Genetic algorithm used to design broadband metamaterial | KurzweilAI

Genetic algorithm used to design broadband metamaterial | KurzweilAI: "...this is the first that can cover a super-octave [more than doubling] bandwidth in the infrared spectrum...”

The new metamaterial is actually made of layers on a silicon substrate or base. The first layer is palladium, followed by a polyimide (plastic) layer and a palladium screen layer on top. The screen has elaborate, complicated cutouts — sub-wavelength geometry — that serve to block the various wavelengths. A polyimide layer caps the whole absorber...

This evolved metamaterial can be easily manufactured because it is simply layers of metal or plastic that do not need complex alignment. The clear cap of polyimide serves to protect the screen, but also helps reduce any impedance mismatch that might occur when the wave moves from the air into the device...

The most complex synthetic biology circuit yet

The most complex synthetic biology circuit yet: Using genes as interchangeable parts, synthetic biologists design cellular circuits that can perform new functions, such as sensing environmental conditions. However, the complexity that can be achieved in such circuits has been limited by a critical bottleneck: the difficulty in assembling genetic components that don’t interfere with each other...


The pathway consists of three components: an activator, a promoter and a chaperone. A promoter is a region of DNA where proteins bind to initiate transcription of a gene. An activator is one such protein. Some activators also require a chaperone protein before they can bind to DNA to initiate transcription.

The researchers found 60 different versions of this pathway in other species of bacteria, and found that most of the proteins involved in each were different enough that they did not interfere with one another. However, there was a small amount of crosstalk between a few of the circuit components, so the researchers used an approach called directed evolution to reduce it. Directed evolution is a trial-and-error process that involves mutating a gene to create thousands of similar variants, then testing them for the desired trait. The best candidates are mutated and screened again, until the optimal gene is created.

Evolution could generate new semiconducting structures

Evolution could generate new semiconducting structures: The team chose silicateins – proteins that build the silica skeletons of marine sponges – as the basis for their work. Using DNA amplification techniques, they grew millions of strands of DNA that code for silicateins. Mutations arise naturally during the process, so the final pool of DNA contained enough variation to ensure that some of the silicateins would build different kinds of mineral structures.

The researchers then attached the DNA to polystyrene microbeads and placed them in a solution containing a silicon-rich compound. Bawazer's team was looking to select proteins that could draw silicon out of the solution to build silica structures around the beads, while still allowing access to the DNA on the surface of the bead. This would make it easy to collect and amplify the DNA that made the most promising structures. The end product? Proteins that built silica structures unlike any seen in nature.

Computer Program 'Evolves' Music From Noise - ScienceNOW

Computer Program 'Evolves' Music From Noise - ScienceNOW: ...adapted DarwinTunes to be accessed online by almost 7000 participants who rated each sound loop, played in a random order, on a 5-point scale from "can't stand it" to "love it." In a musical take on survival of the fittest, the highest-scored loops went on to pair up with others and replicate. Each resulting generation was rated again for its appeal. After about 2500 generations of sound loops, what started out as a cacophony of noise had evolved into pleasant strains of music.

AI designs its own video game

AI designs its own video game: Angelina creates games using a technique known as cooperative co-evolution. The system separately designs different aspects, or species, of the game. In Space Station Invaders - in which players control a scientist who must fend off rogue robots and invading aliens to escape a space station - the species include the layout of each different level, enemy behaviour and the power-ups that give a player extra abilities. Angelina creates a level by randomly selecting from a list, then scattering enemies and power-ups throughout the level. Enemy movements and combat behaviours are also randomly selected from a list, while the effects of the power-ups are also random.

It then combines the species and simulates a human playing the game to see which designs lead to the most fun or interesting results. For example, levels that are initially hard to complete but get easier through clever use of power-ups are considered fun, while those that are impossible to complete are discarded. Angelina then cross-breeds and mutates the most successful members of each species to evolve a new generation, typically 400 times.

Robot biologist solves complex problem from scratch | KurzweilAI

Robot biologist solves complex problem from scratch | KurzweilAI: The biological system that the researchers used to test ABE is glycolysis, the primary process that produces energy in a living cell. They focused on how yeast cells control glycolytic oscillations  because it is one of the most extensively studied biological control systems. ABE derived the equations a priori. The only thing the software knew in advance was addition, subtraction, multiplication and division...

Lipson used genetic programming for the breeding process... However, this process also proved to be too slow.
So Lipson combined the breeding and the debugging processes in an approach he calls co-evolution.

Researchers develop algorithm to predict new superhard crystals

Researchers develop algorithm to predict new super hard crystals: The optimization of hardness is a successful proof-of-principle example, which opens the way for a novel computational technique. “A new era in material design and discovery is about to begin,” said Prof. Oganov. “New materials with desired properties will be routinely discovered using supercomputers, instead of the expensive trial-and-error method that is used today.”

Evolving software inspired by natural selection | Santa Fe Institute

Evolving software inspired by natural selection | Santa Fe Institute: In effect, evolutionary computation starts with a glitchy program, creates a group of slight variations on the original, and keeps the best variations as part of the next generation of the program. Then, repeat until the software does what it’s supposed to do...
Evolutionary computation needs a few additional tricks to make it scalable, but unlike other approaches it works for a wide range of programs, and it’s surprisingly fast. The researchers looked at 16 programs and about 120,000 lines of code with a range of problems from infinite loops to buffer overflows. They found they could repair a program in under six minutes, on average. Humans would take considerably longer -- they would “at least have to read the code” first, Forrest says.

One Per Cent: Evolve your own objects for 3D printing

One Per Cent: Evolve your own objects for 3D printing: Their long-term goal is to create robots that can evolve like biological creatures, so EndlessForms is designed to explore what kind of biological body shapes the model can produce. Forget designing your own objects - what about 3D-printed pets?

Evolution machine: Genetic engineering on fast forward

Evolution machine: Genetic engineering on fast forward - life - 27 June 2011 - New ScientistThe machine let the E. coli multiply, mixed them with the DNA strands, and applied an electric shock to open up the bacterial cells and let the DNA get inside. There, some of the added DNA was swapped with the matching target sequences in the cells' genomes. This process, called homologous recombination, is usually very rare, which is where the viral enzymes come in. They trick cells into treating the added DNA as its own, greatly increasing the chance of homologous recombination.

Under pressure, sodium and hydrogen could undergo a metamorphosis, emerging as a superconductor

Under pressure, sodium and hydrogen could undergo a metamorphosis, emerging as a superconductor: Using an open-source computer program that UB PhD student David Lonie designed, Zurek and Baettig looked for sodium polyhydrides that, under pressure, would be viable superconductor candidates. The program, XtalOpt, is an evolutionary algorithm that incorporates quantum mechanical calculations to determine the most stable geometries or crystal structures of solids.
In analyzing the results, Baettig and Zurek found that NaH9, which contains one sodium atom for every nine hydrogen atoms, is predicted to become metallic at an experimentally achievable pressure of about 250 gigapascals -- about 2.5 million times the Earth's standard atmospheric pressure, but less than the pressure at the Earth's core (about 3.5 million atmospheres).

Move over, Einstein: Machines will take it from here - physics-math - 22 March 2011 - New Scientist

Move over, Einstein: Machines will take it from here: To get an idea of how it works, imagine you find a number of equations that seem to relate the height of the pendulum's midpoint and its horizontal position. While a trivial answer would simply relate these two variables, a physical law of motion describes a much deeper connection between them. This means that it can be used to predict, for example, how the height and position change with time. So Schmidt and Lipson decided to use this to test their equations: not only must an equation match the data, in the pendulum's case it must also describe how the pendulum changes in time...
Today, the algorithm is called Eureqa and has thousands of users all over the world, with people using it for everything from financial forecasting to particle physics.

Evolutionary robotics: for robust robots, let them be babies first | KurzweilAI

Evolutionary robotics: for robust robots, let them be babies first | KurzweilAI: Using a sophisticated computer simulation, Bongard unleashed a series of synthetic beasts that move about in a 3-dimensional space. “It looks like a modern video game,” he says. Each creature — or, rather, generations of the creatures — then run a software routine, called a genetic algorithm, that experiments with various motions until it develops a slither, shuffle, or walking gait — based on its body plan — that can get it to the light source without tipping over.
“The robots have 12 moving parts,” Bongard says. “They look like the simplified skeleton of a mammal: it’s got a jointed spine and then you have four sticks — the legs — sticking out.”

Metamorphosis key to creating stable walking robots - tech - 10 January 2011 - New Scientist

Metamorphosis key to creating stable walking robots: "He found that the four-legged robot was stable when programmed to walk like any of the virtual bots that had metamorphosed with time. 'Metamorphosed robots were able to continue walking even if they were randomly pushed around,' he says. However, when the four-legged robot adopted the walking style of a virtual bot with a fixed body plan, it was far more prone to falling over when pushed. Bongard thinks that's because the morphed robots had to remain balanced and on course through many body plans, so the gait they finally adopted had greater stability."

Scientists develop breakthrough method for crystal structure prediction

Scientists develop breakthrough method for crystal structure prediction: "USPEX provides great opportunities to predict the structure of compounds simply starting from a chemical formula and letting the ‘evolutionary code’ work."...
...Oganov's team made the first major step in solving this problem in 2006 with their development of a powerful evolutionary algorithm that finds the stable structure using ideas inspired by biological evolution. This method has been called "revolutionary" by some scientists...
...the computer generates dozens of initial structures, but only the most preferred ones are allowed to mate and mutate before starting the process, until the best candidates are finally obtained...

Jumping beats moonwalking – for a virtual robot - space - 18 December 2010 - New Scientist

Jumping beats moonwalking – for a virtual robot: "Christopher MacLeod at Robert Gordon University in Aberdeen, UK, is intrigued by the idea of simulating how robots could move on the lunar surface. He uses genetic algorithms to evolve efficient gaits for robots on Earth.

He thinks that if such software is used to evolve the most efficient gait for lunar movement, the results might be quite different from those obtained using Earth's gravity. To move at speed on Earth, going down on all fours and travelling like a chimp provides more power and greater stability than running on two legs – but on the moon that may not apply, he says."

Computer-generated robots

Computer-generated robots: The robots consist of cylinder-shaped tubes with ball-and-socket joints that can assume different shapes depending on external factors and the purpose at hand. “The only input needed is: ‘Move forwards as efficiently as possible along a level surface,’” assures Dipl.-Des. Andreas Fischer, industrial designer and product developer at IPA. Fitness functions within the software algorithm select the movement elements with which the Genetic Robot can advance along this surface the software determines the shape of the tubes, the position of the movement points and the position of the drives (actuators).

The basis for the development is a physic engine in which the most important environmental influences – such as the friction of the ground or gravity – are implemented. If the Genetic Robot is to withstand unevenness, climb stairs or swim in water, these environmental conditions can be simulated.

Video: Swiss Researchers Test Largest-Ever Swarm of Aerial Robots | Popular Science

Video: Swiss Researchers Test Largest-Ever Swarm of Aerial Robots | Popular Science (skip to 1:17 in video): Students built 1-pound plastic foam microdrones with 31-inch wingspans and outfitted them with electric motors, a Linux-based processor, GPS and a WiFi dongle. Then they had to design swarm algorithms.

First, the researchers used artificial evolution models to uncover unique control mechanisms, and reverse-engineered their findings. Then they turned to one of evolution’s best-studied social systems: Ants.

As the project’s Web site explains, the creatures deploy to search for and maintain pheromone paths that lead them to food. This behavior can be an analogue for deploying communication networks that would help rescuers, the researchers say.

CultureLab: Sci Foo: Evolution of music and a dancing cockatoo

CultureLab: Sci Foo: Evolution of music and a dancing cockatoo: He plays us a medley of randomly-generated sequences - the starting point for the experiment - followed by another after 1000 or so (if I remember right) rounds of selection. The first clip is an ugly, clashing affront to our ears but the second, while it has no tune, is rhythmic and strangely mesmerising.

Brain implants evolved to use less energy - New Scientist - New Scientist

Brain implants evolved to use less energy - New Scientist - New Scientist: "After 10,000 generations, the optimum waveform emerged: a bell curve with the lower parts of each side cut short by a vertical line. Tests with implants in cats showed the waveforms were beneficial and minimised energy use. Grill says an implant using this waveform would need its batteries replaced five or six times over 30 years instead of eight to 10 times"