Scientists at universities across the world continue to
experiment with radical ideas in fields such as genetics, quantum
computing, robotics, chemistry and physics. Many of their scientific and
technological developments of 2015, clubbed here under five subheadings
of convenience (Cutting Edge, Science and Medicine, Robotics, The
Universe, and Technology and Society), are clearly works in progress.
Nevertheless,
all of them reflect the potential to change the way we think and work
on our planet, and the universe we live in. A handpicked collection of
some of 2015’s finest moments of epiphany.
Cutting Edge
Invisibility cloaks
While
many would dream of a cloak that makes one invisible to the world, a
true Harry Potter-like invisibility cloak is still a distant dream.
Nevertheless, a lot of research is clearly leading the way towards the
inevitability of making one.
Consider the work of Debashis Chanda at the University of Central Florida. The cover story in the March edition of the journal Advanced Optical Materials
explains how Chanda and his fellow optical and nanotech experts were
able to develop a larger swathe of multi-layer 3D metamaterial operating
in the visible spectral range.
The broad theory
behind invisibility cloaks is to manipulate light by controlling and
bending it around an object to make the latter seem invisible to the
human eye. It is the scattering of light—visible, infrared, X-ray,
etc.—that interacts with matter to help us detect and observe objects.
However, the rules that govern these interactions in natural materials
can be circumvented in metamaterials whose optical properties arise from
their physical structure rather than their chemical composition.
By
improving the technique, Chanda and his team hope to be able to create
larger pieces of the material with engineered optical properties, which
would make it practical to produce for real-life device applications.
In
April, a group of researchers from the Karlsruhe Institute of
Technology (KIT) in Karlsruhe, Germany, said they have developed a
portable invisibility cloak that can be taken into classrooms and used
for demonstrations. It can’t hide a human, but it can make small objects
disappear from sight without specialized equipment.
Scientists
hoping to divert light around an object to render it invisible must
find a way to offset the increased distance by a higher speed limit. To
address this challenge, the KIT team constructed their cloak from a
light-scattering material. By scattering light, the material slows down
the effective propagation speed of the light waves through the medium.
Then the light can be sped up again to make up for the longer path
length around the hidden object. In this cloak, the object to be
concealed is placed inside a hollow metal cylinder coated with acrylic
paint, which diffusely reflects light. The tube is embedded within a
block of polydimethylsiloxane, a commonly used organic polymer, doped
with titanium dioxide nanoparticles that make it scatter light.
There
have been similar attempts. On 17 September, for example, scientists at
the US department of energy’s Lawrence Berkeley National Laboratory
(Berkeley Lab) and the University of California Berkeley said they have
devised an ultra-thin invisibility “skin cloak” that can conform to the
shape of an object and conceal it from detection with visible light.
Working with brick-like blocks of gold nano-antennas, the Berkeley
researchers fashioned a “skin cloak” barely 80 nanometers in thickness.
The surface of the “skin cloak” was meta-engineered to reroute reflected
light waves so that the object was rendered invisible to optical
detection when the cloak is activated.
On 21
September, scientists at the Nanyang Technological University (NTU) in
Singapore said in a statement that they have developed a thermal cloak
that can render an object thermally invisible by actively redirecting
incident heat. To construct the cloak, the researchers deployed 24 small
thermoelectric modules, which are semiconductor heat pumps controlled
by an external input voltage. The modules operate via the Peltier
effect, whereby a current running through the junction between two
conductors can remove or generate heat. When many modules are attached
in series, they can redirect heat flow.
The
researchers also found that their active thermal cloaking was not
limited by the shape of the object being hidden. When applied to a
rectangular air hole, the thermoelectric devices redistributed heat just
as effectively as in the circular one. Baile Zhang and his team plan to
apply the thermal cloaks in electronic systems.
Transparent devices
If you recall all the high-tech transparent technology Tom Cruise used in Minority Report,
you may wonder why they are yet to become a reality even though it’s
well over a decade since that movie was released. It’s batteries that
pose a big problem, according to a 14 November press statement, since
they have thick materials.
With a technique
known as spin-spray layer-by-layer (SSLbL) assembly, Yale researchers
have created ultrathin and transparent films from single-walled carbon
nanotubes (SWNT) and vanadium pentoxide (V2O5) nanowires to serve as
battery anodes and cathodes—in a bid to work around the issue.
The
work was done at the lab of André Taylor, associate professor of
chemical and environmental engineering, and the results were published
online in the journal ACS Nano. Forrest Gittleson, a post-doctoral associate at Yale in chemical and environmental engineering, is the lead author.
The
researchers acknowledge that there are still challenges to overcome
before transparent devices can be mass-produced, the biggest obstacle
being “improving the conductivity of these thin electrodes”. To address
the issue, the researchers created a new “sandwich” architecture that
integrates conductive SWNT layers and active cathode materials to
enhance performance. The next step, Taylor said, is creating a
transparent separator/electrolyte—the third major component of a
battery. It’s how the lithium ions travel between the anode and cathode.
“Nature
has already demonstrated that complex systems can be transparent,”
Gittleson said. “In fact, earlier this year, they discovered a new glass
frog species with translucent skin in Costa Rica. If nature can achieve
it through evolution, we should be able to with careful engineering.”
Companies
are indeed interested in transparent devices. On 18 November 2014,
Apple Inc. was granted a patent for an invention relating to a method
and system for displaying images on a transparent display of an
electronic device. Furthermore, the display screens may allow for
overlaying of images over real-world viewable objects, as well as a
visible window to be present on an otherwise opaque display screen.
Apple credited Aleksandar Pance as the sole inventor of the granted
patent.
In June, Samsung Electronics unveiled
the first commercial use of its mirror and transparent OLEDs (organic
light emitting diodes) at the Retail Asia Expo 2015. Samsung had rolled
out the first mass-produced transparent LCD panels in 2011 and Philips’
HomeLab R&D outfit had demonstrated an LCD mirror TV in 2004.
Meanwhile,
Ubiquitous Energy, a start-up that was spun off from Massachusetts
Institute of Technology in 2014, has developed a “transparent solar cell
technology to market to eliminate the battery life limitations of
mobile devices”, according to a 28 May release by Ubiquitous Energy
co-founder and CEO Miles Barr. Implemented as a fully transparent film
that covers a device’s display area, the company’s “ClearView Power
technology” transmits light visible to the human eye, while selectively
capturing and converting ultraviolet and near-infrared light into
electricity to power the device and extend its battery life.
Face recognition
On 9 June, Microsoft Corp. launched a site, twinsornot.net,
for users to upload their photos, assuring that it would not retain the
pictures that were uploaded. A user simply could upload two photos to
assess how similar the people in these photos are, giving a score from 0
to 100. Running in the background was Microsoft’s Face API (application
programming interface) which, for instance, can detect up to 64 human
faces in an image.
Face recognition typically
provides the functionalities of automatically identifying or verifying a
person from a selection of detected faces. It is widely used in
security systems, celebrity recognition and photo tagging applications.
Optionally,
face detection can also extract a series of face-related attributes
from each face such as pose, gender and age. The inspiration was from an
April launch of Microsoft’s other site: ‘How Old Do I Look?’ (how-old.net) that lets users upload a picture and have the API predict the age and gender of any faces recognized in that picture.
Facial
recognition is not new, and technology companies are very interested in
this for obvious reasons—knowing who their users are and using that
analysis to build better and more marketable products around it.
On
20 June 2014, a University of Central Florida research team said it has
developed a facial recognition tool that promises to be useful in
rapidly matching pictures of children with their biological parents, and
potentially identifying photos of missing children as they age.
Facebook’s
DeepFace uses technology designed by an Israeli start-up called
face.com, a company that Facebook acquired in 2013. According to a paper
published in arxiv.org, a publication owned and operated by Cornell
University, in March, three researchers from Google Inc. have also
developed a similar deep neural net architecture and learning method
that also uses a facial alignment system based on explicit 3D modelling
of faces. Google calls it FaceNet.
Quantum computing
Conventional
computers use bits, the basic unit of information in computing—zeros
and ones. A quantum computer, on the other hand, deals with qubits that
can encode a one and a zero simultaneously—a property that will
eventually allow them to process a lot more information than traditional
computers, and at unimaginable speeds.
Developing
a quantum computer, however, is easier said than done. The main hurdle
is stability since calculations are taking place at the quantum level
because of which the slightest interference can disrupt the process.
Tackling the instability problem is one of the main reasons why the
effort to make a quantum computer is expensive.
The
concept of quantum computing, as conceived by American theoretical
physicist Richard Feynman in 1982, though, is not new to governments and
companies like IBM, Microsoft Corp. and Google Inc. On 2 January 2014, The Washington Post
reported that the US’s National Security Agency, or NSA, was building a
quantum computer that could break nearly every kind of encryption.
Google,
on its part, teamed up with US space agency Nasa’s Quantum Artificial
Intelligence Laboratory in August 2013 (Nasa and Google’s partnership
began 10 years ago and is not restricted to quantum computing) to work
on a machine by D-Wave Systems (whose claims of having built a pure
quantum computer are disputed since the company is not building a “gate
model” device) in the hope that quantum computing may someday
dramatically improve the agency’s ability to solve problems much faster.
On
3 September 2013, Google said it would continue to collaborate with
D-Wave scientists and experiment with the Vesuvius machine at the Nasa
Ames campus in Mountain View, California, “which will be upgraded to a
1000 qubit ‘Washington’ processor”. Bloomberg reported on 9
December this year that Google and D-Wave are developing a new computer
that can solve some types of complex problems that are next to
impossible to solve on conventional computers at Nasa Ames.
In
a related development, Princeton University researchers said they have
built a rice grain-sized laser powered by single electrons tunnelling
through artificial atoms known as quantum dots. It is being touted as a
major step towards building quantum-computing systems out of
semiconductor materials.
The researchers built
the device, which uses about one-billionth the electric current needed
to power a hair dryer, while exploring how to use quantum dots, which
are bits of semiconductor material that act like single atoms, as
components for quantum computers. “It is basically as small as you can
go with these single-electron devices,” said Jason Petta, an associate
professor of physics at Princeton who led the study, which was published
in the journal Science on 16 January.
‘Material’ computers
Manu
Prakash, an assistant professor of bioengineering at Stanford, and his
students have developed a synchronous computer that operates using the
unique physics of moving water droplets. The goal is to design a new
class of computers that can precisely control and manipulate physical
matter, according to an 8 June press statement.
The
team’s aim with the new computer, nearly a decade in the making, is not
to compete with digital computers that process information or operate
word processors but “to build a completely new class of computers that
can precisely control and manipulate physical matter”. “Imagine if when
you run a set of computations that not only information is processed but
physical matter is algorithmically manipulated as well. We have just
made this possible at the mesoscale (between micro- and macro-scale),”
Prakash said.
The ability to precisely control
droplets using fluidic computation could have a number of applications
in high-throughput biology and chemistry, and possibly new applications
in scalable digital manufacturing. The results were published in the
June edition of Nature Physics. Prakash recruited a graduate student, Georgios ‘Yorgos’ Katsikis, who is the first author on the paper.
Developing
a clock for a fluid-based computer implies it has to be easy to
manipulate, and also able to influence multiple droplets at a time, the
researchers point out. The system also needed to be scalable so that in
the future, a large number of droplets can communicate with each other.
Prakash used a rotating magnetic field to do the trick. Katsikis and
Prakash built arrays of tiny iron bars on glass slides.
Every
time the field flips, the polarity of the bars reverses, drawing the
magnetized droplets in a new, predetermined direction. Every rotation of
the field counts as one clock cycle, like a second-hand making a full
circle on a clock face, and every drop marches exactly one step forward
with each cycle.
A camera records the
interactions between individual droplets, allowing the observation of
computation as it occurs in real time. The presence or absence of a
droplet represents the 1s and 0s of binary code, and the clock ensures
that all the droplets move in perfect synchrony, and thus the system can
run virtually forever without any errors.
According
to Prakash, the most immediate application might involve turning the
computer into a high-throughput chemistry and biology laboratory.
Instead of running reactions in bulk test tubes, each droplet can carry
some chemicals and become its own test tube, and the droplet computer
offers unprecedented control over these interactions.
Science and Medicine
Cheap and fast eye check-ups
More
than 4 billion people across the world require eyeglasses. Of these,
more than half lack access to eye tests. Traditional diagnostic tools
are cumbersome, expensive and fail to take advantage of the power of
today’s mobile computing power.
This prompted a
MIT spinout, EyeNetra, to develop smartphone-powered eye-test devices.
The devices are being readied for a commercial launch with the
management wanting to introduce this device in hospitals, optometric
clinics, optical stores and even homes in the US.
EyeNetra
is also pursuing opportunities to collaborate with virtual-reality
companies seeking to use the technology to develop “vision-corrected”
virtual-reality displays. “As much as we want to solve the prescription
glasses market, we could also (help) bring virtual reality to the
masses,” said EyeNetra co-founder Ramesh Raskar, an associate professor
of media arts and sciences at the MIT Media Lab, who co-invented the
device, in a 19 October press statement. The device, called Netra, is a
plastic, binocular-like headset.
Using the
company’s app, users can attach a smartphone to the front and peer
through the headset at the phone’s display. Patterns, such as separate
red and green lines or circles, appear on the screen.
The
app calculates the difference between what a user sees as “aligned” and
the actual alignment of the patterns. This signals any refractive
errors such as nearsightedness, farsightedness and astigmatism (eye
condition that causes blurred or distorted vision), providing the
necessary information for eyeglasses prescriptions.
In
April, EyeNetra launched Blink—an on-demand refractive test service in
New York, where employees bring the start-up’s optometry tools,
including the Netra device, to people’s homes and offices. In India,
EyeNetra has launched Nayantara, a similar programme to provide low-cost
eye tests to the poor and uninsured in remote villages, far from eye
doctors.
One, of course, must mention the work
being done by Adaptive Eyecare, which was founded by Oxford Physics
professor Joshua Silver, who is now director of the non-profit Centre
for Vision in the Developing World at the University of Oxford.
During
a 2009 TED Talk, Silver demonstrated his low-cost glasses, which can be
tuned by the wearer. His spectacles have “adaptive lenses”, which
consist of two thin membranes separated by silicone gel. The wearer
simply looks at an eye chart and pumps in more or less fluid to change
the curvature of the lens, which adjusts the prescription.
A plant-based cancer drug
Elizabeth
Sattely, an assistant professor of chemical engineering at Stanford,
and her graduate student Warren Lau have isolated the machinery for
making a widely-used cancer-fighting drug from an endangered plant,
according to a 10 September press statement.
They
then put that machinery into a common, easily grown laboratory plant in
the Himalayas, which was able to produce the chemical. The drug Sattely
chose to focus on is produced by a leafy Himalayan plant called the
Mayapple.
Within the plant, a series of proteins
work in a step-by-step fashion to churn out a chemical defence against
predators. That chemical defence, after a few modifications in the lab,
becomes a widely-used cancer drug called etoposide. The starting
material for this chemical defence is a harmless molecule commonly
present in the leaf. When the plant senses an attack, it begins
producing proteins that make up the assembly line.
One
by one, those proteins add a little chemical something here, subtract
something there, and after a final molecular nip and tuck, the harmless
starting material is transformed into a chemical defence. The challenge
was figuring out the right proteins for the job.
Sattely
and her team tested various combinations of 31 proteins until they
eventually found 10 that made up the full assembly line. They put genes
that make those 10 proteins into a common laboratory plant, and that
plant began producing the chemical they were seeking.
The
eventual goal is not simply moving molecular machinery from plant to
plant. Now that she’s proven the molecular machinery works outside the
plant, Sattely wants to put the proteins in yeast, which can be grown in
large vats in the lab to better provide a stable source of drugs.
The
technique could potentially be applied to other plants and drugs,
creating a less expensive and more stable source for those drugs, the
researchers say. Sattely’s work was published on 10 September in the
journal Science.
3D-printed heart model may do away with transplants
Work
by a group at Carnegie Mellon could one day lead to a world in which
transplants are no longer necessary to repair damaged organs.
“We
have been able to take MRI (magnetic resonance imaging) images of
coronary arteries and 3D images of embryonic hearts and 3D bioprint them
with unprecedented resolution and quality out of very soft materials
like collagens, alginates and fibrins,” said Adam Feinberg, associate
professor of materials science and engineering and biomedical
engineering at Carnegie Mellon University, in a press statement.
Feinberg
leads the regenerative biomaterials and therapeutics group, and the
group’s study was published in the 23 October issue of the journal Science Advances.
Traditional 3D printers build hard objects typically made of plastic or
metal, and they work by depositing material onto a surface layer by
layer to create the 3D object.
Printing each
layer requires sturdy support from the layers below, so printing with
soft materials like gels has been limited. The challenge with soft
materials is that they collapse under their own weight when 3D printed
in air, explained Feinberg. So, the team developed a method of printing
soft materials inside a support bath material.
“Essentially,
we print one gel inside of another gel, which allows us to accurately
position the soft material as it’s being printed, layer by layer,” he
said.
One of the major advances of this
technique, termed FRESH, or Freeform Reversible Embedding of Suspended
Hydrogels, is that the support gel can be easily melted away and removed
by heating to body temperature, which does not damage the delicate
biological molecules or living cells that were bioprinted.
As
a next step, the group is working towards incorporating real heart
cells into these 3D printed tissue structures, providing a scaffold to
help form contractile muscle. Bioprinting is a growing field, but to
date, most 3D bioprinters cost over $100,000 and/or require specialized
expertise to operate, limiting wider-spread adoption.
Feinberg’s
group, however, has been able to implement their technique on a range
of consumer-level 3D printers, which cost less than $1,000 by utilizing
open-source hardware and software.The 3D printer designs are being
released under an open-source licence.
Can we regrow teeth?
Why
can’t humans regrow teeth lost to injury or disease the way nature
does? By studying how structures in embryonic fish differentiate into
either teeth or taste buds, Georgia Tech researchers hope to one day be
able to turn on the tooth regeneration mechanism in humans.
The
research was conducted by scientists from the Georgia Institute of
Technology in Atlanta and King’s College in London, and published on 19
October in the journal Proceedings of the National Academy of Sciences.
The
studies in fish and mice, according to the researchers, suggest the
possibility that “with the right signals, epithelial tissue in humans
might also be able to regenerate new teeth”.
“We
have uncovered developmental plasticity between teeth and taste buds,
and we are trying to understand the pathways that mediate the fate of
cells towards either dental or sensory development,” said Todd
Streelman, a biology professor at Georgia Tech.
But
growing new teeth wouldn’t be enough, Streelman cautions. Researchers
would also need to understand how nerves and blood vessels grow into
teeth to make them viable.
“The exciting aspect
of this research for understanding human tooth development and
regeneration is being able to identify genes and genetic pathways that
naturally direct continuous tooth and taste bud development in fish, and
study these in mammals,” said professor Paul Sharpe, a co-author from
King’s College.
New way to fix a broken heart?
Coronary
artery disease is the leading cause of death worldwide, but there is
currently no effective method to regenerate new coronary arteries in
diseased or injured hearts. Stanford researchers, according to a 19
October study published in the journal eLife, have identified a progenitor cell type that could make it possible.
The
study was carried out with mice but, as the blood vessels of the human
heart are similar, it could lead to new treatments for the disease or to
restore blood flow after a heart attack, the researchers say.
“Current
methods to grow new blood vessels in the heart stimulate fine blood
vessels rather than re-establishing the strong supply of blood provided
by the main arteries. We need arteries to restore normal function,” said
senior author Kristy Red-Horse from the department of biological
sciences at Stanford.
The researchers reveal
that the smooth muscle of the arteries is derived from cells called
pericytes. The small capillary blood vessels throughout the developing
heart are covered in pericytes. Pericytes are also found throughout the
adult heart, which suggests that they could be used to trigger a
self-repair mechanism.
A problem with cell or
tissue transplantation can be that the cells don’t integrate or they
differentiate into slightly different cells types than intended. As
pericytes are spread all over the heart on all the small blood vessels,
they could be used as a target to stimulate artery formation without the
need for transplantation, the researchers point out.
The
team is now investigating whether pericytes differentiate into smooth
muscle as part of this process and whether it can be activated or sped
up by introducing Notch 3 (a protein) signalling molecules.
“Now
that we are beginning to really understand coronary artery development,
we have initiated studies to reactivate it in injury models and hope to
some day use these same methods to help treat coronary artery disease,”
said Red-Horse.
Measuring the ageing process
There
are times when we intuitively feel that even people born within months
of each other are ageing differently. Indeed they are, say the
researchers of a long-term health study in New Zealand that sought clues
to the ageing process in young adults.
In a paper appearing the week of 6 July in the Proceedings of the National Academy of Sciences,
the team from the US, UK, Israel and New Zealand introduced a panel of
18 biological measures (known as biomarkers) that may be combined to
determine whether people are ageing faster or slower than their peers.
The
data came from the Dunedin Study, a landmark longitudinal study that
has tracked more than a thousand people born in 1972-73 in the same town
from birth. Health measures like blood pressure and liver function were
taken regularly, along with interviews and other assessments.
According
to first author Dan Belsky, an assistant professor of geriatrics at
Duke University’s Centre for Ageing, the progress of ageing shows in
human organs just as it does in eyes, joints and hair—but sooner.
Based
on a subset of these biomarkers, the research team set a “biological
age” for each participant, which ranged from under 30 to nearly 60 in
the 38-year olds. Most participants clustered around an ageing rate of
one year per year, but others were found to be ageing as fast as three
years per chronological year.
As the team
expected, those who were biologically older at age 38 also appeared to
have been ageing at a faster pace. A biological age of 40, for example,
meant that the person was ageing at a rate of 1.2 years per year over
the 12 years the study examined.
The ageing
process, according to the researchers, isn’t all genetic. Studies of
twins have found that only about 20% of ageing can be attributed to
genes, Belsky said. “There’s a great deal of environmental influence,”
he added.
This gives “us some hope that medicine
might be able to slow ageing and give people more healthy active
years”, said senior author Terrie Moffitt, the Nannerl O Keohane
professor of psychology and neuroscience at Duke.
Sunblock that stays on the outside
Researchers
at Yale University have developed a sunscreen that doesn’t penetrate
the skin, eliminating serious health concerns associated with commercial
sunscreens, according to a 28 September news release.
Most
commercial sunblocks may prevent sunburn, but they can go below the
skin’s surface and enter the bloodstream, and are likely to trigger
hormonal side-effects and could even be promote the kind of skin cancers
they are designed to prevent.
The new sunblock
made by the Yale researchers uses bio-adhesive nanoparticles that stay
on the surface of the skin. The results of the research appeared in the
28 September online edition of the journal Nature Materials. Using mouse models, the researchers tested their sunblock against direct ultraviolet rays and their ability to cause sunburn.
Altering brain chemistry to raise the pain threshold
Scientists
at the University of Manchester have shown for the first time that the
numbers of opiate (akin to sedatives) receptors in the brain increases
to combat severe pain in those suffering from arthritis.
Receptors
in our brains respond to natural painkilling opiates such as
endorphins, but the researchers in Manchester have now shown that these
receptors increase in number to help cope with long-term, severe pain.
By applying heat to the skin using a laser stimulator, Dr Christopher
Brown and his colleagues showed that the more opiate receptors there are
in the brain, the higher the ability to withstand the pain, according
to a 23 October study.
The researchers used
positron emission tomography (PET) imaging on 17 patients with arthritis
and nine healthy controls to show the spread of the opioid receptors.
Val
Derbyshire, a patient with arthritis, said in a press statement: “As a
patient who suffers chronic pain from osteoarthritis, I am extremely
interested in this research. I feel I have developed coping mechanisms
to deal with my pain over the years, yet still have to take opioid
medication to relieve my symptoms.”
A cranial fingerprint?
A
person’s brain activity appears to be as unique as his or her
fingerprints, a new Yale University-led imaging study shows. These brain
“connectivity profiles” alone allow researchers to identify individuals
from the functional magnetic resonance imaging (fMRI) images of brain
activity of more than 100 people, according to the study published on 12
October in the journal Nature Neuroscience.
The
researchers compiled fMRI data from 126 subjects who underwent six scan
sessions over two days. Subjects performed different cognitive tasks
during four of the sessions. In the other two, they simply rested.
Researchers looked at activity in 268 brain regions—specifically,
coordinated activity between pairs of regions.
Highly
coordinated activity implies two regions are functionally connected.
Using the strength of these connections across the whole brain, the
researchers were able to identify individuals from fMRI data alone,
whether the subject was at rest or engaged in a task. They were also
able to predict how subjects would perform on tasks.
The
researchers hope that this ability might one day help clinicians
predict or even treat neuro-psychiatric diseases based on individual
brain connectivity profiles. Data for the study came from the Human
Connectome Project led by the WU-Minn Consortium, which is funded by the
16 National Institutes of Health (NIH) Institutes and Centers that
support the NIH Blueprint for Neuroscience Research and by the McDonnell
Center for Systems Neuroscience at Washington University.
Robotics
Robots that crowdsource learning
In
July, scientists from Cornell University led by Ashutosh Saxena
announced the development of a Robo Brain—a large computational system
that learns from publicly available Internet resources.
The
system, according to a 25 August statement by Cornell, was downloading
and processing about 1 billion images, 120,000 YouTube videos and 100
million how-to documents and appliance manuals. Information from the
system, which Saxena had described at the 2014 Robotics: Science and
Systems Conference in Berkeley, is being translated and stored in a
robot-friendly format that robots will be able to draw on when needed.
The India-born, Indian Institute of Technology Kanpur graduate, also launched a website for the project at robobrain.me,
which displays things the brain has learnt, and visitors are able to
make additions and corrections. Robo Brain employs what computer
scientists call structured deep learning, where information is stored in
many levels of abstraction. Deep learning is a set of algorithms, or
instruction steps for calculations, in machine learning.
There
have been similar attempts to make computers understand context and
learn from the Internet. For instance, since January 2010, scientists at
the Carnegie Mellon University have been working to build a
never-ending machine learning system that acquires the ability to
extract structured information from unstructured Web pages.
If
successful, the scientists say it will result in a knowledge base (or
relational database) of structured information that mirrors the content
of the Web. They call this system the never-ending language learner, or
NELL.
We also have IBM’s Watson, which beat
Jeopardy players in 2011, and now has joined hands with the US
Automobile Association to help members of the military prepare for
civilian life. In January 2014, IBM said it will spend $1 billion to
launch the Watson Group, including a $100 million venture fund to
support start-ups and businesses that are building Watson-powered apps
using the “Watson Developers Cloud”.
Can robots, or cobots, make good teammates?
At
the Yale Social Robotics Lab, run by professor of computer science
Brian Scassellati, robots are learning the skills needed to be good
teammates, allowing people to work more safely, more efficiently and
more effectively. The skills include stabilizing parts, handing over
items, organizing a workspace, or helping people use a tool better,
according to a 21 September press release.
Sharing a workspace with most robots can be dangerous, the researchers point out.
“It’s
only now that this is becoming feasible, to develop robots that could
safely operate near and around people,” said Brad Hayes, the PhD
candidate who headed the project. “We are trying to move robots away
from being machines in isolation, developing them to be co-workers that
amplify the strengths and abilities of each member of the team they are
on.”
One way of building team skills is by
guiding the robot to figure out how to help during tasks by simulating
hundreds of thousands of different possibilities and then guessing if
that’s going to be helpful. Given that such tasks tend to take very
long, the researchers suggest that the other approach is to show the
robot directly how to build team skills.
“Here
you are naturally demonstrating to the robot and having it retain that
knowledge,” Hayes explained. “It can then save that example and figure
out if it’s a good idea to generalize that skill to use in new
situations.”
Hayes thinks the technology has
value for both the workplace and the home, particularly for small-scale,
flexible manufacturing or for people who have lost some of their
autonomy and could use help with the dishes or other chores.
Collaborative
robots, also known as cobots, is the new buzzword in robotics. They are
complementary to industrial robots. Safely working alongside humans in
an uncaged environment, they are opening up new opportunities for
industry. They have to be safe, easy to use, flexible and affordable.
Examples include the Baxter robot by Rethink Robotics; the UR5 arm by
Universal Robots and Robonaut2 (by GE).
And, of
course, the YuMi robot that is short for ‘you and me’. It was unveiled
by industrial robotics company ABB at the Hannover Messe on 13 April.
ABB touts it as the world’s first truly collaborative robot, able to
work side-by-side on the same tasks as humans while still ensuring the
safety of those around it.
The company says
YuMi is “capable of handling anything from a watch to a tablet PC and
with the level of accuracy that could thread a needle”.
Universe
Earth to Mars via a petrol station on the moon
Living
on Mars will certainly not be an easy task for human beings. They will
have to tackle major issues such as higher radiation, lack of a
semblance of an atmosphere, lower gravity pull that can affect our
skeletal structure, likely infection from unknown microbes, lack of food
and the effect of loneliness on the mind. But first, they have to reach
Mars, a journey that will take about 180 days. This means they will
need enough fuel, or will have to refuel somewhere.
Studies
have suggested that lunar soil and water ice in certain craters of the
moon may be mined and converted to fuel. Assuming that such technologies
are established at the time of a mission to Mars, a 14 October MIT
study has found that taking a detour to the moon to refuel would reduce
the mass of a mission upon launch by 68%.
The
researchers developed a model to determine the best route to Mars,
assuming the availability of resources and fuel-generating
infrastructure on the moon. Based on their calculations, they found the
most mass-efficient path involves launching a crew from Earth with just
enough fuel to get into orbit around the Earth.
A
fuel-producing plant on the surface of the moon would then launch
tankers of fuel into space, where they would enter gravitational orbit.
The tankers would eventually be picked up by the Mars-bound crew, which
would then head to a nearby fuelling station to gas up before ultimately
heading to Mars.
Olivier de Weck, a professor
of aeronautics and astronautics and of engineering systems at MIT, says
the plan deviates from Nasa’s more direct “carry-along” route.
The results, which are based on the PhD thesis of Takuto Ishimatsu, now a post-doctorate student at MIT, are published in the Journal of Spacecraft and Rockets.
Ishimatsu’s network flow model to explore various routes to
Mars—ranging from a direct carry-along flight to a series of refuelling
pit stops along the way—assumes a future scenario in which fuel can be
processed on, and transported from, the moon to rendezvous points in
space.
When did life on Earth begin?
Geochemists
at the University of California, Los Angeles (UCLA) have found evidence
that life likely existed on Earth at least 4.1 billion years ago—300
million years earlier than previous research suggested, according to a
19 October press release.
The discovery
indicates that life may have begun shortly after the planet formed 4.54
billion years ago. The research was published in the online edition of
the journal Proceedings of the National Academy of Sciences.
“Twenty years ago, this would have been heretical; finding evidence of life 3.8 billion years ago was shocking,” said Mark Harrison, co-author of the research and a professor of geochemistry at UCLA. The new research suggests that life existed prior to the massive bombardment of the inner solar system that formed the moon’s large craters 3.9 billion years ago.
Scientists had long believed the Earth was
dry and desolate during that time period. Harrison’s research—including a
2008 study in Nature he co-authored with Craig Manning, a
professor of geology and geochemistry at UCLA, and former UCLA graduate
student Michelle Hopkins—is proving otherwise.
The
researchers, led by Elizabeth Bell—a post-doctoral scholar in
Harrison’s laboratory—studied more than 10,000 zircons originally formed
from molten rocks, or magmas, from Western Australia.
Zircons
are heavy, durable minerals related to the synthetic cubic zirconium
used for imitation diamonds. They capture and preserve their immediate
environment, meaning they can serve as time capsules.
The
scientists identified 656 zircons containing dark specks that could be
revealing and closely analysed 79 of them with Raman spectroscopy—a
technique that shows the molecular and chemical structure of ancient
microorganisms in three dimensions.
One of the
79 zircons contained graphite (pure carbon) in two locations. The
graphite is older than the zircon containing it, the researchers said.
They know the zircon is 4.1 billion years old, based on its ratio of
uranium to lead but they don’t know how much older the graphite is.
Water on Mars!
Researchers
have discovered an enormous slab of ice just beneath the surface of
Mars, measuring 130 feet thick and covering an area equivalent to that
of California and Texas combined. The ice may be the result of snowfall
tens of millions of years ago on Mars, scientists said in a 15 September
press statement. The research was published in the journal Geophysical Research Letters.
Combining
data gleaned from two powerful instruments aboard Nasa’s Mars
Reconnaissance Orbiter, or MRO, researchers determined why a
“crazy-looking crater” on Mars’ surface is terraced, and not bowl-shaped
like most craters of this size.
Although
scientists have known for some time about Mars’s icy deposits at its
poles and have used them to look at its climatic history, knowledge of
icy layers at the planet’s mid-latitudes, analogous to earthly latitudes
falling between the Canadian-US border and Kansas, is something new.
On
16 October, Nasa confirmed that new findings from MRO had provided “the
strongest evidence yet that liquid water flows intermittently on
present-day Mars”. The findings came five months after scientists found
the first evidence for liquid water on the red planet.
The
findings are in sync with Nasa’s ambitious project to send humans to
Mars in the 2030s, in accordance with the Nasa Authorization Act, 2010,
and the US National Space Policy. The discovery of liquid water,
therefore, could be a big boost for astronauts visiting the planet.
Based
on further research and findings, humans could well be drinking water
on Mars, or use it for creating oxygen and rocket fuel, or to water
plants in greenhouses.
Most Earth-like worlds yet to be born. Boo
When
our solar system was born 4.6 billion years ago only 8% of the
potentially habitable planets that will ever form in the universe
existed. The bulk of those planets—92%—are yet to be born, so will
continue to do so much after the sun burns out 6 billion years hence.
This conclusion is based on an assessment of data collected by Nasa’s
Hubble Space Telescope and the prolific planet-hunting Kepler space
observatory.
“Our main motivation was
understanding the Earth’s place in the context of the rest of the
universe,” said the 20 October study’s author Peter Behroozi of the
Space Telescope Science Institute (STScI) in Baltimore, Maryland.
The
data show that the universe was making stars at a fast rate 10 billion
years ago, but the fraction of the universe’s hydrogen and helium gas
that was involved was very low. Today, star birth is happening at a much
slower rate than long ago, but there is so much leftover gas available
that the universe will keep cooking up stars and planets for a very long
time to come.
Kepler’s planet survey indicates
that Earth-sized planets in a star’s habitable zone, the perfect
distance that could allow water to pool on the surface, are ubiquitous
in our galaxy.
Based on the survey, scientists
predict that there should be 1 billion Earth-sized worlds in the Milky
Way galaxy at present, a good portion of them presumed to be rocky. That
estimate skyrockets when you include the other 100 billion galaxies in
the observable universe.
This leaves plenty of
opportunity for untold more Earth-sized planets in the habitable zone to
arise in the future. The last star isn’t expected to burn out until 100
trillion years from now. That’s plenty of time for literally anything
to happen on the planet landscape, the researchers say.
Ocean rise: How much? How soon? Oh no
Seas
around the world have risen an average of nearly 3 inches since 1992,
with some locations rising more than 9 inches due to natural variation,
according to the latest satellite measurements from Nasa and its
partners.
In 2013, the UN Intergovernmental
Panel on Climate Change issued an assessment based on a consensus of
international researchers that stated global sea levels would likely
rise from 1 to 3 feet by the end of the century.
The
new data, according to a 26 August Nasa press statement, reveal that
the height of the sea surface is not rising uniformly everywhere.
Regional differences in sea level rise are dominated by the effects of
ocean currents and natural cycles such as the Pacific Decadal
Oscillation.
But, as these natural cycles wax
and wane, they can have major impacts on local coastlines. Scientists
estimate that about one-third of sea level rise is caused by the
expansion of warmer ocean water, one-third is due to ice loss from the
massive Greenland and Antarctic ice sheets, and the remaining third
results from melting mountain glaciers.
However, the fate of the polar ice sheets could change that ratio and produce more rapid increases in the coming decades.
Technology and Society
Plastic-eating worms that can annihilate waste
The
world over, people throw away billions of plastic cups and a very small
percentage of that gets recycled. How does one tackle this plastic
menace? With the help of a mealworm, say Stanford researchers.
A
mealworm, the larvae form of the darkling beetle, can subsist on a diet
of styrofoam and other forms of polystyrene, according to two companion
studies co-authored by Wei-Min Wu, a senior research engineer in the
department of civil and environmental engineering at Stanford. It’s the
microorganisms in the mealworm’s gut that biodegrade the plastic in the
process.
The papers, published in Environmental Science and Technology in September, are the first to provide detailed evidence of bacterial degradation of plastic in an animal’s gut.
“There’s
a possibility of really important research coming out of bizarre
places,” said Craig Criddle, a professor of civil and environmental
engineering who supervises plastics research by Wu and others at
Stanford. “Sometimes, science surprises us. This is a shock.”
The
new research on mealworms is significant because styrofoam was thought
to have been non-biodegradable and more problematic for the environment.
Researchers led by Criddle, a senior fellow at the Stanford Woods
Institute for the Environment, are collaborating on ongoing studies with
the project leader and papers’ lead author, Jun Yang of Beihang
University in China, and other Chinese researchers.
Together,
they plan to study whether microorganisms within mealworms and other
insects can biodegrade plastics such as polypropylene (used in products
ranging from textiles to automotive components), microbeads (tiny bits
used as exfoliants) and bioplastics (derived from renewable biomass
sources such as corn or biogas methane).
The
researchers plan to explore the fate of these materials when consumed by
small animals, which are, in turn, consumed by other animals. Another
area of research could involve searching for a marine equivalent of the
mealworm to digest plastics, Criddle said. Plastic waste is a particular
concern in the ocean, where it fouls habitats and kills countless
seabirds, fish, turtles and other marine life.
Fresh milk, off the grid
How
does one preserve milk? Most of us do this by refrigeration and
boiling, but what does one do if there’s sporadic electricity? The
answers may lie in a 19 May study by Tel Aviv University (TAU)
researchers.
Published in the journal Technology,
the study finds that short-pulsed electric fields can be used to kill
milk-contaminating bacteria. Through a process called electroporation,
bacterial cell membranes are selectively damaged.
According
to lead investigator Alexander Golberg of TAU’s Porter School of
Environmental Studies, applying this process intermittently prevents
bacteria proliferation in stored milk, potentially increasing its shelf
life.
According to the study, pulsed electric
fields, an emerging technology in the food industry that has been shown
to effectively kill multiple food-born microorganisms, could provide an
alternative, non-thermal pasteurization process.
The
stored milk is periodically exposed to high-voltage, short pulsed
electric fields that kill the bacteria. The energy required can come
from conventional sources or from the sun. The technology is three times
more energy-efficient than boiling and almost twice as energy efficient
as refrigeration, the researchers say.
Crowdsourcing interactive story plots with AI
Researchers
at the Georgia Institute of Technology have developed an artificial
intelligence (AI) system that crowdsources plots for interactive
stories. While current AI models for games have a limited number of
scenarios and depend on a data set already programmed into a model by
experts, Georgia Tech’s AI system generates numerous scenes for players
to adopt.
“Our open interactive narrative system
learns genre models from crowdsourced example stories so that the
player can perform different actions and still receive a coherent story
experience,” Mark Riedl, lead investigator and associate professor of
interactive computing at Georgia Tech, said in a 19 September news
release.
A test of the AI system, called
Scheherazade IF (Interactive Fiction)—a reference to the fabled Persian
queen and storyteller—showed that it can achieve near human-level
authoring.
“When enough data is available and
that data sufficiently covers all aspects of the game experience, the
system was able to meet or come close to meeting human performance in
creating a playable story,” Riedl added.
The
researchers evaluated the AI system by measuring the number of “common
sense” errors (e.g. scenes out of sequence) found by players, as well as
players’ subjective experiences for things such as enjoyment and
coherence of story. The creators say that they are seeking to inject
more creative scenarios into the system.
Right
now, AI plays it safe with the crowdsourced content, producing what one
might expect in different genres. But opportunities exist to train
Scheherazade (just like its namesake implies) to surprise and immerse
those in future interactive experiences.
The
impact of this research can support not only online storytelling for
entertainment, but also digital storytelling used in online course
education or corporate training. The research paper Crowdsourcing Open Interactive Narrative
(co-authored by Matthew Guzdial, Brent Harrison, Boyang Li and Mark
Riedl) was presented at the 2015 Foundations of Digital Games Conference
in Pacific Grove, California.
Teaching computers to ‘see’ what humans do
Researchers
from Georgia Tech’s school of interactive computing and institute for
robotics and intelligent machines have developed a new method that
teaches computers to “see” and understand what humans do in a typical
day, according to a 28 September news release.
The
researchers gathered more than 40,000 pictures taken every 30 to 60
seconds, over a six-month period, by a wearable camera and predicted
with 83% accuracy what activity that person was doing. The idea,
according to the researchers, is to give users the ability to be able to
track all of their activities—not just physical ones like walking and
running, which most wearables like Fitbit do.
The ability to literally see and recognize human activities has implications in a number of areas—from developing improved personal assistant applications like Siri to helping researchers explain links between health and behaviour, the researchers say.
No comments:
Post a Comment