I spent 10 weeks working on the EKg program with Richard, Ed, Dave and Mr. Lu in summer 2004. I began work on a program to control the incredibly acurrate voltage reference used in the apparatus. The voltage standard used is based on the Josephson Effect and it is the same standard used to define voltage around the world.

The thing about that article is that it provides a very brief and dumbed down idea of what these scientists are actually trying to accomplish. It is not so much that they wish to replace the standard of the kilogram but that they are answering questions about what our fundamental units should be. Many people feel that if they could make the whole SI system all over again, they would never let mass be a fundamental unit.

If you put on your physicists hat for a second, you can use the EKg to get the most accurate readings for Planck's constant, which they have done. If you like, you can stop adding more accurate decimal places on the end and just define it, such as the scientific community has done with the speed of light.

If you put on your metrologist hat, you have two ways to define mass using this device. You can take the plank constant that the Ekg group has found and define the mass of an electron as some readers suggested. Here electon mass = [2*(Rydburg constant)*(planck constant)]/[(speed of light)*(fine structure constant)]. This definition is kind of messy because it depends on a lot of other people's work.

The other way is to use Einstein's equations: Energy = planck's constaant * frequency = m * c^2. Solve for m. In this case you have the kilogram being defined as the mass of a body at rest whose energy equals a collection of photons with frequencies that sum to 135639274x10^42 Hz. This leaves out gravity all together.

To answer a few other comments: Yes this thing looks like a rube goldberg device, but there really is no other way to make a device of this type. It is so large in order to minimize error due to vibration, stray magenetic and electric fields and other such things. And you have to remember that these guys make this stuff up as they go along. There is no one to say how one makes an apparatus to give you a mass standard that is accurate to within 1 part per billion per year. Once they achieve their accuracy, they can probably go back and redesign the system. These guys are trained physicists, not engineerings.

Other countries are working on developing a standard based on Avogadro's constant using silicon atoms. Currently the russians are producing the purest collection of silicon atoms ever made by humans. This is good. You want multiple efforts trying different ways because all of these systems back eachother up or disagree which means you go back to the drawing board.

You would implement this new mass standard in the same way you do any standard. You designate one place as holding the standard and everyone goes back there once in a while to check their own national standards against it.