System Bits: Oct. 20

Automating big-data analysis
Until now, big-data analysis consisted of searching for buried patterns that had some kind of predictive power but picking which “features” of the data to analyze usually required some human intuition.

Now, however, MIT researchers are aiming to take the human element out of big-data analysis with a new system that they say not only searches for patterns but designs the feature set, too. To test the first prototype of their system, they enrolled it in three data science competitions, in which it competed against human teams to find predictive patterns in unfamiliar data sets. Of the 906 teams participating in the three competitions, the researchers said their “Data Science Machine” finished ahead of 615.

Further, in two of the three competitions, the predictions made by the Data Science Machine were 94 percent and 96 percent as accurate as the winning submissions. In the third, the figure was a more modest 87 percent. But where the teams of humans typically labored over their prediction algorithms for months, the Data Science Machine took somewhere between two and 12 hours to produce each of its entries.

To Max Kanter, whose MIT master’s thesis in computer science is the basis of the Data Science Machine, the Data Science Machine is a natural complement to human intelligence. “There’s so much data out there to be analyzed. And right now it’s just sitting there not doing anything. So maybe we can come up with a solution that will at least get us started on it, at least get us moving.”

Graphene-based inks for printed electronics
A low-cost, high-speed method for printing graphene inks using a conventional roll-to-roll printing process, like that used to print newspapers and crisp packets, could open up a wide range of practical applications, including inexpensive printed electronics, intelligent packaging and disposable sensors, according to University of Cambridge researchers.

The team collaborated with Cambridge-based technology company Novalia, to develop the method that allows graphene and other electrically conducting materials to be added to conventional water-based inks and printed using typical commercial equipment, believed to be the first time that graphene has been used for printing on a large-scale commercial printing press at high speed.

Roll-to-roll printing of graphene ink
(Source: University of Cambridge)

The researchers said graphene’s flexibility, optical transparency and electrical conductivity make it suitable for a wide range of applications, including printed electronics. And while numerous laboratory prototypes have been demonstrated around the world, widespread commercial use of graphene is yet to be realized.

Potential applications include printed, disposable biosensors, energy harvesters and RFID tags.

Natural electromagnets
In the drive to miniaturize electronics, solenoids have become way too big, according to Rice University researchers who discovered the essential component can be scaled down to nano-size with macro-scale performance, and discovered graphene spirals could challenge macro solenoids.

The secret, according to Rice theoretical physicist Boris Yakobson and his colleagues, is in a spiral form of atom-thin graphene that, remarkably, can be found in nature.

Yakobson explained, “Usually, we determine the characteristics for materials we think might be possible to make, but this time we’re looking at a configuration that already exists. These spirals, or screw dislocations, form naturally in graphite during its growth, even in common coal.”

The team determined that when a voltage is applied, current will flow around the helical path and produce a magnetic field, as it does in macro inductor-solenoids.

Solenoids are wires coiled around a metallic core. They produce a magnetic field when carrying current, turning them into electromagnets. These are widespread in electronic and mechanical devices, from circuit boards to transformers to cars. They also serve as inductors, primary components in electric circuits that regulate current, and in their smallest form are part of integrated circuits. (The lump in power cables that feed electronic devices contains inductors.)

While transistors get steadily smaller, basic inductors in electronics have become relatively bulky, and it’s the same inside the circuits. Commercial spiral inductors on silicon occupy excessive area. If realized, graphene nano-solenoids could change that.

A nano-coil made of graphene could be an effective solenoid inductor for electronic applications, according to researchers at Rice University. (Source: Yakobson Research Group/Rice University)

The nano-solenoids analyzed through computer models at Rice should be capable of producing powerful magnetic fields of about 1 tesla, about the same as the coils found in typical loudspeakers, according to Yakobson and his team.