Computers Math

Researchers at the U.S. Department of Energy’s (DOE) Princeton Plasma Physics Laboratory (PPPL) have found that updating a mathematical model to include a physical property known as resistivity could lead to the improved design of doughnut-shaped fusion facilities known as tokamaks.
“Resistivity is the property of any substance that inhibits the flow of electricity,” said PPPL physicist Nathaniel Ferraro, one of the collaborating researchers. “It’s kind of like the viscosity of a fluid, which inhibits things moving through it. For example, a stone will move more slowly through molasses than water, and more slowly through water than through air.”
Scientists have discovered a new way that resistivity can cause instabilities in the plasma edge, where temperatures and pressures rise sharply. By incorporating resistivity into models that predict the behavior of plasma, a soup of electrons and atomic nuclei that makes up 99% of the visible universe, scientists can design systems for future fusion facilities that make the plasma more stable.
“We want to use this knowledge to figure out how to develop a model that allows us to plug in certain plasma characteristics and predict whether the plasma will be stable before we actually do an experiment,” said Andreas Kleiner, a PPPL physicist who was the lead author of a paper reporting the results in Nuclear Fusion. “Basically, in this research, we saw that resistivity matters and our models ought to include it,” Kleiner said.
Fusion, the power that drives the sun and stars, combines light elements in the form of plasma — the hot, charged state of matter composed of free electrons and atomic nuclei — and generates massive amounts of energy. Scientists seek to harness fusion on Earth for a virtually inexhaustible supply of power to generate electricity.
Scientists want the plasma to be stable because instabilities can lead to plasma eruptions known as edge-localized modes (ELMs) that can damage internal components of the tokamak over time, requiring those components to be replaced more frequently. Future fusion reactors will have to run without stopping for repairs, however, for months at a time. More

Loss of permanent teeth is usually caused by dental diseases or trauma and is common in the global population, especially among the elderly due to aging and relatively poorer oral health.
Failure to replace a missing tooth not only affects facial aesthetic and chewing function, but it may also lead to jawbone loss and shifting to teeth, which may cause malocclusion and bite irregularities that could have a significant impact on the health of the remaining teeth, gums, jaw muscles and jaw points.
Artificial teeth, also known as bridges and dentures, are prosthetic devices used to replace missing teeth. It is essential for the false teeth to resemble the patient’s original tooth so that the patient can retain his or her original appearance, chewing function, oral and physical health.
Currently, the process of designing and creating dentures is highly time-consuming as the existing computerised design process requires tedious manual inputs, teeth occlusion information collection as well as multiple denture fitting procedures due to limited accuracy of exciting technologies.
Researchers from the Faculty of Dentistry at the University of Hong Kong (HKU) and the Department of Computer Science of Chu Hai College of Higher Education, collaborated to develop a new approach using artificial intelligence to automate the design of individualised dentures, in order to enhance the treatment efficiency and improve patient experience.
The AI technology used in the process was based on 3D Generative Adversarial Network (3D-GAN) algorithm and tested on 175 participants recruited at HKU. The study shows that AI technology could reconstruct the shape of a natural healthy tooth and automate the process of false teeth design with high accuracy.
“The 3D GAN algorithm was selected due to its superior performance on 3D object reconstruction compared to other AI algorithms. In the preliminary study, 3D GAN was able to rebuild similar shapes to the original teeth for 60% of the cases. It is expected to mature with more AI training data,” co-Investigator, Dr Reinhard Chau explained.
The new approach only requires the digital model of a patient’s dentition to function. It can learn the features of an individual’s teeth from the rest of the dentition and generate a false tooth that looks like the missing tooth.
“This will facilitate the treatment workflow for dentists in replacing a missing tooth, as the preparation and fitting process will require minimal time, and a patient will not need to stay at the clinic for long hours,” said Principal Investigator Dr Walter Lam.
The study entitled “Artificial intelligence-designed single molar dental prostheses: A protocol of prospective experimental study” is published in the journal PLoS ONE. The preliminary results of the study were presented in the recent International Association of Dental Research (IADR) General Session. The study won the IADR Neal Garrett Clinical Research Prize and First runner-up in the 2022 IADR-SEA Hatton Award — Senior Category.
Story Source:
Materials provided by The University of Hong Kong. Note: Content may be edited for style and length. More

New research led by University of Massachusetts Amherst assistant professor Jinglei Ping has overcome a major challenge to isolating and detecting molecules at the same time and at the same location in a microdevice. The work, recently published in ACSNano, demonstrates an important advance in using graphene for electrokinetic biosample processing and analysis and could allow lab-on-a-chip devices to become smaller and achieve results faster.
The process of detecting biomolecules has been complicated and time consuming. “We usually first have to isolate them in a complex medium in a device and then send them to another device or another spot in the same device for detection,” says Ping, who is in the College of Engineering’s Mechanical and Industrial Engineering Department and is also affiliated with the university’s Institute of Applied Life Sciences. “Now we can isolate them and detect them at the same microscale spot in a microfluidic device at the same time — no one has ever demonstrated this before.”
His lab achieved this advance by using graphene, a one-atom-thick honeycomb lattice of carbon atoms, as microelectrodes in a microfluidic device.
“We found that, compared to typical inert-metal microelectrodes, the electrolysis stability for graphene microelectrodes is more than 1,000 times improved, making them ideal for high-performance electrokinetic analysis,” he says.
Also, Ping added, since monolayer graphene is transparent, “we developed a three-dimensional multi-stream microfluidic strategy to microscopically detect the isolated molecules and calibrate the detection at the same time from a direction normal to the graphene microelectrodes.”
The new approach developed in the work paves the way to the creation of lab-on-a-chip devices of maximal time and size efficiencies, Ping says. Also, the approach is not limited to analyzing biomolecules and can potentially be used to separate, detect and stimulate microorganisms such as cells and bacteria.
Story Source:
Materials provided by University of Massachusetts Amherst. Note: Content may be edited for style and length. More

A newborn giraffe or foal must learn to walk on its legs as fast as possible to avoid predators. Animals are born with muscle coordination networks located in their spinal cord. However, learning the precise coordination of leg muscles and tendons takes some time. Initially, baby animals rely heavily on hard-wired spinal cord reflexes. While somewhat more basic, motor control reflexes help the animal to avoid falling and hurting themselves during their first walking attempts. The following, more advanced and precise muscle control must be practiced, until eventually the nervous system is well adapted to the young animal’s leg muscles and tendons. No more uncontrolled stumbling — the young animal can now keep up with the adults.
Researchers at the Max Planck Institute for Intelligent Systems (MPI-IS) in Stuttgart conducted a research study to find out how animals learn to walk and learn from stumbling. They built a four-legged, dog-sized robot, that helped them figure out the details.
“As engineers and roboticists, we sought the answer by building a robot that features reflexes just like an animal and learns from mistakes,” says Felix Ruppert, a former doctoral student in the Dynamic Locomotion research group at MPI-IS. “If an animal stumbles, is that a mistake? Not if it happens once. But if it stumbles frequently, it gives us a measure of how well the robot walks.”
Felix Ruppert is first author of “Learning Plastic Matching of Robot Dynamics in Closed-loop Central Pattern Generators,” which will be published July 18, 2022 in the journal Nature Machine Intelligence.
Learning algorithm optimizes virtual spinal cord
After learning to walk in just one hour, Ruppert’s robot makes good use of its complex leg mechanics. A Bayesian optimization algorithm guides the learning: the measured foot sensor information is matched with target data from the modeled virtual spinal cord running as a program in the robot’s computer. The robot learns to walk by continuously comparing sent and expected sensor information, running reflex loops, and adapting its motor control patterns. More

Thanks to high-tech, low-cost tracking devices, the study of wildlife movement is having its Big Data moment. But so far, only people with data science skills have been able to glean meaningful insights from this ‘golden age’ of tracking. A new system from the Max Planck Institute of Animal Behavior (MPI-AB) and the University of Konstanz is changing that. MoveApps is a platform that lets scientists and wildlife managers explore animal movement data — with little more than a device and a browser — to tackle real-world issues.
The system is linked to the database Movebank, developed by MPI-AB and hosted at the Max Planck Computing and Data Facility (MPCDF), which stores tracking data for over one thousand animal species worldwide. This tracking data can be pulled into MoveApps where owners of these data can then run complex analyses to find meaning in the numbers. A ranger could use the system to keep an eye on tracked animals in the park, creating a daily map showing where animals are located. Or, a conservation agency working with endangered species could receive an alert when a sudden clustering of GPS points suggests that an animal might have died.
In a paper published in Movement Ecology, the authors detail how MoveApps unites programmers with data owners needing analytical tools on an open, serverless platform. While programmers develop tools that become openly available on the platform, users can browse these tools and run analyses with a few simple clicks on a user-friendly web-based interface.
The aim is to turn animal tracking Big Data towards solving big problems — by making it possible to analyze and make sense of movement data quickly and easily.
“You don’t need a data science degree, you don’t need to work at a university, you don’t need a software license or a big computer,” says first author Andrea Kölzsch, MoveApps project lead and postdoc at MPI-AB “You just need to have a question that can be answered with animal tracking data.”
Near real-time analysis for rapid response
When the North Carolina Zoo began tagging and tracking African vultures seven years ago, zoo staff would spend hours each day analyzing the GPS data to figure out if they needed to check on the birds. African vultures are the fastest declining group of birds globally and are particularly susceptible to poisoning by feeding on carcasses laced with pesticides. Because a poisoned carcass can kill over a hundred vultures in a matter of hours, zoo staff need to quickly identify feeding events, indicated by a clustering of GPS points from tagged birds. More

The Tin Man didn’t have one. The Grinch’s was three sizes too small. And for soft robots, the electronically powered pumps that function as their “hearts” are so bulky and rigid, they must be decoupled from the robot’s body — a separation that can leak energy and render the bots less efficient.
Now, a collaboration between Cornell researchers and the U.S. Army Research Laboratory has leveraged hydrodynamic and magnetic forces to drive a rubbery, deformable pump that can provide soft robots with a circulatory system, in effect mimicking the biology of animals.
“These distributed soft pumps operate much more like human hearts and the arteries from which the blood is delivered,” said Rob Shepherd, associate professor of mechanical and aerospace engineering in the College of Engineering, who led the Cornell team. “We’ve had robot blood that we published from our group, and now we have robot hearts. The combination of the two will make more lifelike machines.”
The group’s paper, “Magnetohydrodynamic Levitation for High-Performance Flexible Pumps,” published July 11 in Proceedings of the National Academy of Sciences. The paper’s lead author was postdoctoral researcher Yoav Matia.
Shepherd’s Organic Robotics Lab has previously used soft material composites to design everything from stretchable sensor “skin” to combustion-driven braille displays and clothing that monitors athletic performance — plus a menagerie of soft robots that can walk and crawl and swim and sweat. Many of the lab’s creations could have practical applications in the fields of patient care and rehabilitation.
Like animals, soft robots need a circulatory system to store energy and power their appendages and movements to complete complex tasks.
The new elastomeric pump consists of a soft silicone tube fitted with coils of wire — known as solenoids — that are spaced around its exterior. Gaps between the coils allow the tube to bend and stretch. Inside the tube is a solid core magnet surrounded by magnetorheological fluid — a fluid that stiffens when exposed to a magnetic field, which keeps the core centered and creates a crucial seal. Depending on how the magnetic field is applied, the core magnet can be moved back and forth, much like a floating piston, to push fluids — such as water and low-viscosity oils — forward with continuous force and without jamming.
“We’re operating at pressures and flow rates that are 100 times what has been done in other soft pumps,” said Shepherd, who served as the paper’s co-senior author with Nathan Lazarus of the U.S. Army Research Laboratory. “Compared to hard pumps, we’re still about 10 times lower in performance. So that means we can’t push really viscous oils at very high flow rates.”
The researchers conducted an experiment to demonstrate that the pump system can maintain a continuous performance under large deformations, and they tracked the performance parameters so future iterations can be custom-tailored for different types of robots.
“We thought it was important to have scaling relationships for all the different parameters of the pump, so that when we design something new, with different tube diameters and different lengths, we would know how we should tune the pump for the performance we want,” Shepherd said.
Postdoctoral researcher Hyeon Seok An contributed to the paper.
The research was supported by the U.S. Army Research Laboratory.
Story Source:
Materials provided by Cornell University. Original written by David Nutt, courtesy of the Cornell Chronicle. Note: Content may be edited for style and length. More

Changes in atmospheric density after sunset can cause hot pockets of gas called ‘plasma bubbles’ to form over the Earth’s equator, resulting in communication disruptions between satellites and the Earth. New AI models are now helping scientists to predict plasma bubble events and create a forecast. The work was presented this week at the National Astronomy Meeting (NAM 2022) by Sachin Reddy, a PhD student at University College London.
Shortly after sunset, pockets of super-heated gas called ‘plasma bubbles’ form in the upper atmosphere and stretch into space (up to 900km above the Earth’s surface). These bubbles start small and grow rapidly — from the size of a football pitch to that of a small country in just a couple of hours. As the bubbles grow bigger, they can prevent satellites from communicating with the Earth by blocking and warping their radio signals.
To predict plasma bubbles, a team of researchers has collated 8 years of data from the SWARM satellite mission. The spacecraft has an automatic bubble detector on-board called the Ionospheric Bubble Index. This compares changes in the density of electrons and the magnetic field strength to check if bubbles are present: a strong correlation between the two indicates the presence of a plasma bubble.
The satellite flies at an altitude of 460km (about 30x higher than a commercial plane) through the middle of most plasma bubbles. The model combines the data collection from SWARM with a machine learning approach to make predictions on the likelihood of a plasma bubble event occurring at any time.
The results show that the number of plasma bubble events varies from season to season, just like the weather, and that the number of events increases with solar activity. Despite this, the model finds location to be a far more crucial element in predicting plasma bubbles than the time of year, with most events occurring over a region in the Atlantic called the South Atlantic Anomaly. The AI model predicts events with an accuracy of 91% across different tests.
Reddy says: “Just like the weather forecast on earth, we need to be able to forecast bubbles to prevent major disruptions to satellite services. Our aim is to be able to say something like: “At 8pm tomorrow there is a 30% chance of a bubble appearing over the Horn of Africa.” This kind of information is extremely useful for spacecraft operators and for people who depend on satellite data every day, just like you and me.”
Story Source:
Materials provided by Royal Astronomical Society. Note: Content may be edited for style and length. More

Researchers at Nanyang Technological University, Singapore (NTU Singapore) have found that the increased use of videoconferencing platforms during the COVID-19 pandemic contributed to a higher level of fatigue, as reported by workers.
Following work-from-home orders issued by governments worldwide during the pandemic, many employees attended meetings virtually using technologies such as Zoom or Microsoft Teams, instead of meeting face-to-face.
In a survey conducted in December 2020, the NTU research team found that 46.2% of all respondents reported feelings of fatigue or being overwhelmed, tired, or drained from the use of videoconferencing applications.
The researchers derived the results through an analysis of a survey of 1,145 Singapore residents in full-time employment and who had indicated that they use videoconferencing apps frequently.
The researchers from the NTU Wee Kim Wee School of Communication and Information (WKWSCI) and its Centre for Information Integrity and the Internet (IN-cube), published their findings in the journal Computers in Human Behavior Reports in June 2022.
Assistant Professor Benjamin Li, from NTU’s WKWSCI, who led the study, said: “We were motivated to conduct our study after hearing of increasing reports of fatigue from the use of videoconferencing applications during the pandemic. We found that there was a clear relation between the increased use of videoconferencing and fatigue in Singaporean workers. Our findings are even more relevant in today’s context, as the use of videoconferencing tools is here to stay, due to flexible work arrangements being a continuing trend.” He is also a member of IN-cube. More