When you look at your pictures, do they all look alike? Do you want to move in a new direction, but don't know how? When I'm in this predicament I force myself out of my comfort zone by randomly choosing the design elements I will use to build my picture.
There are several ways of accomplishing this. If you know how to program, it is fairly easy to write one that generates a random list of design element combinations. If programming isn't for you, simply write out a list of design elements by hand and roll dice to select one element from each category. Either way, build a composition based on the list. Most of the time your results will be distinctive and edgy, but sometimes, no matter what you try, the combinations just won't work for a given concept. Should this happen, create another list and try again.
You can also base compositions on abstract images. This is nothing new, but it is still an effective way of working. Source material can come from almost anywhere—abstract paintings, stains and rust, close-up photos, particle generator programs, photo effect filters, etc. The only requirements are strong composition and unrecognizable subject matter.For maximum benefit pick an abstract image that differs sharply from your typical work.
Often what's obvious to us isn't to other people. We draw a tree and think that it's vastly different from the one we drew in our last picture, but our viewers don't see it that way. If we want something to be unique, we need to study its characteristics and determine which are most distinctive, then emphasize them through exaggeration. Push the exaggeration to extremes (just how far varies depending on how realistic you work) to be sure your viewers get the message.
Flipping an idea on its ear is another good way of seeing things differently. It's a simple, but powerful technique—develop a new design by reversing some or all of your dominant design elements. If it's a night scene, make it a day scene. If your center of interest is small, make it huge. Challenge yourself to communicate the exact same idea in a completely different visual way. It doesn't always work, but the results are impressive when it does.
You can try this technique at any point in the design process—in the beginning while you are searching for a direction, in the middle when you are evaluating ideas, or at the end after you have a design you really like. I know it's tempting, especially for students, to settle for the first good idea that comes along, but I encourage you to resist this and open your mind to more creative solutions.