Aurelia Butler

MIT researchers have developed a method for 3D printing materials with tunable mechanical properties, which can sense how they are moving and interacting with the environment. The researchers create these sensing structures using just one material and a single run on a 3D printer.
To accomplish this, the researchers began with 3D-printed lattice materials and incorporated networks of air-filled channels into the structure during the printing process. By measuring how the pressure changes within these channels when the structure is squeezed, bent, or stretched, engineers can receive feedback on how the material is moving.
These lattice materials are composed of single cells in a repeating pattern. Changing the size or shape of the cells alters the material’s mechanical properties, such as stiffness or hardness. For instance, a denser network of cells makes a stiffer structure.
This technique could someday be used to create flexible soft robots with embedded sensors that enable the robots understand their posture and movements. It might also be used to produce wearable smart devices, like customized running shoes that provide feedback on how an athlete’s foot is impacting the ground.
“The idea with this work is that we can take any material that can be 3D-printed and have a simple way to route channels throughout it so we can get sensorization with structure. And if you use really complex materials, then you can have motion, perception, and structure all in one,” says co-lead author Lillian Chin, a graduate student in the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL).
Joining Chin on the paper are co-lead author Ryan Truby, a former CSAIL postdoc who is now as assistant professor at Northwestern University; Annan Zhang, a CSAIL graduate student; and senior author Daniela Rus, the Andrew and Erna Viterbi Professor of Electrical Engineering and Computer Science and director of CSAIL. The paper is published in Science Advances. More

In firefighting, the worst flames are the ones you don’t see coming. Amid the chaos of a burning building, it is difficult to notice the signs of impending flashover — a deadly fire phenomenon wherein nearly all combustible items in a room ignite suddenly. Flashover is one of the leading causes of firefighter deaths, but new research suggests that artificial intelligence (AI) could provide first responders with a much-needed heads-up.
Researchers at the National Institute of Standards and Technology (NIST), the Hong Kong Polytechnic University and other institutions have developed a Flashover Prediction Neural Network (FlashNet) model to forecast the lethal events precious seconds before they erupt. In a new study published in Engineering Applications of Artificial Intelligence, FlashNet boasted an accuracy of up to 92.1% across more than a dozen common residential floorplans in the U.S. and came out on top when going head-to-head with other AI-based flashover predicting programs.
Flashovers tend to suddenly flare up at approximately 600 degrees Celsius (1,100 degrees Fahrenheit) and can then cause temperatures to shoot up further. To anticipate these events, existing research tools either rely on constant streams of temperature data from burning buildings or use machine learning to fill in the missing data in the likely event that heat detectors succumb to high temperatures.
Until now, most machine learning-based prediction tools, including one the authors previously developed, have been trained to operate in a single, familiar environment. In reality, firefighters are not afforded such luxury. As they charge into hostile territory, they may know little to nothing about the floorplan, the location of fire or whether doors are open or closed.
“Our previous model only had to consider four or five rooms in one layout, but when the layout switches and you have 13 or 14 rooms, it can be a nightmare for the model,” said NIST mechanical engineer Wai Cheong Tam, co-first author of the new study. “For real-world application, we believe the key is to move to a generalized model that works for many different buildings.”
To cope with the variability of real fires, the researchers beefed up their approach with graph neural networks (GNN), a kind of machine learning algorithm good at making judgments based on graphs of nodes and lines, representing different data points and their relationships with one another.
“GNNs are frequently used for estimated time of arrival, or ETA, in traffic where you can be analyzing 10 to 50 different roads. It’s very complicated to properly make use of that kind of information simultaneously, so that’s where we got the idea to use GNNs,” said Eugene Yujun Fu, a research assistant professor at the Hong Kong Polytechnic University and study co-first author. “Except for our application, we’re looking at rooms instead of roads and are predicting flashover events instead of ETA in traffic.”
The researchers digitally simulated more than 41,000 fires in 17 kinds of buildings, representing a majority of the U.S. residential building stock. In addition to layout, factors such as the origin of the fire, types of furniture and whether doors and windows were open or closed varied throughout. They provided the GNN model with a set of nearly 25,000 fire cases to use as study material and then 16,000 for fine tuning and final testing.
Across the 17 kinds of homes, the new model’s accuracy depended on the amount of data it had to chew on and the lead time it sought to provide firefighters. However, the model’s accuracy — at best, 92.1% with 30 seconds of lead time — outperformed five other machine-learning-based tools, including the authors’ previous model. Critically, the tool produced the least false negatives, dangerous cases where the models fail to predict an imminent flashover.
The authors threw FlashNet into scenarios where it had no prior information about the specifics of a building and the fire burning inside it, similar to the situation firefighters often find themselves in. Given those constraints, the tool’s performance was quite promising, Tam said. However, the authors still have a ways to go before they can take FlashNet across the finish line. As a next step, they plan to battle-test the model with real-world, rather than simulated, data.
“In order to fully test our model’s performance, we actually need to build and burn our own structures and include some real sensors in them,” Tam said. “At the end of the day, that’s a must if we want to deploy this model in real fire scenarios.” More

Our modern understanding of the physical world is based on gauge theories: mathematical models from theoretical physics that describe the interactions between elementary particles (such as electrons or quarks) and explain quantum mechanically three of the fundamental forces of nature: the electromagnetic, weak, and strong forces. The fourth fundamental force, gravity, is described by Einstein’s theory of general relativity, which, while not yet understood in the quantum regime, is also a gauge theory. Gauge theories can also be used to explain the exotic quantum behavior of electrons in certain materials or the error correction codes that future quantum computers will need to work reliably, and are the workhorse of modern physics.
In order to better understand these theories, one possibility is to realize them using artificial and highly controllable quantum systems. This strategy is called quantum simulation and constitutes a special type of quantum computing. It was first proposed by the physicist Richard Feynman in the 80s, more than fifteen years after being awarded the Nobel prize in physics for his pioneering theoretical work on gauge theories. Quantum simulation can be seen as a quantum LEGO game where experimental physicists give reality to abstract theoretical models. They build them in the laboratory “quantum brick by quantum brick,” using very well controlled quantum systems such as ultracold atoms or ions. After assembling one quantum LEGO prototype for a specific model, the researchers can measure its properties very precisely in the lab, and use their results to understand better the theory that it mimics. During the last decade, quantum simulation has been intensively exploited to investigate quantum materials. However, playing the quantum LEGO game with gauge theories is fundamentally more challenging. Until now, only the electromagnetic force could be investigated in this way.
In a recent study published in Nature, ICFO experimental researchers Anika Frölian, Craig Chisholm, Ramón Ramos, Elettra Neri, and Cesar Cabrera, led by ICREA Prof. at ICFO Leticia Tarruell, in collaboration with Alessio Celi, a theoretical researcher from the Talent program at the Autonomous University of Barcelona, were able to simulate a gauge theory other than electromagnetism for the first time, using ultracold atoms.
A gauge theory for very heavy photons
The team set out to realize in the laboratory a gauge theory belonging to the class of topological gauge theories, different from the class of dynamical gauge theories to which electromagnetism belongs.
In the gauge theory language, the electromagnetic force between two electrons arises when they exchange a photon: a particle of light that can propagate even when matter is absent. However, in two-dimensional quantum materials subjected to very strong magnetic fields, the photons exchanged by the electrons behave as if they were extremely heavy and can only move as long as they are attached to matter. As a result, the electrons have very peculiar properties: they can only flow through the edges of the material, in a direction that is set by the orientation of the magnetic field, and their charge becomes apparently fractional. This behavior is known as the fractional quantum Hall effect, and is described by the Chern-Simons gauge theory (named after the mathematicians that developed one of its key elements). The behavior of the electrons restricted to a single edge of the material should also be described by a gauge theory, in this case called chiral BF, which was proposed in the 90s but not realized in a laboratory until the ICFO and UAB researchers pulled it out of the freezer. More

Researchers at NYU Tandon have reached a key milestone in their quest to develop wearable technology that manages to measure key brain mechanisms through the skin.
Rose Faghih, Associate Professor of Biomedical Engineering, has been working for the last seven years on a technology that can measure mental activity using electrodermal activity (EDA) — an electrical phenomenon of the skin that is influenced by brain activity related to emotional status. Internal stresses, whether caused by pain, exhaustion, or a particularly packed schedule, can cause changes in the EDA — changes that are directly correlated to mental states.
The overarching goal — a Multimodal Intelligent Noninvasive brain state Decoder for Wearable AdapTive Closed-loop arcHitectures, or MINDWATCH, as Faghih calls it — would act as a way to monitor a wearer’s mental state, and offer nudges that would help them revert back to a more neutral state of mind. For example, if a person was experiencing a particularly severe bout of work-related stress, the MINDWATCH could pick up on this and automatically play some relaxing music.
Now Faghih — along with Rafiul Amin, her former PhD student — has accomplished a crucial task required for monitoring this information. For the first time, they have developed a novel inference engine that can monitor brain activity through the skin in real time with high scalability and accuracy. The results are featured in a new paper, “Physiological Characterization of Electrodermal Activity Enables Scalable Near Real-Time Autonomic Nervous System Activation Inference,” published in PLOS Computational Biology.
“Inferring autonomic nervous system activation from wearable devices in real-time opens new opportunities for monitoring and improving mental health and cognitive engagement,” according to Faghih.
Previous methods measuring sympathetic nervous system activation through the skin took minutes, which is not practical for wearable devices. While her earlier work focused on inferring brain activity through sweat activation and other factors, the new study additionally models the sweat glands themselves. The model includes a 3D state-space representation of the direct secretion of sweat via pore opening, as well as diffusion followed by corresponding evaporation and reabsorption. This detailed model of the glands provides exceptional insight into inferring the brain activity. More

An international collaboration of scientists has created and observed an entirely new class of vortices — the whirling masses of fluid or air.
Led by researchers from Amherst College in the US and the University of East Anglia and Lancaster University in the UK, their new paper details the first laboratory studies of these ‘exotic’ whirlpools in an ultracold gas of atoms at temperatures as low as tens of billionths of a degree above absolute zero.
The discovery, announced this week in the journal Nature Communications, may have exciting future implications for implementations of quantum information and computing.
Vortices are familiar objects in nature, from the whirlpools of water down a bathtub drain to the airflow around a hurricane.
In quantum-mechanical systems, such as an atomic Bose-Einstein condensate, the vortices tend to be tiny and their circulation comes in discrete, quantized units. Such vortices have long been objects of fascination for physicists and have helped to illuminate the unusual properties of superfluidity and superconductivity.
The unusual nature of the observed whirlpools here, however, is due to symmetries in the quantum gas. One especially fascinating property of physical theories, from cosmology to elementary particles, is the appearance of asymmetric worlds despite perfect underlying symmetries. For example, when water freezes to ice, disordered molecules in a liquid arrange themselves into a periodic array. More

Wearable human-machine interface devices, HMIs, can be used to control machines, computers, music players, and other systems. A challenge for conventional HMIs is the presence of sweat on human skin.
In Applied Physics Reviews, by AIP Publishing, scientists at UCLA describe their development of a type of HMI that is stretchable, inexpensive, and waterproof. The device is based on a soft magnetoelastic sensor array that converts mechanical pressure from the press of a finger into an electrical signal.
The device involves two main components. The first component is a layer that translates mechanical movement to a magnetic response. It consists of a set of micromagnets in a porous silicone matrix that can convert the gentle fingertip pressure into a magnetic field variation.
The second component is a magnetic induction layer consisting of patterned liquid metal coils. These coils respond to the magnetic field changes and generate electricity through the phenomenon of electromagnetic induction.
“Owing to the material’s flexibility and durability, the magnetoelastic sensor array can generate stable power under deformations, such as rolling, folding, and stretching,” said author Jun Chen, from UCLA. “Due to these compelling features, the device can be adopted for human-body powered HMI by transforming human biomechanical activities into electrical signals.”
The power required to run the HMI comes from the wearer’s movements. This means no batteries or other external power components are required, rendering the HMI more environmentally friendly and sustainable.
The device was tested in a variety of real-world situations, including in the presence of a water spray, such as might exist in the shower, a rainstorm, or during vigorous athletic activity. The device worked well when wet, since the magnetic field was not greatly affected by the presence of water.
The investigators studied a range of fabrication and assembly techniques to optimize the biomechanical-to-electrical energy conversion of the device. They found they could achieve a balance between performance and flexibility by controlling the thickness of the flexible film and the concentration of the magnetic particles.
To test their system, the investigators carried out a series of experiments in which a subject applied finger taps to turn a lamp off and on and control a music player.
“Our magnetoelastic sensor array not only wirelessly functions as the on and off buttons of a lamp but also controls a music player’s command features, representing the actions of play, pause, next, and previous,” Chen said.
These tests promise new applications for versatile water-resistant HMIs that can be used to control many types of smart devices.
Story Source:
Materials provided by American Institute of Physics. Note: Content may be edited for style and length. More

Cedars-Sinai investigators have created the most bio-realistic and complex computer models of individual brain cells — in unparalleled quantity. Their research, published today in the peer-reviewed journal Cell Reports, details how these models could one day answer questions about neurological disorders — and even human intellect — that aren’t possible to explore through biological experiments.
“These models capture the shape, timing and speed of the electrical signals that neurons fire in order to communicate with each other, which is considered the basis of brain function,” said Costas Anastassiou, PhD, a research scientist in the Department of Neurosurgery at Cedars-Sinai, and senior author of the study. “This lets us replicate brain activity at the single-cell level.”
The models are the first to combine data sets from different types of laboratory experiments to present a complete picture of the electrical, genetic and biological activity of single neurons. The models can be used to test theories that would require dozens of experiments to examine in the lab, Anastassiou said.
“Imagine that you wanted to investigate how 50 different genes affect a cell’s biological processes,” Anastassiou said. “You would need to create a separate experiment to ‘knock out’ each gene and see what happens. With our computational models, we will be able to change the recipes of these gene markers for as many genes as we like and predict what will happen.”
Another advantage of the models is that they allow researchers to completely control experimental conditions. This opens the possibility of establishing that one parameter, such as a protein expressed by a neuron, causes a change in the cell or a disease condition, such as epileptic seizures, Anastassiou said. In the lab, investigators can often show an association, but it is difficult to prove a cause.
“In laboratory experiments, the researcher doesn’t control everything,” Anastassiou said. “Biology controls a lot. But in a computational simulation, all the parameters are under the creator’s control. In a model, I can change one parameter and see how it affects another, something that is very hard to do in a biological experiment.”
To create their models, Anastassiou and his team from the Anastassiou Lab — members of the Departments of Neurology and Neurosurgery, the Board of Governors Regenerative Medicine Institute and the Center for Neural Science and Medicine at Cedars-Sinai, used two different sets of data on the mouse primary visual cortex, the area of the brain that processes information coming from the eyes. More