No weekly update last week, and a bare minimum one this week – not due to lack of progress, but lack of time.
I’ve been plugging away at the new Auto-pilot, which has involved some intricate work with n-point splines and a lot of experimental research. I want a hybrid system where the Auto-pilot will follow a smooth course from A to C, avoiding B – yet still be velocity based so it can react to moving objects (asteroids, pirates, missiles et al). Achieving this ain’t so easy 🙂
Pass 1 has a great velocity only system which looks good, is smooth – but it’s darned hard to get the thing to reliably stop at the destination without overshooting (that or my brain can’t handle the mechanics!). But, regardless – it’s not the complete system I want.
Pass 2 has a Hermite spline solver, which computes the ‘ideal smooth course’ from A to B – guaranteeing arrival at the destination. This is far more flexible, and useful (ie. looping paths, orbital paths, complex courses). However, it’s ‘rigid’ in that the vessel has to follow a pre-ordained point – like a roller-coaster car on the track. It doesn’t consider dynamic things getting in the way.
Pass 3 now has both – with the auto-pilot managing it’s velocity by aiming at the point being driven along the spline. It’s still WIP – but it will allow the vessels to react to various external stimuli and then get back on track – or at worst recompute a new course when free to do so. And it means they can fly around things now instead of straight through them – so docking looks a little more realistic now 😉
Of course, once I get it working – then I have to make it work reliably for all vessel sizes and masses, with all drive systems, in all situations and with lots of stuff flying around trying to get in their way… 🙂
Comments are closed.