Space fans are well aware that there are plans for NASA to build a fully electromagnetic drive, much to the skepticism of critics. However, a leaked paper suggests they actually did it -- and it works.

A leaked, unpublished paper from NASA engineers have revealed what appears to be a fully functional EM Drive. This is a form of space drive that can produce thrust through electricity alone, which totally violates newton's Third Law of Physics.

The EM Drive is an invention of British Roger Shawyer and was controversial due to its nature. Recent "tests" have shown that the thrust these drives produce is small or even nonexistent. However, this leaked paper from NASA's Johnson Space Center appears to have made a functioning EM Drive. They even concluded it's viable for space propulsion.

According to the Register, while NASA did build a functioning EM Drive, it is in fact extremely slow. The EM Drive that was "tested" only produced 1.2 millinewtons of thrust per kilowatt. This is significantly low compared with the 60 millinewtons per kilowatt produced by the best ion drives.

While this seems tiny -- and it is -- it appears to be twice as powerful as the thrust from solar sails. An ion drive alone requires power supply and fuel (usually xenon gas). However, an EM Drive can simply run using solar power.

The EM Drive works in a frictionless vacuum where tiny amounts of thrust add up very quickly. Given that it only needs electricity, the drive could keep satellites floating almost indefinitely and journey to other planets, provided they are together.

The tests in the paper were done in a vacuum so air current didn't influence the readings. They were also conducted in a chamber that is shielded from magnetic and RF interference. If the readings are to be believed, then it's so far, so good. A theory is that protons fired around the nozzle may be providing the thrust, but this has to be proven with more tests.

Given that the paper is still unpublished, this means the paper is not yet peer reviewed. This means it may even have errors that have not been spotted by the common reader.