I did start off using a relatively simple PIC-based controllers for the air suspension, done by my friends at eLabtronics.
This sort of programming:
...and this sort of hardware:
But I found I really struggled with it. One issue was that the re-programming wasn't 'live' (you had to upload the complete program to the controller before you saw the new settings in action) and another was that it didn't have enough software 'building blocks' for my purposes (eg no PID block). A final issue was that there were no gauges in the software - so you couldn't see what was actually happening.
I got more and more frustrated, and then called on someone I know who has developed full programmable engine management ECUs.
He said: "Sure!" and then helped me massively in churning out custom stuff for a brilliant ECU* - but programs written in flow-chart software that I can understand (I cannot write code, so I need flow charts).
So I tried his approach, and very quickly got better results than previously. When I got stuck, my buddy would help me out with some new programming - eg averaging a road speed input. (My buddy's name is Andy Wyatt, and he has now moved to Haltech.)
Within an unbelievably short time (weeks, not months) the system was up and running and working well.
*the ECU is designed from the ground up as a completely customizable controller. It doesn't seem to have been a commercial success, but it is absolutely awesome. If I could afford to, I'd buy five - control my (yet to be built) solar house, any industrial process you name, etc.
As I said, I haven't had to now touch the programming in literally years - something that I was starting to think was a bloody impossible dream!
(I think the air suspension system (engineering, hardware, electronics, software) is the single hardest car project I have ever done - but air suspension
per se doesn't have to be
this hard!)