The Fox Report
Barry Fox’s technology column
Got a sec?
T
he International Bureau of
Weights and Measures wants to
get rid of leap seconds. What are
they? And why banish them? It’s all
about time. The time it takes for the
Earth to move round the Sun is not
exactly 365 days. So, we have leap
years. Every four years February gets
an extra day to even things out.
Slowing earth
Likewise, the second was originally
defined as a tiny fraction of the time it
takes for the Earth to turn once on its
axis. But the Earth’s rotational speed
is not rigidly fixed. The length of the
day is increasing by a few milliseconds
a century – but why? If you have ever
seen ‘ice ponds’ in a field, you have seen
the reason why. The ponds are bowls
in the ground caused by the weight of
lingering balls of ice. A physicist friend
calls it ‘post-glacial rebound’. ‘Remarkably’ he says, ‘10,000 years after the
last Ice Age finished, the Earth’s crust
is still re-bounding after the removal
of the enormous weight of ice. Like an
ice skater throwing out her arms, the
planet has spun ever more slowly.’ All
thanks to expansion and the conservation of angular momentum.
Atomic time
To reflect this lack of terrestrial precision in defining the second, in 1967 the
second was re-defined by tying it to the
vibration frequency of the caesium-133
atom in an atomic clock. In turn, the
day has since been re-defined by tying
it to the atomically defined second.
Because the Earth’s rotation time varies
unpredictably, our earthly time lords
(The International Earth Rotation and
Reference Systems Service (IERS) in
Paris, France) have, since 1972, been
adding a leap second whenever they
thought it necessary. This has happened 27 times, on no particular day
of the year. This random time warping
would crash computers, power grids
and satellite navigations systems were
it not for the fact that they get their time
from publicly accessible time signals,
10
which in turn derive from atomic
clocks. Nevertheless, to avoid confusion, leap seconds will be abolished
in around ten years’ time.
NPL and time
But what would happen if the publicly
accessible time signals were suddenly
lost or blocked? This is one of the topics now being researched by the UK’s
NPL (National Physical Laboratory).
The NPL occasionally opens its doors
for press events and puts on a public
relations show, with a few simple science demonstrations and presentations.
I went along to one recently, hoping
to plug big holes in my understanding
of how so much of today’s technology
relies on precise timekeeping – think
GPS location, video time code, audio
sampling frequencies and so on.
NPL was founded in 1900 at Bushy
House in Teddington by the Royal Society. It is now a commercial company,
with a Chief Executive Officer, but is
owned by the government Department
for Science, Innovation and Technology. It’s home to a thousand scientists
and engineers, and has the feel of a
modern university campus. But, as I
found out first-hand, impressive appearances can be misleading. (See the
part-pain, part-farce footnote about
what happened when I offered NPL’s
Official Public Relations Coordinator
and CIPR-accredited PR practitioner
sight of my draft text, before publication, along with a few specific questions. Apart from being ludicrously
slow to respond, NPL seemed mainly
interested in injecting a dose of civilservice-speak semantics while quietly
ripping out passages of likely reader
interest – without identifying them
and without offering explanation or
justification. Only a few points were
usefully handled and my technical
queries were simply ignored.)
Anyway, back to the NPL, and what
do those thousand boffins at NPL
actually do? I could broadly sum up
NPL’s mission by quoting the open-day
presenters as seeking, ‘measurement
– independent and impartial’. However,
NPL’s PR department now officially
prefers to say, ‘to provide the measurement capability that underpins the UK’s
prosperity and quality of life’. OK, so be
it. What matters is that current research
topics include ionising radiation, 5G,
6G, graphene, underwater acoustics,
quantum physics, and time.
Navigation and time
It’s easy to forget that longitude measurement (ie, how far east/west you
are) depends on knowing the time.
Pendulum clocks don’t work at sea,
which is why John Harrison spent 40
years building the no-pendulum clocks
and watch which eventually won him
the prize offered by the Longitude Act
of 1714. Anyone interested should
climb the steep hill in Greenwich Park
and visit the observatory’s exhibition of
clocks. These include all five of Harrison’s magnificent time measurement
devices. You can also see the Greenwich
Meridian, on which is based Greenwich
Meantime, which dates from 1880. (And
also, Sir Sandford Fleming’s global 24hour zones which start with Greenwich
as zero longitude.)
Simple modern clocks and watches
rely on the way quartz resonates when
an electric pulse is applied. But quartz
materials vary, so their time does
slightly too. Atoms from any one element, however, are all same, so an
atomic clock based on caesium was
developed in 1955. Atomic time became
the reference standard in 1967, and
in 1972 UTC (Co-ordinated Universal
Time) replaced GMT. There are now
500 atomic clocks around world, some
using hydrogen masers rather than
caesium. UTC is disseminated (for
instance from a transmitter in Rugby
transmitting at 60kHz) and is used to
control radio clocks, Internet time and
NPL Time, a commercial time service
that NPL sells and delivers by fibre to
the business and financial sector.
The newest generation of optical
atomic clocks are accurate to around
one second over the lifetime of the
Practical Electronics | July | 2023
Universe. This precision means that
losing leap seconds is now firmly on
the cards, and will remove the last link
between UTC and GMT.
GPS vulnerability and time
The GPS systems we all now rely on
were originally developed for military
use (notably, targeting missiles) but
declassified for civilian use. GPS relies
on time codes and the loss of a reliable
signal can have serious consequences.
For example, autonomous vehicles
can’t function without GPS – missiles miss. Such loss of signal may be
due to criminal or military jamming,
terrorist attack, solar events (flares),
building obstruction or unintentional
RF interference.
‘Tetra 2-way radio is worst of all
because the GPS signal is weak and ‘in
the noise’’ explained a senior product
manager at NPL. ‘The US is extremely
worried about security. Donald Trump
talked about the need for alternatives,
and the recent jamming of drones in
Ukraine has concentrated minds. The
UK economy would lose £1bn over a
five-day period if GPS were lost.’
Fibre is of course immune from jamming. But it’s vulnerable to workmen
digging up the road. And fibre isn’t
portable. One solution/idea is to use LEO
(low earth orbiting) satellites to improve
the robustness of time distribution.
‘Resiliency through diversity’ is the
NPL approach. ‘The better the time,
the better the data. No single solution
can provide a fully resilient assured
capability’. Arguably one of the most
significant, and least talked about,
research projects at Teddington is
development of a miniature atomic
clock which can provide ‘holdover’ –
keeping precise time if the GPS signal
is lost. Currently, the larger clocks can
function with microsecond accuracy
for around a month before time drift
starts to compromise location accuracy.
Fitting drones with atomic clocks could
be a military game-changer. Very likely
it’s already being done, but I suspect
NPL regrets talking openly about it.
Mystery time
Sometimes it’s not clear where devices
are getting their time from. For example,
I still use one of the Joggler devices sold
well over ten years ago by UK mobile
phone operator and Internet service
provider O2. Made by OpenPeak and
powered by Ubuntu, the Joggler looked
like (and worked as) a photo frame but
also connected to the Internet and ran
some Google Apps and simple games.
Practical Electronics | July | 2023
Joggler was a commercial flop and
most of the online functionality disappeared long ago. But my Joggler still
connects to the Internet and works
as a clock, getting its time code from
some unknown service. The time
refreshes after a reboot and correctly
observes Summertime shifts. But it’s
not a radio clock getting time off-air.
Somehow, Joggler must still be connecting to some ghost time server,
somewhere. NPL either doesn’t know,
or does not want to say, where devices
like old Jogglers are managing to find
their time.
NPL Coda
Journalists are often criticised for not
letting companies, or individuals, see
articles ahead of publication. Some editors expressly forbid it. That’s not an
attitude I support – fact-checking can
help stop rubbish ending up in print
or on the Internet. However, it’s true
that there is a risk that the subject will
take so long checking the draft and/or
asking for unreasonable changes that
deadlines are missed.
Despite this, I thought it would be
nice to offer NPL the opportunity to
check what I’d written for PE for factual accuracy; and while they are at
it, answer a couple of queries. I used
NPL’s press contact form to offer sight
of my draft and immediately got back
an acknowledgement email assuring,
‘we are working to get back in touch
with you as quickly as possible.’
Four days later NPL’s PRC (public
relations coordinator) and CIPR-accredited PRP (PR practitioner) sprung
into action with the re-assurance ‘Yes
please! I can get it back to you fairly
quickly too.’ I immediately sent the
draft, but I heard nothing back so after
eight days politely reminded.
‘Apologies for the delay’ said acronymheavy NPL’s PRC and CIPR-APRP, a day
later ‘I’ve asked a colleague in the T&F
department at NPL to review your article
and answer your questions below. I will
get back to you as soon as possible.’
After eight more days I’d heard nothing so politely reminded NPL that some
three weeks had clocked up (ironically
on the topic of time) ‘since my original
contact and I am still waiting for NPL
to comment.’
After another four days had elapsed
when NPL’s PRC and CIPR-APRP offered, ‘Huge apologies for the delay.
There has been a lot of colleagues out
of the office recently. Let me follow
this up now and I will get it back to
you early this week.’
Silence and inaction reigned supreme, so after six weeks, or a full
lunar month and a half, I spent the now
princely price of a First Class Stamp
on snail-mailing NPL’s CEO Dr Peter
Thompson, with the suggestion, ‘If ever
you are upset by any writing about NPL
that you consider inaccurate, you might
like to refer to this timeline.’
A week later an obviously embarrassed NPL staffer phoned to admit
that NPL’s PR kingpin press contact
had gone on leave, without finishing or
delegating the job in hand and without
even putting an auto-answer message
on her email account.
The ‘fact-checked’ draft finally arrived, with a couple of useful minor
corrections, some semantic changes,
chunks of text excised and my queries
completely ignored. Initially, there was
none of the customary highlighting
used to identify changes in a document. When highlighting was added it
somehow managed to miss the censored
excisions. At that stage I gave up on
NPL’s front-of-house press machine.
Silver lining
The upside to the whole time-wasting
pantomime is that it will make a very
good oven-baked rebuttal the next
time anyone pushes that original
question – ‘why don’t journalists
check before publishing?’.
NEW!
5-year
collection
2017-2021
All 60 issues from Jan 2017
to Dec 2021 for just £44.95
PDF files ready for
immediate download
See page 6 for further
details and other great
back-issue offers.
Purchase and download at:
www.electronpublishing.com
11
