Computers Math

The quest for high throughput intelligent computing paradigms — for big data and artificial intelligence — and the ever-increasing volume of digital information has led to an intensified demand for high-speed and low-power consuming next-generation electronic devices. The “forgotten” world of antiferromagnets (AFM), a class of magnetic materials, offers promise in future electronic device development and complements present-day ferromagnet-based spintronic technologies.
Formidable challenges for AFM-based functional spintronic device development are high-speed electrical manipulation (recording), detection (retrieval), and ensuring the stability of the recorded information — all in a semiconductor industry-friendly material system.
Researchers at Tohoku University, University of New South Wales (Australia), ETH Zürich (Switzerland), and Diamond Light Source (United Kingdom) successfully demonstrated current-induced switching in a polycrystalline metallic antiferromagnetic heterostructure with high thermal stability. The established findings show potential for information storage and processing technologies.
The research group used a Mn-based metallic AFM (PtMn)/heavy metal (HM) heterostructure — attractive because of its significant antiferromagnetic anisotropy and its compatibility with PtMn Silicon-based electronics. Electrical recording of resistance states (1 or 0) was obtained through the spin-orbit interaction of the HM layer; a charge current in the adjacent HM resulted in spin-orbit torques acting on the AFM, leading to a change in the resistance level down to a microsecond regime.
“Interestingly, the switching degree is controllable by the strength of the current in the HM layer and shows long-term data retention capabilities,” said Samik DuttaGupta, corresponding author of the study. “The experimental results from electrical measurements were supplemented by a magnetic X-ray imaging, helping to clarify the reversible nature of switching dynamics localized within nm-sized AFM domains.”
The results are the first demonstration of current-induced switching of an industry-compatible AFM down to the microsecond regime within the field of metallic antiferromagnetic spintronics. These findings are expected to initiate new avenues for research and encourage further investigations towards the realization of functional devices using metallic AFMs for information storage and processing technologies.

Story Source:
Materials provided by Tohoku University. Note: Content may be edited for style and length. More

Current research and applications in the field of artificial intelligence (AI) include several key challenges. These include: (a) A priori estimation of the required dataset size to achieve a desired test accuracy. For example, how many handwritten digits does a machine have to learn before being able to predict a new one with a success rate of 99%? Similarly, how many specific types of circumstances does an autonomous vehicle have to learn before its reaction will not lead to an accident? (b) The achievement of reliable decision-making under a limited number of examples, where each example can be trained only once, i.e., observed only for a short period. This type of realization of fast on-line decision making is representative of many aspects of human activity, robotic control and network optimization.
In an article published today in the journal Scientific Reports, researchers show how these two challenges are solved by adopting a physical concept that was introduced a century ago to describe the formation of a magnet during a process of iron bulk cooling.
Using a careful optimization procedure and exhaustive simulations, a group of scientists from Bar-Ilan University has demonstrated the usefulness of the physical concept of power-law scaling to deep learning. This central concept in physics, which arises from diverse phenomena, including the timing and magnitude of earthquakes, Internet topology and social networks, stock price fluctuations, word frequencies in linguistics, and signal amplitudes in brain activity, has also been found to be applicable in the ever-growing field of AI, and especially deep learning.
“Test errors with online learning, where each example is trained only once, are in close agreement with state-of-the-art algorithms consisting of a very large number of epochs, where each example is trained many times. This result has an important implication on rapid decision making such as robotic control,” said Prof. Ido Kanter, of Bar-Ilan’s Department of Physics and Gonda (Goldshmied) Multidisciplinary Brain Research Center, who led the research. “The power-law scaling, governing different dynamical rules and network architectures, enables the classification and hierarchy creation among the different examined classification or decision problems,” he added.
“One of the important ingredients of the advanced deep learning algorithm is the recent new bridge between experimental neuroscience and advanced artificial intelligence learning algorithms,” said PhD student Shira Sardi, a co-author of the study. Our new type of experiments on neuronal cultures indicate that an increase in the training frequency enables us to significantly accelerate the neuronal adaptation process. “This accelerated brain-inspired mechanism enables building advanced deep learning algorithms which outperform existing ones,” said PhD student Yuval Meir, another co-author.
The reconstructed bridge from physics and experimental neuroscience to machine learning is expected to advance artificial intelligence and especially ultrafast decision making under limited training examples as to contribute to the formation of a theoretical framework of the field of deep learning.

Story Source:
Materials provided by Bar-Ilan University. Note: Content may be edited for style and length. More

Think about your favorite t-shirt, the one you’ve worn a hundred times, and all the abuse you’ve put it through. You’ve washed it more times than you can remember, spilled on it, stretched it, crumbled it up, maybe even singed it leaning over the stove once.
We put our clothes through a lot and if the smart textiles of the future are going to survive all that we throw at them, their components are going to need to be resilient.
Now, researchers from the Harvard John A. Paulson School of Engineering and Applied Sciences and the Wyss Institute for Biologically Inspired Engineering have developed an ultra-sensitive, seriously resilient strain sensor that can be embedded in textiles and soft robotic systems.
The research is published in Nature.
“Current soft strain gauges are really sensitive but also really fragile,” said Oluwaseun Araromi, a Research Associate in Materials Science and Mechanical Engineering at SEAS and the Wyss Institute and first author of the paper. “The problem is that we’re working in an oxymoronic paradigm — highly sensitivity sensors are usually very fragile and very strong sensors aren’t usually very sensitive. So, we needed to find mechanisms that could give us enough of each property.”
In the end, the researchers created a design that looks and behaves very much like a Slinky.

advertisement

“A Slinky is a solid cylinder of rigid metal but if you pattern it into this spiral shape, it becomes stretchable,” said Araromi. “That is essentially what we did here. We started with a rigid bulk material, in this case carbon fiber, and patterned it in such a way that the material becomes stretchable.”
The pattern is known as a serpentine meander, because its sharp ups and downs resemble the slithering of a snake. The patterned conductive carbon fibers are then sandwiched between two prestrained elastic substrates.
The overall electrical conductivity of the sensor changes as the edges of the patterned carbon fiber come out of contact with each other, similar to the way the individual spirals of a slinky come out of contact with each other when you pull both ends. This process happens even with small amounts of strain, which is the key to the sensor’s high sensitivity.
Unlike current highly sensitive stretchable sensors, which rely on exotic materials such as silicon or gold nanowires, this sensor doesn’t require special manufacturing techniques or even a clean room. It could be made using any conductive material.
The researchers tested the resiliency of the sensor by stabbing it with a scalpel, hitting it with a hammer, running it over with a car, and throwing it in a washing machine ten times. The sensor emerged from each test unscathed.

advertisement

To demonstrate its sensitivity, the researchers embedded the sensor in a fabric arm sleeve and asked a participant to make different gestures with their hand, including a fist, open palm, and pinching motion. The sensors detected the small changes in the subject’s forearm muscle through the fabric and a machine learning algorithm was able to successfully classify these gestures.
“These features of resilience and the mechanical robustness put this sensor in a whole new camp,” said Araromi.
Such a sleeve could be used in everything from virtual reality simulations and sportswear to clinical diagnostics for neurodegenerative diseases like Parkinson’s Disease.
Harvard’s Office of Technology Development has filed to protect the intellectual property associated with this project.
“The combination of high sensitivity and resilience are clear benefits of this type of sensor,” said Robert Wood, the Charles River Professor of Engineering and Applied Sciences at SEAS and senior author of the study. “But another aspect that differentiates this technology is the low cost of the constituent materials and assembly methods. This will hopefully reduce the barriers to get this technology widespread in smart textiles and beyond.”
“We are currently exploring how this sensor can be integrated into apparel due to the intimate interface to the human body it provides,” says Conor Walsh, the Paul A. Maeder Professor of Engineering and Applied Sciences at SEAS and co-author of the study. “This will enable exciting new applications by being able to make biomechanical and physiological measurements throughout a person’s day, not possible with current approaches.”
The research was co-authored by Moritz A. Graule, Kristen L. Dorsey, Sam Castellanos, Jonathan R. Foster, Wen-Hao Hsu, Arthur E. Passy, James C. Weaver, Senior Staff Scientist at SEAS and Joost J. Vlassak, the Abbott and James Lawrence Professor of Materials Engineering at SEAS.
It was funded through the university’s strategic research alliance with Tata. The 6-year, $8.4M alliance was established in 2016 to advance Harvard innovation in fields including robotics, wearable technologies, and the internet of things (IoT). More

The effects of climate change are sometimes difficult to grasp, but now a virtual reality forest, created by geographers, can let people walk through a simulated forest of today and see what various futures may hold for the trees.
“The main problem that needs to be addressed is that climate change is abstract,” said Alexander Klippel, professor of geography, Penn State. “Its meaning only unfolds in 10, 15 or 100 years. It is very hard for people to understand and plan and make decisions.”
The researchers combined information on forest composition with information on forest ecology to create a forest similar to those found in Wisconsin.
“As part of an NSF-funded CNH program grant with Erica Smithwick (E. Willard and Ruby S. Miller Professor of Geography at Penn State) we are working with the Menominee Indian Tribe of Wisconsin,” said Klippel, who also is director of Penn State’s Center for Immersive Experience. “Inspired by the Menominee’s deeper connection to the environment we believe that experiencing the future is essential for all environmental decision making.”
The virtual-reality experience takes the extensive climate change models, sophisticated vegetation models and ecological models and creates a 2050 forest that people can experience by walking through it, investigating the tree types and understory, and seeing the changes.
Visualizing Forest Futures.
The first step, of course, was to create a forest of today. Using data on a typical Wisconsin forest, the researchers could have used strict or deterministic rules and placed trees in the forest. However, they chose to use a procedural method that would populate the forest using a set of ecological rules, creating a more organic, natural feel.

advertisement

“Orientation and small details of the trees are also randomized in the approach so that the trees don’t look exactly the same,” said Jiawei Huang, graduate student in geography, Penn State.
The researchers report today (Nov. 11) in the International Journal of Geographical Information Science that, “Procedural rules allowed us to efficiently and reproducibly translate the parameters into a simulated forest.” They used analytical modeling to convert the data for procedural modeling. They also worked with ecological experts to provide feedback and evaluate the results.
To capture the ecology of the forest, the researchers used LANDIS II, a well-established, powerful model.
“Our ecologist colleagues, coauthors on this paper — Melissa S. Lucash, research assistant professor of geography, University of Oregon, and Robert M. Scheller, professor of geography, North Carolina State University — ensured the expertise that is necessary to make the predictions accurate,” said Klippel.
The researchers note that the model is powerful enough to deal with events such as windstorms, fire and flooding, and, of course, climate change.

advertisement

A virtual walk through this Wisconsin forest shows tall trees and understory. Strollers, using VR headsets and controllers, can reveal the types of trees in the forest, change elevations from forest floor to birds-eye view and in-between, and more closely examine the forest composition.
The researchers chose two future scenarios, a base scenario and a hot and dry scenario. Using VR, visitors to the forest can see the changes in tree types and abundance and compare the base scenario to the hot and dry scenario.
“Our approach to create visceral experiences of forests under climate change can facilitate communication among experts, policymakers and the general public,” the researchers report.
The researchers aim is to create a medium to communicate things in the future or the past that allows for a more holistic and visceral access so that non-experts can see the changes brought on by climate change.
The National Science Foundation supported this research.

Story Source:
Materials provided by Penn State. Original written by A’ndrea Elyse Messer. Note: Content may be edited for style and length. More

Electron microscopes provide deep insight into the smallest details of matter and can reveal, for example, the atomic configuration of materials, the structure of proteins or the shape of virus particles. However, most materials in nature are not static and rather interact, move and reshape all the time. One of the most common phenomena is the interaction between light and matter, which is ubiquitous in plants as well as in optical components, solar cells, displays or lasers. These interactions — which are defined by electrons being moved around by the field cycles of a light wave — happen at ultrafast time scales of femtoseconds (10-15 seconds) or even attoseconds (10-18 seconds, a billionth of a billionth of a second). While ultrafast electron microscopy can provide some insight into femtosecond processes, it has not been possible, until now, to visualize the reaction dynamics of light and matter occurring at attosecond speeds.
Now, a team of physicists from the University of Konstanz and Ludwig-Maximilians-Universität München have succeeded in combining a transmission electron microscope with a continuous-wave laser to create a prototypical attosecond electron microscope (A-TEM). The results are reported in the latest issue of Science Advances.
Modulating the electron beam
“Basic phenomena in optics, nanophotonics or metamaterials happen at attosecond times, shorter than a cycle of light,” explains Professor Peter Baum, lead author on the study and head of the Light and Matter research group at University of Konstanz’s Department of Physics. “To be able to visualize ultrafast interactions between light and matter requires a time resolution below the oscillation period of light.” Conventional transmission electron microscopes use a continuous electron beam to illuminate a specimen and create an image. To achieve attosecond time resolution, the team led by Baum uses the rapid oscillations of a continuous-wave laser to modulate the electron beam inside the microscope in time.
Ultra-short electron pulses
Key to their experimental approach is a thin membrane which the researchers use to break the symmetry of the optical cycles of the laser wave. This causes the electrons to accelerate and decelerate in rapid succession. “As a result, the electron beam inside the electron microscope is transformed into a series of ultrashort electron pulses, shorter than half an optical cycle of the laser light,” says first author Andrey Ryabov, a postdoctoral researcher on the study. Another laser wave, which is split from the first one, is used to excite an optical phenomenon in a specimen of interest. The ultrashort electron pulses then probe the sample and its reaction to the laser light. By scanning the optical delay between the two laser waves, the researchers are then able to obtain attosecond resolution footage of the electromagnetic dynamics inside the specimen.
Simple modifications, large impact
“The main advantage of our method is that we are able to use the available continuous electron beam inside the electron microscope rather than having to modify the electron source. This means that we have a million times more electrons per second, basically the full brightness of the source, which is key to any practical applications,” continues Ryabov. Another advantage is that the necessary technical modifications are rather simple and do not require electron gun modifications.
As a result, it is now possible to achieve attosecond resolution in a whole range of space-time imaging techniques such as time-resolved holography, waveform electron microscopy or laser-assisted electron spectroscopy, amongst others. In the long term, attosecond electron microscopy may help to uncover the atomistic origins of light-matter interactions in complex materials and biological substances.
Facts:
Ultrafast imaging breakthrough: Physicists from the University of Konstanz and Ludwig-Maximilians-Universität München in Germany achieve attosecond time resolution in a transmission electron microscope by combining it with a continuous-wave laser.
Research team led by Professor Peter Baum (University of Konstanz) modify a transmission electron microscope to create time-resolved images of light-matter interactions at attosecond speeds (10-18 seconds).
Potential boost for a range of imaging techniques and the further exploration of the atomistic origins of light-matter interactions.

Story Source:
Materials provided by University of Konstanz. Note: Content may be edited for style and length. More

Autonomous functions for robots, such as spontaneity, are highly sought after. Many control mechanisms for autonomous robots are inspired by the functions of animals, including humans. Roboticists often design robot behaviors using predefined modules and control methodologies, which makes them task-specific, limiting their flexibility. Researchers offer an alternative machine learning-based method for designing spontaneous behaviors by capitalizing on complex temporal patterns, like neural activities of animal brains. They hope to see their design implemented in robotic platforms to improve their autonomous capabilities.
Robots and their control software can be classified as a dynamical system, a mathematical model that describes the ever-changing internal states of something. There is a class of dynamical system called high-dimensional chaos, which has attracted many researchers as it is a powerful way to model animal brains. However, it is generally hard to gain control over high-dimensional chaos owing to the complexity of the system parameters and its sensitivity to varying initial conditions, a phenomenon popularized by the term “butterfly effect.” Researchers from the Intelligent Systems and Informatics Laboratory and the Next Generation Artificial Intelligence Research Center at the University of Tokyo explore novel ways for exploiting the dynamics of high-dimensional chaos to implement humanlike cognitive functions.
“There is an aspect of high-dimensional chaos called chaotic itinerancy (CI) which can explain brain activity during memory recall and association,” said doctoral student Katsuma Inoue. “In robotics, CI has been a key tool for implementing spontaneous behavioral patterns. In this study, we propose a recipe for implementing CI in a simple and systematic fashion only using complicated time-series patterns generated by high-dimensional chaos. We felt our approach holds potential for more robust and versatile applications when it comes to designing cognitive architectures. It allows us to design spontaneous behaviors without any predefined explicit structures in the controller, which would otherwise serve as a hindrance.”
Reservoir computing (RC) is a machine learning technique that builds on dynamical systems theory and provides the basis of the team’s approach. RC is used to control a type of neural network called a recurrent neural network (RNN). Unlike other machine learning approaches that tune all neural connections within a neural network, RC only tweaks some parameters while keeping all other connections of an RNN fixed, which makes it possible to train the system faster. When the researchers applied principles of RC to a chaotic RNN, it exhibited the kind of spontaneous behavioral patterns they were hoping for. For some time, this has proven a challenging task in the field of robotics and artificial intelligence. Furthermore, the training for the network takes place prior to execution and in a short amount of time.
“Animal brains yield high-dimensional chaos in their activities, but how and why they utilize chaos remains unexplained. Our proposed model could offer insight into how chaos contributes to information processing in our brains,” said Associate Professor Kohei Nakajima. “Also, our recipe would have a broader impact outside the field of neuroscience since it can potentially be applied to other chaotic systems too. For example, next-generation neuromorphic devices inspired by biological neurons potentially exhibit high-dimensional chaos and would be excellent candidates for implementing our recipe. I hope we will see artificial implementations of brain functions before too long.”

Story Source:
Materials provided by University of Tokyo. Note: Content may be edited for style and length. More

Thermoelectric materials can turn a temperature difference into electricity. Organic thermoelectric materials could be used to power wearable electronics or sensors; however, the power output is still very low. An international team led by Jan Anton Koster, Professor of Semiconductor Physics at the University of Groningen, has now produced an n-type organic semiconductor with superior properties that brings these applications a big step closer. Their results were published in the journal Nature Communications on 10 November.
The thermoelectric generator is the only human-made power source outside our solar system: both Voyager space probes, which were launched in 1977 and are now in interstellar space, are powered by generators that convert heat (in this case, provided by a radioactive source) into an electric current. ‘The great thing about such generators is that they are solid-state devices, without any moving parts,’ explains Koster.
Conductivity
However, the inorganic thermoelectric material used in the Voyager’s generators is not suitable for more mundane applications. These inorganic materials contain toxic or very rare elements. Furthermore, they are usually rigid and brittle. ‘That is why interest in organic thermoelectric materials is increasing,’ says Koster. Yet, these materials have their own problems. The optimal thermoelectric material is a phonon glass, which has a very low thermal conductivity (so that it can maintain a temperature difference) and also an electron crystal with high electrical conductivity (to transport the generated current). Koster: ‘The problem with organic semiconductors is that they usually have a low electrical conductivity.’
Nevertheless, over a decade of experience in developing organic photovoltaic materials at the University of Groningen has led the team on a path to a better organic thermoelectric material. They focused their attention on an n-type semiconductor, which carries a negative charge. For a thermoelectric generator, both n-type and p-type (carrying positive charge) semiconductors are needed, although the efficiency of organic p-type semiconductors is already quite good.
Buckyballs
The team used fullerenes (buckyballs, made up of sixty carbon atoms) with a double-triethylene glycol-type side-chain added to them. To increase the electrical conductivity, an n-dopant was added. ‘The fullerenes already have a low thermal conductivity, but adding the side chains makes it even lower, so the material is a very good phonon glass,’ says Koster. ‘Furthermore, these chains also incorporate the dopant and create a very ordered structure during annealing.’ The latter makes the material an electric crystal, with an electrical conductivity similar to that of pure fullerenes.
‘We have now made the first organic phonon glass electric crystal,’ Koster says. ‘But the most exciting part for me is its thermoelectric properties.’ These are expressed by the ZT value. The T refers to the temperature at which the material operates, while Z incorporates the other material properties. The new material increases the highest ZT value in its class from 0.2 to over 0.3, a sizeable improvement.
Sensors
‘A ZT value of 1 is considered a commercially viable efficiency, but we believe that our material could already be used in applications that require a low output,’ says Koster. To power sensors, for example, a few microwatts of power are required and these could be produced by a couple of square centimetres of the new material. ‘Our collaborators in Milan are already creating thermoelectric generators using fullerenes with a single side chain, which have a lower ZT value than we now have.’
The fullerenes, side chain and dopant are all readily available and the production of the new material can likely be scaled up without too many problems, according to Koster. He is extremely happy with the results of this study. ‘The paper has twenty authors from nine different research groups. We used our combined knowledge of synthetic organic chemistry, organic semiconductors, molecular dynamics, thermal conductivity and X-ray structural studies to get this result. And we already have some ideas on how to further increase the efficiency.’

Story Source:
Materials provided by University of Groningen. Note: Content may be edited for style and length. More

The ongoing global pandemic has created an urgent need for rapid tests that can diagnose the presence of the SARS-CoV-2 virus, the pathogen that causes COVID-19, and distinguish it from other respiratory viruses. Now, researchers from Japan have demonstrated a new system for single-virion identification of common respiratory pathogens using a machine learning algorithm trained on changes in current across silicon nanopores. This work may lead to fast and accurate screening tests for diseases like COVID-19 and influenza.
In a study published this month in ACS Sensors scientists at Osaka University have introduced a new system using silicon nanopores sensitive enough to detect even a single virus particle when coupled with a machine learning algorithm.
In this method, a silicon nitride layer just 50 nm thick suspended on a silicon wafer has tiny nanopores added, which are themselves only 300 nm in diameter. When a voltage difference is applied to the solution on either side of the wafer, ions travel through the nanopores in a process called electrophoresis.
The motion of the ions can be monitored by the current they generate, and when a viral particle enters a nanopore, it blocks some of the ions from passing through, leading to a transient dip in current. Each dip reflects the physical properties of the particle, such as volume, surface charge, and shape, so they can be used to identify the kind of virus.
The natural variation in the physical properties of virus particles had previously hindered implementation of this approach, however, using machine learning, the team built a classification algorithm trained with signals from known viruses to determine the identity of new samples. “By combining single-particle nanopore sensing with artificial intelligence, we were able to achieve highly accurate identification of multiple viral species,” explains senior author Makusu Tsutsui.
The computer can discriminate the differences in electrical current waveforms that cannot be identified by human eyes, which enables highly accurate virus classification. In addition to coronavirus, the system was tested with similar pathogens — respiratory syncytial virus, adenovirus, influenza A, and influenza B.
The team believes that coronaviruses are especially well suited for this technique since their spiky outer proteins may even allow different strains to be classified separately. “This work will help with the development of a virus test kit that outperforms conventional viral inspection methods,” says last author Tomoji Kawai.
Compared with other rapid viral tests like polymerase chain reaction or antibody-based screens, the new method is much faster and does not require costly reagents, which may lead to improved diagnostic tests for emerging viral particles that cause infectious diseases such as COVID-19.

Story Source:
Materials provided by Osaka University. Note: Content may be edited for style and length. More