SKYNET comes one step closer

Look at the “dynamic re-grasping of the cell phone” at 2:39 in the video. [Link]

Concept

A human being recognizes external environment by using many kinds of sensory information. By integrating these information and making up lack of information for each other, a more reliable and multilateral recognition can be achieved. The purpose of Sensor Fusion Project is to realize new sensing architecture by integrating multi-sensor information and to develop hierarchical and decentralized architecture for recognizing human beings further. As a result, more reliable and multilateral information can be extracted, which can realize high level recognition mechanism.

Robot Fish from MIT

That sounds it emits is eerie. [Link]

The robofish bodies are continuous (i.e., not divided into different segments), flexible and made from soft polymers. This makes them more maneuverable and better able to mimic the swimming motion of real fish, which propel themselves by contracting muscles on either side of their bodies, generating a wave that travels from head to tail.

"Most swimming techniques can be copied by exploiting natural vibrations of soft structures," says Valdivia Y Alvarado.

As part of his doctoral thesis, Valdivia Y Alvarado created a model to calculate the optimal material properties distributions along the robot's body to create a fish with the desired speed and swimming motion. The model, which the researchers initially proposed in the journal Dynamic Systems Measurements and Control (ASME), also takes into account the robot's mass and volume. A more detailed model is described in Valdivia Y Alvarado's thesis and will soon be published along with new applications by the group.

Other researchers, including a team at the University of Essex, have developed new generations of robotic fish using traditional assembly of rigid components to replicate the motions of fish, but the MIT team is the only one using controlled vibrations of flexible bodies to mimic biological locomotion.

"With these polymers, you can specify stiffness in different sections, rather than building a robot with discrete sections," says Youcef-Toumi. "This philosophy can be used for more than just fish" - for example, in robotic prosthetic limbs.

The most notorious counterfeiter

Interesting story about a guy who produced $7 Million worth of fake US Dollars using common scanners and printers between 2005 and 2008. Just got caught late last year. [Link]

When Talton set out to circumvent the U.S. Treasury's security measures, he had no experience in counterfeiting, printing, or graphic design, and he didn't even own a computer. His first attempts were made with a Hewlett-Packard all-in-one ink-jet printer/scanner/fax/copier, which could be picked up at the time for less than $150. Early experiments, printed on regular copy paper, were fuzzy, so he cleaned up the original image on a computer. But there was a problem, Talton says: "It wouldn't take the mark." Counterfeit-detection pens mark yellow on genuine currency but brown or black on fake. Talton didn't know why. At first he thought the Treasury treated the paper, so he experimented with chemicals he found at the body shop and even tried dipping his notes in fabric softener. Nothing worked. Frustrated, he began taking a detection pen everywhere he went, trying it on whatever paper he came across. He was about to give up when one day, sitting on the toilet, he found himself staring at the roll of tissue beside him. He took out the pen: The mark showed up yellow.

Talton discovered that toilet paper, the pages of Bibles and dictionaries, and newsprint are all made from the same kind of recycled paper pulp, and all take the mark. Newsprint is strong, and it has an additional advantage for the large-scale buyer: "Newsprint is real cheap," Talton says.

The simplest method of making counterfeit money is to scan both sides of a bill and print them on either side of a single piece of paper. But in a genuine bill the security strip and watermark are embedded, so this type of counterfeit is never convincing. Talton realized he could solve the problem by using two sheets of tissue-thin newsprint: He printed imitation watermarks and security strips on the back of one, then glued the sheets together with the security features inside. Next he printed the front and back faces of the bills—five at a time—on either side of the two-ply sheets, which he hung from clotheslines and coated with hair spray, creating a texture similar to that of genuine currency and a barrier that helped the paper take the mark of a counterfeit pen. Finally, he cut the notes to size. For all his scrupulousness, though, Talton used the same scan for every $100 bill he printed, so the alphanumeric codes to the left and right of the portrait of Ben Franklin never changed. These are the quadrant number and the face-plate number, which indicate which plate at the Bureau of Engraving and Printing was used to make the bill: Talton's hundred came from plate No. 38, spot H, quadrant No. 2, and was thus marked h2 and h38. H2-H38 would quickly become the name by which the Secret Service identified a new family of counterfeit notes.

A new kind of router

Lawrence Roberts wrote one of the first RFC’s for the predecessor of internet, ARPANET. In an article in IEEE describes why he he thinks the current generation of routers is not suited to the internet in its current form. He has started a new company to design a new kind of router that identifies and operates on packet sequences (called flows) instead of the packets themselves. [Link]

Since the early days of the ARPANET, I’ve always thought that routers should manage flows rather than individual packets. Why hasn’t it been done before? The reason is that memory chips were too expensive until not long ago. You need lots of memory to store the hash table with routing parameters of each flow. (A 1 gigabit-per-second data trunk often carries about 100 000 flows.) If you were to keep a flow table on one IMP of 40 years ago, you’d spend US $1 million in memory. But about a decade ago, as memory cost kept falling, it started to make sense economically to design flow-management equipment.

In early 2004, I decided to leave Caspian and start Anagran, focusing on smaller flow-management equipment to solve the overload and fairness problems. We designed the equipment to operate at the edge of networks, the point where an Internet service provider aggregates traffic from its broadband subscribers or where a corporate network connects to the outside world. Virtually all network overload occurs at the edge.

Anagran’s flow manager, the FR-1000, can replace routers and DPI systems or may simply be added to existing networks. It supports up to 4 million simultaneous flows—a combined 80 Gb/s in throughput. Its hardware consists of inexpensive, off-the-shelf components as opposed to ASICs, which increase development costs. We implemented our flow-routing algorithms in a field-programmable gate array, or FPGA, and the router’s memory consists of standard high-speed DRAM. The FR-1000 sells in different models, starting at less than $30 000.

Like a regular router, the FR-1000 has input and output ports. But the similarities end there. Recall that in a traditional router the routing and queuing chips consume 80 percent of the power and space. By routing only the first packet of a flow, the FR-1000’s chips do much less work, consuming about 1 percent of the power that a conventional router requires.

Even more significant, the FR-1000 does away entirely with the queuing chips. During congestion, it adjusts each flow rate at its input instead. If an incoming flow has a rate deemed too high, the equipment discards a single packet to signal the transmission to slow down. And rather than just delaying or dropping packets as in regular routers, in the FR-1000 the output provides feedback to the input. If there’s bandwidth available, the equipment increases the flow rates or accepts more flows at the input; if bandwidth is scarce, the router reduces flow rates or discards packets.

By eliminating power-hungry circuitry, the FR-1000 consumes about 300 watts, or one-fifth the total power of a comparable router, and occupies one unit in a standard rack, a tenth of the space that other routers fill. We estimate that the equipment allows network operators to reduce their operating costs per gigabit per second by a factor of 10.

Translation Party

This website takes a phrase and translates it repeatedly between English and Japanese until the input and output English phrases match exactly. [Link]

First, the message that one sees if the Java script is turned off :)

and now, the translation party:

until …

Imagining the 10th dimension

Well, I would say it is entertaining even though it sounds like pseudoscience. Plus you get to hear what a whole multi-verse collapsing into a point would sound like.

Caution: multiple infinities are discussed in this video.

Invention Overload

Slide show of all the inventions entered in the James Dyson Award. [Link]

Parking Patent?

This is really clever :)

Seth Godin: Marketing or Engineering

What’s more important?

GigaPan: Munich Airport

Gigapan.org sells a robotic tripod that automatically takes and stitches pictures using any point and shoot camera to take panoramic pictures that contain breath-taking detail. Below, a 150 degree panoramic view of the Munich Airport. More Panoramas at GigaPan.

Self-Healing Surfaces

Scientists have invented surfaces that can “heal” themselves of scratches (like human skin). [Link]

Mechanical bearings are one example of possible applications – the materials of the bearings usually have a electroplated coating, in which the capsules can be embedded. If there is a temporary shortage of lubricant, part of the bearing’s coating is lost, the capsules at the top of the layer burst and release lubricant. The bearing is not therefore damaged if it temporarily runs dry. The researchers have produced the first copper, nickel and zinc coatings with the new capsules, although surface coverage does not extend beyond the centimeter scale. Experts estimate that it will be another one and a half to two years before whole components can be coated. In a further step the team worked on more complex systems – involving differently filled capsules, for example, whose fluids react with one another like a two component adhesive.