They will ask for a working model (not normally a requirement).
My understanding is that they will reject the application as not being "useful" or inoperative. However frankly I have never filed such an application so I don't have any firsthand knowledge. (Patent, jump in any time)
My understanding is that they will reject the application as not being "useful" or inoperative. However frankly I have never filed such an application so I don't have any firsthand knowledge. (Patent, jump in any time)I haven’t filed an application like this either. You are correct that the rejection would be on utility grounds, but I suspect you could have enablement and written description issues as well. At least I would make those objections, wanting the applicant to prove it.
I have filed applications where you wouldn’t immediately assume the widget worked or was better then previous widgets. Generally, if you just have a widget with lots of little widgets and you claim it’s efficient, and submit an affidavit with the data to back it up, I would bet you can get a patent on it. If you write your claims to a perpetual motion machine or a perpetual energy machine, I suspect they would want more then an affidavit with some data attached. I would think they would want to see it or have an independent source test it.
In the end, even with the data, the examiner could just get obstinate and refuse to let it issue, forcing you up the chain to the board where it probably would issue, absent any real proof it doesn’t work.
Then the infomercials for the “patented” perpetual motion machine” start. ;-)
patent +AMDG