April 14, 2006

Machine optimization, brute force innovation and artificial intelligence

John Koza has combined genetic algorithms to search solutions and optimize around a problem space with artificial intelligence to create computer generated patentable solutions He is examining existing patents for electronic circuits and using those as parameters to be fed into genetic algorithms.

Applying these genetic algorithm principles to improving the design of project teams

Combining dynamic computational chemistry with genetic algorithms to accelerate the discovery and advance of scientific research

These approaches can work together to rapidly explore the area of a defined problem space. As these techniques are perfected innovation will move to asking unique and interesting questions. Automation will accelerate the finding of solutions.

April 13, 2006

Other tech: artificial gill closer

Artificial Gills: One Big Stroke Closer To Reality The key to artificial gills, explains Harihara Baskaran, an engineer at Case Western Reserve University, is getting water to flow easily through microchannels, which mimic a real gill. Working with Infoscitex of Waltham, Mass., Baskaran is using chipmaking techniques to create tiny devices with hundreds of channels. Bump that up to 250,000 channels, and the device could generate enough oxygen for a person. More on Harihara Baskaran's work

Other artificial gill work by an Israeli company More on the isreali work Details and interview with the Israeli inventor Alan Bodner on his artificial gill Better batteries and waterflow are the keys to better artificial gills.

other tech: new robotic military vehicle, Crusher

"Crusher," a 6.5-ton, six-wheeled robotic vehicle designed to negotiate harsh terrain, will be presented along with its predecessor, "Spinner," at Carnegie Mellon's National Robotics Engineering Center on April 28, spokeswoman Anne Watzman said. Although Crusher is designed to carry weapons, the university has worked only on the machinery of the vehicle, Watzman said.

Crusher combines some capabilities of Spinner -- an invertible machine able to right itself -- with mobility and autonomy technology, such as the use of terrain data, developed under a program called PerceptOR.

Carnegie Mellon had two successful entries in the DARPA Grand Challenge. They were led by William Whittaker Here is a list of his current and completed projects

The current projects are

Robotic Search for Antarctic Meteorites: autonomous mobile robot and sensors to locate and recover meteorites in Antarctica

Pioneer: a mobile mapping and reconnaissance machine for structural assessment of the damaged Chernobyl nuclear power plant

RoboHost: robotic tour guides for museums

Demeter: mobile robot for unmanned grain harvesting

AutoLoad: technologies to improve productivity and reduce costs of excavation in earthmoving projects

Lunar Rover Initiative: a pair of mobile robots for the first privately funded lunar mission with telepresence for public participation and education

Carnegie Mellon Robotics Institute are leaders in robotics work Here is a directory of projects at the robotics institute with pictures

Larger nanoscale: Improved electrospinning

Techniques improved could be useful on smaller scales.

Daoheng Sun and Liwei Lin's at the university of Berkeley have improved electrospinning. They enable fibers ranging from 50 to 500 nanometers in diameter to be deposited onto a collector plate in a directed, controlled manner. In reference to the shortened distance between the ejector and collection points that it used, the team named the new process "near-field electrospinning."

First, instead of applying the polymer solution into the electric field with a syringe, they used a fine-tipped tungsten electrode, which they dipped into the solution like a pen into ink. Then, positioning the electrode above a collection plate, they applied electrical voltage to it, creating the electric field and initiating the process of electrospinning with the tiny drop of polymer on the electrode's tip. This allowed the team to reduce the initial diameter of the polymer stream as it leaves the electrode far below the diameter of the stream produced by the conventional syringe.

Second, the researchers shortened the distance the polymer travels in the electric field from the conventional 10 to 30 centimeters to between one-half millimeter and three millimeters. This allowed them to take advantage of the brief period of stability that polymer fibers exhibit when the electrospinning process begins. Just like the exhaust of a jet engine that shoots out in a straight line before billowing into random patterns, the fibers move in a relatively straight line for a fleeting moment when they enter the electric field. In Sun and Lin's near-field technique, the fibers are captured before their billowing begins.

The shortened distance also meant that Lin and Sun could dramatically reduce the voltage required from 30,000 volts to as low as 600 volts. Finally, rather than using a screen fixed in place to capture the fibers, Sun and Lin let the fibers land on a plate that could be moved in various patterns at various speeds. This allowed the researchers to pattern the fibers onto the plate the way a quilter creates a design by maneuvering fabric under her sewing machine's needle.

Lin said he foresees the possibility of two immediate directions for the new process. One is for device applications that require precise deposition of the nanofibers, such as making nanosensors for biological measurements – a glucose monitor, for instance. The other will be to make non-woven fabrics with organized patterns that can have many applications, such as scaffolds for living cells. Near-field electrospinning may also be useful in nanolithography for making next-generation microchips, Lin predicted. But, he said, this will require more effort to develop.

Lin is currently working on two improvements to the near-field process: an electrode that can provide a continuous supply of polymer and a movable stage with good planar control to capture the fibers.

Light powered engine added the 4 nanometer long Rice University nanocar

This work is continued refinement of increasingly complex and useful nanodevices.

Light powered engine added the 4 nanometer long Rice University nanocar light powered nanocar The car’s light-powered motor is attached mid-chassis. When struck by light, it rotates in one direction, pushing the car along like a paddlewheel. There is other coverage at cnet
The nanocar consists of a rigid chassis and four alkyne axles that spin freely and swivel independently of one another. The four buckyball wheels that were used in the original version of the nanocar drained energy from the motor and were replaced with spherical molecules of carbon, hydrogen and boron called p-carborane. Rice University has more pictures and diagrams

April 12, 2006

Oak Ridge develops method for optically trapping and moving objects smaller than wavelength of light

A team from the Department of Energy's Oak Ridge National Laboratory, California Institute of Technology and Protein Discovery developed a new method to use a beam of light to trap protein molecules and make them dance in space The technique, called photoelectrophoretic localization and transport, or PELT, involves shining a highly focused beam of light on semiconductor material and using electric fields to move the proteins. Force-field traps are created by a photocurrent focused at the illuminated areas of the semiconductor. In contrast to traditional electrophoresis, which uses high voltage, this approach permits researchers to dynamically change characteristics of the electric field in three dimensions in real time using computer-controlled software and low voltage.

This new method also overcomes limitations of conventional optical trapping techniques, commonly called optical tweezers, which are versatile but unable to transport objects smaller than the wavelength of light. Included in this category are many biomolecules such as DNA fragments, oligonucleotides, proteins and peptides. Instead, such small molecules must first be attached to larger particles called "handles." This and other techniques have significant limitations, according to authors of the PNAS paper.

April 11, 2006

Other tech: State of the art superconductor electronics 2006

A large number of articles on superconductor electronics as of 2005-6 In general higher speed (up to 1000GHz) and more sensitive for detection and communication.

April 10, 2006

Nanoscale tech: Nanopore DNA sequencing status

DNA sequencing by passing a DNA strand through a “nanopore” to detect electrical changes as the strand of DNA is passed through it could speed up DNA sequencing more than 200 times. US physicists at the University of California, San Diego have made a detailed computer simulation of over 100,000 interacting atoms and shown theoretically how to get the necessary resolution on the sensing. The challenge lies in refining the method to improve its resolution to be able to detect single bases. Using the current methods, a nucleotide needs to be repeated in a sequence about 100 times successively in order to produce a measurable characteristic change. The concept of nanopore sequencing has been around since at least 2000.

Previous attempts to sequence DNA using nanopores were not successful because the twisting and turning of the DNA strand introduced too much noise into the signal being recorded. The new idea takes advantage of the electric field that drives the current perpendicular to the DNA strand to reduce the structural fluctuations of DNA while it moves through the pore, thus minimizing the noise. The University of California, San Diego researchers caution that there are still hurdles to overcome because no one has yet made a nanopore with the required configuration of electrodes, but they think it is only a matter of time before someone successfully assembles the device. The nanopore and the electrodes have been made separately, and although it is technically challenging to bring them together, the field is advancing so rapidly that they think it should be possible in the near future.

The work is part of a project to sequence a mammal-sized genome for $1,000.