80 Science & technology The EconomistDecember 7th 2019
21produce five-metre sections of propellant
tank. Its most precise ones create engine
parts with an accuracy of 40 millionths of a
metre. A conventionally manufactured
rocket of similar size would contain, the
firm says, nearly 100,000 parts. Terran 1 has
less than 1,000. That simplifies the supply
chain and accelerates the testing of parts.Bright ideas
Speeding up launcher production in this
way helps. But it will not be enough if
America is to fulfil its goal of launching re-
placement satellites with a day’s notice.
That is one reason, says Mr Roper, why the
air force is now buying, at a series of pitch-
ing events that started in March, ideas for
ways of prevailing in “high end” orbital
combat. Encouragingly for proposers of
such ideas, little bureaucracy is involved.
Settlement for those accepted is immedi-
ate—the air force sidesteps its lumbering
payments system by using official credit
cards to transfer money instantly to peo-
ple’s PayPal accounts. Those who present
clever proposals can thus pocket awards
exceeding $100,000 within minutes. The
latest of these pitching meetings, on No-
vember 5th and 6th, resulted in on-the-
spot contracts worth $22.5m.
Meanwhile darpa, America’s main mil-
itary-research organisation, is trying to
organise a responsive-space competition
of its own. Next year it hopes to hold a
challenge in which teams will attemptlaunches twice in a matter of days or
weeks, each time learning only shortly be-
forehand of the mission’s location, desti-
nation orbit and payload characteristics.
This has never been done before. Program-
ming the computers takes time, and the
rocket must be trimmed in advance for the
particular trajectory, taking into account
such factors as the weather. Prizes of up to
$10m will be awarded.
It is a measure of the task’s difficulty
that, of the 55 teams which signed up ini-
tially, only three qualified, and two have
subsequently dropped out. The name of
the remaining competitor is secret.
At least one of the dropouts has not giv-
en up completely, though. That firm, Virgin
Orbit, has turned a Boeing 747-400 into a
flying launch pad. At an altitude of about
10.7km, the aircraft releases a rocket called
LauncherOne. This rocket’s engine ignites
after 4.8 seconds of freefall.
Such launches, Virgin Orbit says, can
take place above nasty weather. They also
make it easier to reach east-to-west “retro-
grade” orbits, because the launching plane
can fly in the opposite direction to Earth’s
spin, reducing the launch velocity required
for such an orbit. Though Virgin Orbit’s
system has yet to put a satellite into orbit,
Britain’s Royal Air Force seems interested.
In July it announced a deal to launch small
satellites on notices possibly as short as a
week. By today’s standards, that is, indeed,
pretty responsive.^7These oak planks, once part of the portico of a property just outside Imperial Rome,
travelled a long way before the builders got their hands on them. The science of dating
trees by looking at their growth rings is now so good that Mauro Bernabei of Italy’s
National Research Council and his colleagues were able to say, in a paper just published
in PLOS One, where the trees that provided the planks had grown, and when they were
cut. Rings’ thicknesses are affected by the local climate. Comparison with samples of
known origin showed that the trees grew in what is now eastern France, and were felled
between 40 and 60AD. That speaks of a sophisticated timber trade, which floated the
logs down the Saône and Rhône to the Mediterranean, and thence to the Eternal City.You can’t get the wood, you knowA
few years ago it looked as if malaria
might be on the way out. From 2000 to
2014 the number of cases and deaths fell. As
the World Health Organisation’s annual re-
port on the disease shows, though, the de-
cline in cases has ended (see chart overleaf )
and that in deaths has slowed. The report,
published on December 4th, says there
were 228m cases of malaria in 2018, which
resulted in 400,000 deaths. Most victims
were young children in Africa. That is a far
cry from targets set in 2015 for the near-
elimination of malaria by 2030.
That strategy of elimination had count-
ed on $6bn a year being poured into malar-
ia-control efforts. Funding in recent years,
however, has been about $3bn a year. More
money would surely help. But substantial
gains can be made by doing things more ef-
ficiently—something at which malaria
programmes have been dismal.
Stopping malaria relies on three things:
insecticide-treated bed nets to prevent
nocturnal mosquito bites; the spraying of
homes with insecticides; and the treating
of pregnant women and children with
rounds of preventive medication. These
are all “imperfect tools, often used imper-
fectly”, says Pedro Alonso, head of the ma-
laria programme at the World Health Orga-
nisation. Countries usually deploy the
same package of measures everywhere,
even though infection rates and their sea-
sonal patterns vary a lot between regions,
and particularly between cities and the
countryside. Transmission reaches a peak
in the rainy season, when mosquitoes are
abundant, so preventive mass-treatment
of children then can make a huge differ-
ence. Regional variations are particularly
pronounced in large countries like Nige-
ria—a place that, by itself, accounts for a
quarter of the world’s malaria cases.
The typical approach of a malaria-con-
trol programme is to bombard a country
with bed nets and then use whatever cash
remains for sporadic rounds of preventive
medication. But in many big cities, such as
Dar es Salaam and Nairobi, cases are few
and far between, so deploying nets there is
a waste. Overspending on nets at the ex-
pense of other things happens partly be-
cause nets are easy to count—a feature that
aid programmes are particularly fond of.
Results which cannot be attributed directly
to money a donor spends tend to fall fur-
ther down that donor’s list of priorities.
This kind of reasoning tips the scales, be-Cases of malaria have stopped falling.
Better targeting is neededMalariaOff track