There’s a great article on Ars Technica about the design of the first stealth fighter. An airplane that’s invisible to radar has all sorts of interesting constraints. The shape of its surfaces isn’t dictated by what is most aerodynamic, but what will deflect radar. This results in an odd-looking airplane that is less stable than conventional ones.
Of course, the airplane must still fly without crashing. The solution? Fly-by-wire controls that compensate for the airplane’s inherent instability. With such a cybernetic system in place, designers could push the plane’s form. They could offset the stability requirement to a different subsystem, and so optimize surfaces for radar-deflection.
This has me thinking about my projects. How many of the constraints I’m dealing with could be offset to cybernetic systems? (After all, every interaction in a digital information environment is fly-by-wire.)