Artists who work with 3D modeling and drawing programs have long been hindered by the fact that the computer still requires them to input their work in two dimensions--using a mouse and drawing pad. Now, Daniel Keefe, along with colleagues Robert C. Zelenik and David H. Laidlaw, has developed a new system that promises to change that, letting artists “sculpt” complex 3D illustrations in midair.

“Drawing on Air” is a haptic-aided 3D drawing system that gives artists, designers, and illustrators the ability to use their hands to create detailed, precise images and then transfer them to a computer’s 3D modeling and design programs. Although the target audience for the program includes artists and illustrators, there are also interesting appli-cations in the scientific world, helping researchers design virtual reality displays of difficult-to-understand scientific data.

To use the program, a user wears virtual reality goggles that sync with the computer screen in order to create the illusion that the image she is drawing is floating in midair. In the user’s dominant hand, a stylus gives force feedback; the other hand wears a tracking device. This allows two-handed drawing that provides a great deal of control and precision in drawing lines and curves. It also permits one-handed drawing with the stylus only.

The haptic feedback that the artist receives through the stylus provides additional control over line thickness and color. Applying more pressure produces thicker lines. And once the artist draws the object, she can grab it and rotate it to examine and edit it from different angles. Keefe has no current plans to market his system commercially but will continue investigating other applications of technology.

“The hardware used is quite expensive, but developing alternative means of realizing the tool and/or waiting a few years for the hardware to become more mainstream might be first steps toward making the technology more widely available,” Keefe says.



MIT Develops Tiny Tractor Beam

How do you move microscopic objects around on the surface of a silicon chip? If you’re MIT researcher Matthew Lang or David Appleyard, you simply pick them up and move them with a beam of light. The new technology they’ve developed could soon be used as a tool in biological and materials research and may one day even be applied in the microelectronics industry as a way of building sensors electrical components.

Although it might not sound like anything new or remarkable (after all, optical tweezers that use light to move tiny objects have been around for decades), the system that Lang and Appleyard have developed is different in that it works on silicon, an opaque substance. Previously, a laser produced the light beams, requiring a transparent glass surface to let the laser shine through. The new method produces infrared light beams, which pass directly through silicon, instead of visible light.

The infrared light holds the object and can lift it slightly, pushing and pulling it around the surface wherever a user wants. In tests, the system was able to move beads and cells that ranged in size from a few nanometers to a whopping 20 micro-meters across. This size range encompasses nearly all types of living cells.

Because the tool lets someone precisely control the placement of materials on a chip, it could soon be a valuable tool for biological research in which researchers must place cells directly next to sensors in order to monitor them.


Online Game Helps Search Engine Learn

Despite the tremendous amount of music now available online, it can still be hard to find new tunes that you really like. Wouldn’t it be great if there were a search engine that returned a list of songs that matched a description of the type of music you wanted, say, “high-energy acoustic with tight harmonies and fun lyrics”?

A group of researchers at the University of California - San Diego are working toward creating just that kind of search engine, and they’re doing it in a very interesting way: by using a multiplayer online game to gather the data needed to train their “computer audition system” to automatically annotate new music.

The Listen Game presents players with musical categories such as “instrumentation,” “emotional content,” “musical genre,” and “song characteristics.” It then plays a series of seven song clips. For each clip, players receive six words and are asked to select the one that best describes the music and the one that least describes it, in relation to the given category. Players compete against others, who are signed in concurrently, and the game awards points when players agree on the words; it awards more points as the number of players who agree on the words increases. There is also a freestyle round where a user enters his own descriptive word, which is then presented in the next round as one of the choices; if others pick that word, the user who entered it receives points.

Around 500 people have played the prototype game since its inception, spending anywhere from a few minutes to more than an hour matching their responses against other players.

Lankriet says that the team is getting ready to release a new, more graphically appealing version to release to the broader public. “Listen Game was a prototype game, mainly presented to our research community (a limited public) to investigate the possibility of collecting information from humans playing games around music. The answer to this turned out to be ‘Yes, we can do so.’”

Although the current Listen Game is no longer maintained, the new version is nearly finished and will let users either play anonymously or log in to customize the game and interact with other players.



Old Memory, New Tricks

In the quest for memory of the future, sometimes it doesn’t hurt to remember the old chestnut “everything old is new again.” That’s the path that scientists at Arizona State University’s CANi (Center for Applied Nanoionics) took when they decided to modify materials already used in manufacturing present-day silicon chips to create a new type of memory that promises more storage, faster performance, and significantly less energy consumption than current technologies.

PMC (programmable-metallization cell), or nanoionic memory, developed at ASU, is created by “doping” silicon oxide with a small amount of copper, both of which are materials that are used in current-day memory chips. By mixing the copper with the silicon oxide, the copper becomes mobile. When current is applied, the copper atoms rearrange, forming a nanoscale switch.

Because nanoionic memory uses differences in resistance states rather than differences in charge to store information, the ASU research team can stack memory cells on top of one another, and each cell can hold multiple bits of information. This leads to the potential for significantly higher storage density; a 1TB memory card or thumb drive is well within the realm of possibility.

A few companies, including Micron and Qimonda, have already licensed the technology, and it’s possible that the first nanoionic memory chips could be produced within the next two years.

“We are beyond the ‘science project’ stage of the development of this technology,” says Michael Kozicki, CANi’s director. “Real products are currently in development, but whether they survive the tortures of the real world or not has still to be determined. We are highly optimistic, though. This technology has attributes, including a ‘self-healing’ capability, that make survival and eventual widespread deployment high-ly likely.”