One of the really big problems in the engineering world is estimating the difficulty, time, and cost of a new project. In many cases, estimating the project if far more work than actually doing it. Here is one way to get a handle on how much trouble any specific project may be.
First, divide the whole project into parts. If needed, divide the parts into smaller parts until you can put each of them into one of the following categories:
If your parts are all Class-I, no problem. Just cobble together all the bits and pieces from previous projects.
Class-II means you do some research into how it has been done before. Google around and see what you can find that fits what you need.
Class-III is where it begins to get sticky. You will have to 'reinvent the wheel' to do this part, but you have reason to believe it is possible.
You can be in trouble estimating the project if you have any Class-IV's. If you do, work on these first to see if its a deal killer. Let your client know that until you determine if this part is possible, you really don't know about how much total effort the whole project will take.
Unless you are engaged in pure research for a University or something, don't do any project with a Class-V item. The is called a 'Science Project'. If it is part of something you have already been hired to do, you may want to have a real sit-down with the client and give them the bad news.
Class-VI means you avoid the whole project entirely unless you are planning to move to an alternate universe where the physical laws are more amenable to what you need to do.
The overall class of a project is the highest number of any of its component parts. If you have just one class-IV part, the project itself is considered class-IV.