Please don't do this, you will starve. Very few people are able to make enough money from open source in this way and I don't think this is the kind of thing that's suited to it.
Commercialize this technology, market it, sell it.
The main considerations revolve around:
a. Is it patentable? (OP should contact a former USPTO examiner or a reputable legal shop like Orrick.) Provisional patent a minimum, seems a must, or it's not worth going forward if Mr. Wonderful can knock it off in China.
b. Is it worth the effort to source and offer components which are commodities? Some folks want "batteries included" and will pay for it, so mark those up to where it makes sense to make it worth the while.
Disclaimer: I've worked inside ivy school research and non-research depts and enterprise shops too.