You're making an effort to figure out how to figure stuff out, without providing details of what you want to figure out...makes sense.
As an example I intend to build side mirrors for my motorcycle and car that have "dim" or "night vision" capabilities (someone already patented this, FYI).
These are a few methods I use when looking at problems requiring a new and unique design:
1) See if the problem has already been solved for a different application. Then verify that the solution will work for yours.
In the case of side mirrors we know that the rearview mirror mounted on the inside of the windshield has this capability. My design improves upon that slightly so that it will work well for side mirrors.
2) Break down the problem into it's simplest form, solve it and see if it can be reassembled
to solve the aggregate problem.
If I were building it from scratch I would note that I need a method of A) receiving the image, B) displaying the image and C) setting the display contrast.
I could do this by A) mounting cameras where the mirrors are and B) placing 4" LED display panels at some viewable location. C) A video circuit would have to automatically adjust contrast and brightness.
3) Look at the broad picture and see if there is a behavior that can be mimicked to solve the
problem.
If you look at chameleons they change color based upon their surroundings. Perhaps something where the silver pigment on the mirror could be darkened would reduce glare from headlights at night...
[It turns out a company already makes mirrors that do just that. You run a current through the mirror in one direction for daytime use and another for night time use.]
On the off chance that you have experience with digital logic, I discuss that here in one of last years class projects:
http://amazingbackpack.webs.com/parity%20checker.pdf