When you start, you never know how hard development will be, but the best rule of thumb I have come up with is to look at the cost and time to do A-B comparison testing of two design iterations.  Sure, sometimes you knock it out of the park on your first shot, but that's not something you can plan on. 

​

Strategies for making incremental progress efficiently:

​

"Fail Fast, Fail Forward".  In this approach, you build a partial design and test it to see what happens, with a focus on reducing the time between iterations.  Each time, you learn something to move forward with.  The tools you have now to embrace this strategy include desktop CNC, 3d printing, pre-made validated anatomy models (like Syndaver), and a culture that says when it's time to go home, start a 3d print so you start the morning with parts you can pick up and test to see what you're missing.

​

"Design, Build, Test" loops.  This approach is more for advanced work, where you have a design that is maybe 70% of the way there, and you need to take it to the next step.  Each loop is more formal, testing is more defined, and you expect a certain level of success.