|
Chipping in for More Solar Power
Advocates of clean, renewable energy have long eyed solar power
as a way to move forward. The technology's problems, however, include
inefficiencies that cause loss of power. Partial shading and the
power lost from converting direct current to household alternating
current reduce the yield of solar panels, effectively raising the
price of solar energy. A three-year-old startup company thinks it
can solve these issues by harnessing a different kind of power -
namely, computer power.
Israeli-based SolarEdge produces a junction box for solar panels, an inverter, and web-based monitoring software. The junction box features a computer chip that fixes the voltage coming from the panels, making it steady rather that erratic. This reduces issues with power degradation. The inverter is smarter than your typical device; it not only converts electricity, but it also collects data on the temperature and output of each panel. All of this information can be viewed with the company's software, so users can better understand the dynamics of their array, including degradation issues.
SolarEdge believes that by using its products, solar power consumers at both the single household level and above could see their solar arrays become up to 25 percent more productive. Several solar companies are testing the systems, and CEO Guy Sella hopes to line up deals with manufacturers to fit the junction boxes onto the panels during production, rather than afterward. SolarEdge has competition from other firms trying to tackle the lost energy issue, but Sella says his company's system costs less and solves the same problem.
Read
more about this
Flying in the Face of Flu
When people fly, their germs fly with them, which means that flu
variations can spread around the world within hours, sometimes even
before the carrier shows any symptoms or suspects they might be
infectious. Epidemiologists believe that future pandemics will spread
via air travel, leaving many countries faced with the challenge
of containing the spread of disease. But how can a country such
as Mexico deal with this threat
CAPSCA "the Cooperative Arrangement for the Prevention of
the Spread of Communicable diseases by Air travel" thinks they
can help. The organization tries to help airports in developing
nations work out plans for containing the potential spread of communicable
diseases. It helps them think ahead. When faced with an aircraft
suspected of harboring a disease, for example, CAPSCA makes sure
the airport has thought about where to park the plane, how to handle
the luggage, what to do to maintain contact with passengers who
appear symptom-free, and so forth. This is in addition to figuring
out which passengers are most likely to be infected and whether
they need preventive treatment, or even to be taken to a hospital.
The protection of customs officials and airline personnel must also
be considered.
CAPSCA hopes to see the development of automated systems that can be used at airports to detect potentially infected people. Turkey succeeded in spotting its first case of swine flu with such a system; infrared cameras installed at its airport detected a passenger with a high temperature. It's not an ideal solution because infections develop at different rates depending on the bug producing them; some flus become infectious before the host develops a fever, for example. One company is working on a microphone system that can detect and locate people suffering from persistent coughs, and determine whether that cough indicates the person may be ill. More precise technologies, such as a breath test, however, remain to be developed.
Read
more about this
IBM's New Product: Could it Win at Jeopardy
A machine that could truly understand human language - that could
"get" the meaning of words - would allow us to make tremendous
advances in areas such as search. So far, though search algorithms
have improved tremendously over the years, we haven't achieved this
holy grail. But IBM thinks Watson, its new computer system, could
bring us much closer. In fact, it's so sure that it plans to pit
the system against human contestants on a televised episode of the
game show Jeopardy!
The show will be hosted by Alex Trebeck, as always, who will read
the questions out loud while the text is fed into the machine. David
Ferrucci, the IBM computer scientist who leads the Watson effort,
said the system works by breaking a question down into its component
parts, searching its databases for related information, and then
putting the connections together to get the result. As an example,
he explained that the system would break the Jeopardy! Answer "It's
the opera mentioned in the lyrics of a 1970 number-one hit by Smokey
Robinson and the Miracles" into four parts initially: an opera;
an opera mentioned in a song; a song that was a hit in 1970; a hit
by Smokey Robinson and the Miracles.
In its current incarnation, Watson does not search the web; IBM hopes to sell the product to its corporate clients, who often have to find specific information in large data warehouses. Ferrucci says that Watson is much more likely to get the right answer than a system based on natural-language processing, which could be strongly influenced by keywords. Even so, that field is advancing rapidly as well, to the point that the Defense Advanced Research Projects Agency (DARPA) is launching a five-year research effort to advance the state of natural-language processing.
Read
more about this
|
