Electronic Voting Machines
While I don't think that electronic voting machine fraud actually happened this election, any system which uses closed source software that is controlled by one company and leaves no paper trail is just too easy to abuse. And this is before reading alarming statements made by the CEO of the company that controls the machines. But the problem is not that CEO's political views -- I'd be just as horrified by an extreme left-leaning CEO running a company making closed-source voting machines. The problem is the closed source voting machines.
The scientific community has something called peer review. It's the process by which members of the community check each other's work to keep each other honest. Anything from honest mistakes to sloppy methods to not caring to outright malice can distort results. When the only thing you see are the results, you have no idea how valid those results are. But when not only the results, but also the methods, data, records, equipment, etc, used to produce those results are available to the public where they are studied by other informed members of the community, we can gain a sense of confidence in those results. That's why you don't hear much about those people that thought they found cold fusion in the early 90s anymore.
Similarly, the software that runs a voting machine needs to be out in the open. There are many well-respected members of the software and computer security industry that would be eager to review any such software, and the security of the machines would benefit greatly from it. In something as important as a presidential election, we should demand no less. For more information, see: this eff page.