that openness makes the successes mean more.



The open question is tuning. The abort vehicle is short enough that overcooked gains make it correct too much, and the maker hopes to develop an empirical tuning model based on rocket parameters. A new nose cone is also needed before the next Falcon flight, all on a budget of roughly thirty to fifty dollars a month.



Project by T-Zero Systems. Watch it on their YouTube channel:

formatting link


Setting the flight controller up like an airplane in iNav seems like the smartest call in the whole project. The maker notes these controllers are robust and pretty much work immediately, which beats hand-rolling stabilization from scratch. The trade-off is that gain behavior does not transfer cleanly between airframes, and the short abort vehicle proves it by correcting too hard.

How is the tuning actually done for the abort vehicle right now? Is it trial and error at the pad, or does the maker lean on the Excel ascent and descent simulations to pick starting gains before a flight? I am curious how far the spreadsheet work carries the process.

From what the maker has said, the simulations of rocket ascent and descent are built in Excel, and tuning so far has come out rough: the abort vehicle is so short that the gains were overcooked and it overcorrects. The stated goal is to develop an empirical tuning model based on rocket parameters, so a repeatable method is still ahead rather than in place.

Since the maker has more than a few failures that could be documented, a dedicated failure breakdown series would be a strong addition. Pairing each one with the Excel simulation and the control loop code would show exactly why a flight went wrong, and it would feed straight into that empirical tuning model once it exists. Multistage flights would also make great future content.

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required