Materials science is moving beyond the "perimeter of ignorance" faster than we can keep up with it. And definitely faster than architects, civil engineers, industrial designers, etc. can keep up with it. Materials scientists, and even computer scientists in overlapping fields, are finding lots of "completely unexpected" things that defy our understanding of how matter behaves.
Combine that with the "weird computer" revolution that happens when the matter itself becomes programmable, and the future gets hard to imagine. (Maybe less hard to imagine is the resulting human health and ecological disasters that will happen, kind of like how the industrial revolution created climate change).
A world where every molecule is itself a computer - The farthest I can get when thinking about this is Stanislav Lem's
Solaris (1961) where the planet itself was not only alive but conscious, and trying to communicate with humans.
Shape-shifting worm blob model could inspire future robot swarms
Oct 2021, phys.org
'Entangled active matter collectives' are a hot topic in robotics and materials science...
via Georgia Tech: Chantal Nguyen et al, Emergent Collective Locomotion in an Active Polymer Model of Entangled Worm Blobs, Frontiers in Physics (2021). DOI: 10.3389/fphy.2021.734499
Physicists make square droplets and liquid lattices
Sep 2021, phys.org
Completely unexpected:
In their work, the team used combinations of oils with different dielectric constants and conductivities, then subjected the liquids to an electric field.
As well as being disrupted by the electric field, the liquids were confined into a thin, nearly two-dimensional sheet. This combination led to the oils reshaping into various completely unexpected droplets and patterns.
The droplets in the experiment could be made into squares and hexagons with straight sides, which is almost impossible in nature, where small bubbles and droplets tend to form spheres. The two liquids could be also made to form into interconnected lattices: grid patterns that occur regularly in solid materials but are unheard of in liquid mixtures.
via Aalto University Department of Applied Physics in the Active Matter: Diversity of non-equilibrium patterns and emergence of activity in confined electrohydrodynamically driven liquids, Science Advances (2021). DOI: 10.1126/sciadv.abh1642
The next generation of robots will be shape-shifters
Mar 2022, phys.org
It is hoped that active matter will lead to a new generation of machines whose function will come from the bottom up. So, instead of being governed by a central controller (the way today's robotic arms are controlled in factories), these new machines would be made from many individual active units that cooperate to determine the machine's movement and function. This is akin to the workings of our own biological tissues, such as the fibers in heart muscle.
via University of Bath: Jack Binysh et al, Active elastocapillarity in soft solids with negative surface tension, Science Advances (2022). DOI: 10.1126/sciadv.abk3079
Self-sensing artificial muscle based on liquid crystal elastomer and low-melting point alloys
May 2022, phys.org
Inspired by the coupled behavior of muscles, bones, and nerve systems of mammals and other living organisms to create a multifunctional artificial muscle in the lab.
via Frontier Institute of Science and Technology, Jiaotong University, China: Haoran Liu et al, Shape-programmable, deformation-locking, and self-sensing artificial muscle based on liquid crystal elastomer and low–melting point alloy, Science Advances (2022). DOI: 10.1126/sciadv.abn5722
Ancient art of kirigami meets AI for better materials design
Apr 2022. phys.org
via Argonne National Laboratory: Pankaj Rajak et al, Autonomous reinforcement learning agent for stretchable kirigami design of 2D materials, npj Computational Materials (2021). DOI: 10.1038/s41524-021-00572-y
And: Pankaj Rajak et al, Autonomous reinforcement learning agent for chemical vapor deposition synthesis of quantum materials, npj Computational Materials (2021). DOI: 10.1038/s41524-021-00535-3
A new approach to tackle optimization problems using Boltzmann machines
Apr 2022, phys.org
"Optimization problem" is codeword for 1. slime mold computers, 2. quantum computers, and 3. weird computers in general, like crystals, dust, liquid photons, BECs, you name it, and because the optimization problem, also known as the traveling salesman problem, and which is related to random walks, or the drunkard's walk, is a type of problem that classical computers are really bad at, but quantum computers, slime mold, etc are really good at.
Restricted Boltzmann machines (RBMs) are generative neural networks. They speak the language of big data and show you the patterns in it.
RBMs rely on binary activations, circumventing the direct matrix-vector multiplications that are typically the most computationally demanding for deep learning networks.
"Our algorithm functions by using the basic principles of digital logic in a new way," Patel explained. "Usually, digital gates only function in the forward direction, but by using probabilistic graphical models and machine learning, we have shown ways of operating them in reverse.
Using this principle, we design our probabilistic digital circuits in a way that can solve the forward problem ("Is this set of inputs a valid solution?" or "What is 191 x 223?"), but because the system is reversible, it can also answer the much harder reverse problem ("What are all the sets of inputs that produce a valid solution?" and "What are A and B such that A x B = 42593?" )."
via University of California Berkeley: Saavan Patel et al, Logically synthesized and hardware-accelerated restricted Boltzmann machines for combinatorial optimization and integer factorization, Nature Electronics (2022). DOI: 10.1038/s41928-022-00714-0
A new age of 2.5D materials
May 2022, phys.org
Scientists are exploring new ways to artificially stack two-dimensional (2D) materials, introducing so-called 2.5D materials with unique physical properties.
They're made using chemical vapor deposition, and they're made out of graphene, hexagonal boron nitride, and transition metal dichalcogenides.
via Kyushu University: Hiroki Ago et al, Science of 2.5 dimensional materials: paradigm shift of materials science toward future social innovation, Science and Technology of Advanced Materials (2022). DOI: 10.1080/14686996.2022.2062576
Mathematicians suggest liquid crystals could be used to create building blocks for a new kind of computer
Aug 2022, phys.org
The orientations of LCD molecules could be manipulated using an electric field and perform calculations similar to the way they are done with standard logic gates. The researchers note that, in their approach, calculations would appear as ripples moving through the crystal.
via MIT: Žiga Kos et al, Nematic bits and universal logic gates, Science Advances (2022). DOI: 10.1126/sciadv.abp8371
New programmable materials can sense their own movements
Aug 2022, phys.org
"Sensorizing structures"
Method for 3D printing materials with tunable mechanical properties from incorporated networks of air-filled channels, and which can sense how they are moving and interacting with the environment.
Also "architected materials" have customizable mechanical properties based solely on its geometry.
via MIT: Fluidic innervation sensorizes structures from a single build material, Science Advances (2022). science.org/doi/10.1126/sciadv.abq4385
Researchers engineer novel material capable of 'thinking'
Aug 2022, phys.org
"We discovered how to use mathematics and kinematics in mechanical-electrical networks."
The researchers were stuck, until they rediscovered a 1938 paper published by Claude E. Shannon, who described a way to create an integrated circuit by constructing mechanical-electrical switching networks that follow the laws of Boolean mathematics.
The material is made from conductive and non-conductive rubber materials that sense and react to how forces are applied to them.
via Pennsylvania State University: Ryan Harne, Mechanical integrated circuit materials, Nature (2022). DOI: 10.1038/s41586-022-05004-5.
