How SpaceX's starship caught its booster: Control engineering [video]

youtube.com

28 pointsdtquad4 comments

The video is basically a high-level overview of general control algorithms available for this purpose. Lots of speculation and "might use" qualifiers.

Does anyone know of any talks where engineers who actually worked on these problems discuss the unique and interesting challenges they encountered, and how they worked through them?

Spoiler alert: it was not deep neural nets.

Around the end of the video it says AI could be used to adjust the parameters of control, i.e. PID, algorithms.

PID was what we were doing in the late 1800's. Adding AI to tune it is like applying lipstick to a pig.

We have a ton o theoretical and practical advances in our disposal today

Reply on news.ycombinator.com