It seems not a day goes by in the tech world without a headline appearing about new nanoscale devices that promise to revolutionize medicine or extend Moore’s Law. But when it comes to the development of energy sources to power these tiny devices, the advances have been fewer and farther between. Current power sources such as conventional batteries are too large and their contents too toxic to use in nanoscale systems, especially those designed for use in the body.

Researchers at the Georgia Institute of Technology, however, have developed a technique to convert mechanical, hydraulic, or vibrational energy to electrical energy, creating the potential for future self-powered nanoscale energy sources known as nanogenerators. These devices would harvest the mechanical energy that is readily available in the environment to produce a steady supply of current, capable of providing power to wireless sensors, portable electronics, and implantable biomedical devices.

Professor Zhong Lin Wang and graduate student Jinhui Song created the prototype nanogenerators by growing arrays of zinc oxide nanowires, which are both piezoelectric (meaning they generate an electric charge when mechanically deformed) and semiconducting. The duo then used an atomic-force microscope tip to bend the wires individually. As the nanowires stretched and released, they generated a small amount of measurable electric current.

With further development the nanogenerators could be placed anywhere that there’s enough motion to produce vibrations in the nanowires, such as in shoes, flowing water or wind, a moving body, and even in an acoustic or ultrasonic energy field. Technology is already available in the form of tiny capacitors to store the energy, so constant motion isn’t necessary to sustain electrical flow.

Now that the team has demonstrated that the system can generate electrical charge, the next step is to up the voltage. Wang says, “The biggest technological challenge we will face is the simultaneous generation of electricity from many nanowires and output them in series so that we can have a higher voltage output.”

Wang expects we could start seeing nanogenerators powering nanosensors in three to five years.



Thought-Controlled Robots—Not Just Wishful Thinking

Imagine turning on the lights in your house just by thinking about flipping the switch. Or imagine having a prosthetic arm that performs tasks as effortlessly as a natural one. Once merely the stuff of science fiction, thought-controlled computers are a step closer to reality with a system that Japanese scientists at the ATR Computational Neuroscience Laboratories in Kyoto and Honda Research Institute in Saitama have developed.

Using brain activity that a fMRI (functional magnetic resonance imaging) scanner recorded in real-time, a team that Yukiyasu Kamitani is leading has created a robotic hand that can mimic the motions of a human hand using data from a subject making “rock, paper, scissor” movements. Other computer-brain interface projects have achieved similar results using electrodes implanted directly into the brain or attached to a subject’s skull, but the fMRI is less invasive and produces a much higher-resolution snapshot of a subject’s brain activity.

The system relies on the fact that different thoughts and actions create unique fluctuations in the blood flow to different brain regions. The fMRI scanner captures these while the subject is forming the hand movements and feeds them to an attached computer. Once the computer is trained to recognize which patterns are associated with a shape, the robotic hand can duplicate those shapes when the subject moves and a pattern match is made. There was a seven second lag between the subject and robot’s movements, but researchers achieved an 85% decoding accuracy.

Research could eventually lead developing prosthetic robotic devices that a wearer’s thoughts completely control or to computers recognizing and executing nonverbal, nontactile commands. Much work is left, however, to execute commands based on thought alone.

“A next step will be to decode the intention of an action before the action is initiated,” says Kamitani. “Entertainment products like video games might implement the basic idea in as early as five years,” he says.



A New Take On Network Security

Computer security experts trying to design systems to protect networks from malicious computer viruses and hacker attacks have long used as a model the method that the human immune system handles attacks by viruses and other harmful biological agents. But just as vaccines only work against disease-causing agents that the body has been already made aware of, traditional computer security measures fall short against structurally new types of attacks, leaving the electronic patient also vulnerable.

This is because the model that has dominated both immunology and computer security is one that uses signature matching to detect intruders. A United Kingdom team of researchers that Uwe Aickelin at the University of Nottingham is leading, however, is looking to an alternative theory as a basis for an artificial immune system for computer networks that could be more effective at combating new viruses and DoS attacks.

Called the “danger theory,” it proposes that a simple “non-self” reaction doesn’t trigger an immune system response when a threat enters the body, but instead the immune system is reacting to danger signals that an attack (cell distress, damage, and death) causes. Aickelin has transferred this concept to the security software he’s developing, which monitors a network and sounds an alarm only when it detects unusual activity, such as sudden network traffic spikes or an increase in the number of error messages.

Aieckelin says, “Our system relies on ‘signals’ which are polled regularly. Hence, any type of new attack can be spotted, as long as it stresses the computer.”

Aickelin says the biggest challenge his research faces is getting others to understand it because the work is so different than what has been done previously. “The immunologists themselves are still having problems convincing traditional colleagues that the new theories about danger are useful and able to explain previously inexplicable aspects of immunology.

“We face the same challenge in computer science, as people are quite familiar with traditional pattern-matching computer-security approaches and because they are anchored in their beliefs find it difficult to grasp what we are doing,” Aickelin says.

Aickelin is working closely with HP labs in Bristol, UK, to commercialize a system based on his research. He estimates it could be available in a five-year timeframe.



MIT Takes New Material For A Spin

While not quite poetic enough to be the stuff of dreams, MIT’s newest breakthrough could be the stuff that future spin-based computer chips are made of. A research team that Jagadeesh Moodera leads at MIT’s Francis Bitter Magnet Lab developed the new material, a magnetic semiconductor that is made of indium oxide with a dash of chromium. Operating at room temperature and sitting on top of a conventional silicon semiconductor, the transparent material injects electrons of a certain spin state into the circuit. The electrons travel through, and a spin detector reads them on the other side.

In the field of spin-based electronics, or spintronics, not only does the on/off state used in conventional electronic circuits carry information, but the spin direction (or up and down) of electronics does as well. This allows for more data to be stored in less space, and also, due to the nonvolatile nature of spin states, reduces the power consumption.

Spintronic devices are currently being used for high-density mass data storage, but many scientists see spintronics as a possible savior of Moore’s Law. The material that the MIT researchers developed introduces the possibility of using spin states to carry information through an electronic circuit, applying the benefits of spintronics to computer transistors. This could result in smaller chips that not only process more data and consume less energy but that can also take advantage of reversible spin states to create multifunctional electronic devices using the same circuitry.

“We can carry information in two ways at once, and this will allow us to further reduce the size of electronic circuits,” Moodera stated in an MIT press release.

by Kristina Spencer