British physicists are pushing back the boundaries of super-precise timing with a clock technology that promises, theoretically, billion-year accuracy.
The very best atomic clocks neither lose nor gain the equivalent of one second every 30 million years
Its workings rely on the behaviour of extremely cold atoms watched over by ultra-quick lasers.
Several labs are working on so-called optical atomic clocks, but the National Physical Laboratory team is said to have made significant advances.
The journal Science says the UK work could help redefine the second.
"It is among the most significant things this team has ever done, and one of the most important developments in NPL's 104-year history," said team leader, Professor Patrick Gill.
What Professor Gill and colleagues have done is improve an emerging technology for marking time.
Current atomic clocks count time based on the way caesium atoms jump back and forth between different energy levels.
This occurs at microwave frequencies with 9.2 billion jumps making up the moment of time we know as the second.
The very best of these clocks neither lose nor gain the equivalent of one second every 30 million years.
And much of the modern world is built on these remarkable devices. They synchronise television feeds, calculate bank transfers, organise data over the net, and guide airliners safely across the world.
But the new optical clocks now under development operate at much higher frequencies and thus allow for a much finer description of time.
Pros and cons
The US National Institute of Standards and Technology led the way in 2001 with a set-up based on the energy transitions of a single, cooled mercury ion (an atom with an electron stripped off).
Lasers were used to set the ion jumping and count the ticks - just over a million billion of them every second.
Now the NPL has weighed in with a similar set-up but based on a super-cooled strontium ion. According to the Science journal report, it is "a factor of three more accurate than any previously reported optical frequency measurement".
"Different labs have gone for their own particular ion - there are advantages and disadvantages with all of them," explained NPL's
Dr Helen Margolis, the lead author of the Science paper.
"Strontium is perhaps easier because the lasers you need to make the standard are actually simpler than the ones you need for mercury," she told BBC News.
Experiments suggest the current optical clocks can keep near-perfect time equivalent to one second in 10 million years.
Eventually, however, the NPL team believes we could see clocks that are nearly 1,000 times more accurate than the primary atomic timepieces operating today.
If and when this happens, the research community would then probably move the definition of the second away from a caesium standard to one based on one of the ions under study.
Certainly, the Americans have talked of optical timepieces for which near-perfection is marked on the billion-year timescale.
A more accurate definition of the second will improve satellite navigation services, such as GPS and the soon to be launched Galileo system, Dr Margolis believes.
"Also, if you want to send spacecraft to distant parts of the galaxy, you need very accurate clocks to do the navigation," she said.
A new time standard may also provide a vastly more powerful tool to test the laws of physics and help solve a recent scientific fascination - are the physical constants really constant or do they change over time?