Jefferts’ pursuit is precision
- By Florence Olsen
- Oct 25, 2007
Steven Jefferts’ invention fills a 20-by 12-foot room, but the centerpiece of his contraption, with its computer systems, racks and electronics, is a tower 6 feet tall and 4 feet wide. The tower is bolted to a 1-foot thick table covered with mirrors and lenses, which work in tandem with a titanium sapphire laser to keep the tower cool inside.
The laser is one of several lab components that Jefferts and a small team of government physicists assembled to create what looks like a Rube Goldberg machine. But the apparatus is a precision instrument. The interplay of its cesium atoms and microwave oscillator produces the world’s most accurate standard measurement of a moment in time.
The instrument — NIST-F1 — belies its complexity and demanding nature. Jefferts, a senior staff physicist at the National Institute of Standards and Technology’s Time and Frequency Division in Boulder, Colo., said the instrument sends an alert to a member of the research team at any time — day or night — when it develops a problem that could produce inaccurate time data. “My wife’s comment is, ‘It’s sort of like having a perpetual 2-year-old,’ ” Jefferts said.
Government and industry program officers and information technology officials rely on Jefferts’ invention without realizing it. Jefferts’ group uses the instrument to maintain the accuracy of an ensemble of atomic clocks at NIST that provides Internet Time Service. NIST receives about 2 billion time requests each day — 25,000 a second — from computers, network devices and satellite navigation systems to be automatically synchronized with the Internet Time Service.
The service provides a precise time stamp for financial and other time-sensitive transactions. It also provides an accurate local time-and-date log of events affecting critical equipment, including routers, gateways and firewalls. The Internet Time Service helps managers establish a sequence of events during a postmortem review of any problem, said Judah Levine, a senior staff physicist at NIST. Another group of people relies on the time service to have their e-mail messages correctly time- and date-stamped.
As it was building the calibration instrument, Jefferts’ group had to contend with water problems in the room that originally housed it.
“Here was one of the most accurate measurement devices of any kind in the world, and for many years, Steve and his team had to work in a room that had a leaky roof,” said Tom O’Brian, chief of NIST’s Time and Frequency Division. “He had large sheets of plastic wrapping over the clock to try to prevent the leaky [roof] from damaging it. Fortunately, we’ve been able to move the clock to a better laboratory that doesn’t have a leaky roof and has much better temperature control.”
Jefferts’ team also wrote and now maintains the software that controls the calibration instrument. Jefferts and fellow physicists have used the software to improve the accuracy of NIST-F1 since its installation Dec. 29, 1999. NIST-F1 is at least 10 times more accurate now than it was then. Jefferts said the software “checks for errors and flags suspect data and calls us when it’s convinced there’s an error. That has been a huge part of the improvement.”
But Jefferts isn’t satisfied. The group’s current project is to build a successor, NIST-F2.
“We keep pushing, trying to make these things better and better and better,” Jefferts said. His team is designing a calibration instrument that will be more accurate than NIST-F1 because it will be cooled by liquid nitrogen. “It’s kind of a painful step to have to take, but our leading source of inaccuracy is the room temperature environment,” he said.
Jefferts was leading a project that would have put a calibration instrument on the International Space Station, a project that NASA canceled shortly after the Space Shuttle Columbia disaster in 2003. Microgravity would have enabled the researchers to achieve much greater atomic clock accuracy than possible on Earth, but it would have been a challenging project to complete, he said.
“I have mixed emotions about the fact that it was canceled, in the sense that it was a wonderful project, but I’m not sure, given the realities, that we would have ever gotten there,” he said.
O’Brian said Jefferts’ team has successfully navigated difficult challenges, for which he credits Jefferts’ “broad and deep scientific knowledge and leadership.” The accuracy of Jefferts’ calibration instrument is such that it would lose 1 second in 100 million years. But what kind of person worries about measuring 9,192,631,770 atomic-level oscillations to provide the world’s most precise definition of a second?
Levine said “I don’t think you have to be special. You just have to be worried about that last decimal place.”