Private-sector and government representatives said they are working together more often and more closely in order to secure digital information. The Homeland Security 2002 conference, entitled "Establishing a Culture of Cooperation," emphasized security in the financial sector, which in general is responsible for more information than other industries. Richard Marshall of the Critical Infrastructure Assurance Office said the problem would not be solved by government mandate, but that it required proactive cooperation from both sides. He said his group's collection of cybersecurity "best practices," called the National Strategy to Secure Cyberspace, would probably be approved by President Bush early next year. Meanwhile, Fidelity Investments chief privacy officer Leigh Williams said his company was now spending about the same amount on security-related technology as before Sept. 11, but that it was taking a more unified approach in regards to deployment. Whereas before Sept. 11, Fidelity viewed physical and computer security as different issues, the company is now trying to consolidate its efforts. Williams said that unification extended beyond company boundaries as well, and that many different financial associations had joined together in the Financial Services Sector Coordinating Council to present a united front in dealing with government and formulating industry-wide solutions. Getting financial institutions to secure their data, which are often measured in terabytes, is a formidable challenge, according to Convera's Dale Hazel.
http://www.pcworld.com/news/article/0,aid,107889,00.asp

IBM researcher Scott Fahlman recently authored a proposal to charge spammers and telemarketers for sending unsolicited emails or making intrusive phone calls through the deployment of new software and telephones. Familiar correspondents such as friends and family would be able to circumvent the system by using assigned "interrupt tokens," and Fahlman hopes that phone companies will contribute to the adoption of the technology. The IronPort Internet messaging company certifies whether email is authorized or unauthorized in return for a monetary bond that is debited when recipients report unsolicited messages, but this money goes to antispam organizations; the fees charged using Fahlman's system would go to spam recipients. Still, Fahlman acknowledges that implementing his plan will require someone to manage the fee structure and the development of interoperable email software and phone sets. Skeptics include Coalition Against Unsolicited Commercial Email co-founder John Mozena, who says the widespread adoption of interrupt rights faces too many hurdles, including the fact that spam is a social trend rather than a technical problem. Meanwhile, computer language designer Paul Graham says such a plan suffers from a "chicken-and-egg" situation--"It's inconvenient until everyone's using it, but no one uses it until it's convenient." In addition, there are many other antispam products--filters, whitelists, and the like--that could hinder the progress of Fahlman's proposal.
http://www.wired.com/news/technology/0,1282,56788,00.html

Inspired by the work of mathematician Alan Turing, who postulated that an intelligent computer can fool a human interrogator into thinking it is also human, former Yahoo! chief scientist Dr. Udi Manber challenged a group of Carnegie Mellon University researchers two years ago to develop techniques to sniff out computer programs disguised as people in order to prevent them from entering portals to collect personal data on users. A team led by Carnegie Mellon professor and cryptographer Dr. Manuel Blum, the 1995 ACM Turing Award winner, devised a series of cognitive puzzles called Captchas that could easily be deciphered by people, but not by computers because of the limitations of artificial intelligence. The Gimpy puzzle, for instance, relies on the unique qualities of human vision that computers lack; one version consists of a distorted word superimposed on a cluttered background, and solving it requires the viewer to identify the word and type it out in a box. The puzzle has become a standard registration tool for Yahoo!. Another puzzle generates a distorted sound clip of a word or a string of numbers, which the user must decipher and type out as well. The Captcha effort has sparked research to develop better puzzles as well as programs that can solve existing ones. Dr. Jitendra Malik of the University of California at Berkeley believes that such programs could be useful in detecting trademark infringements online or automatically identifying military targets. Palo Alto Research Center (PARC) scientists led by Dr. Henry Baird have tweaked a text-scanning program to crack certain Yahoo-Gimpy Captchas, and at the same time are working on a text-based Captcha that PARC plans to market to e-commerce sites. Dr. Malik also used Captchas to test an object recognition method he developed with Dr. Serge Belongie.
http://www.nytimes.com/2002/12/10/science/physical/10COMP.html
(Access to this site is free; however, first-time visitors must register.)

The U.S. Defense Department is planning to develop an "enduring personalized cognitive assistant" (Epca) that reasons, uses represented knowledge, has behavioral and functionality awareness, accumulates knowledge, can explain itself, and is capable of responding to unexpected changes, according to a Broad Agency Announcement (BAA) from the Defense Advanced Research Projects Agency (DARPA). "We're not looking for superhuman behavior, like reading minds, but just commonsense reasoning that one would expect even from a child," notes project co-leader Ronald Brachman, who adds that Epca's first testbed will probably be the office, where it could learn to assist by observing and interacting with personnel. The BAA is calling for scientists to submit Epca design research proposals by Dec. 19; it is hoped that such proposals will detail a systems architecture that takes reactive, deliberative, and reflective processes into account. Integrating this architecture with long-term and short-term memories, perception, representation, reasoning, communications, and actuation will hopefully imbue the systems with commonsense. The BAA says that merely mimicking natural systems such as neural networks is inadequate--a better solution is to devise new architectures that can "harness raw computing power in powerful ways that brains do." The document also suggests using intelligent, user-adaptable interfaces. In addition to Epca, DARPA's Information Technology Processing Office will consider funding any proposed project that fulfills its requirements for a "new class of cognitive system." Brachman says that increasing computer speed and advancements in human brain research are some of the reasons why DARPA thinks artificial intelligence is a good bet.
http://www.eetonline.com/at/news/OEG20021209S0062

The U.S. military developed mesh networking--the creation of resilient, self-healing, and fast ad hoc wireless networks with no fixed points of failure--to improve battlefield conditions, but plans are underway to apply mesh network technology to the civilian sector, particularly in the areas of transportation and communications. Military products with mesh networking functionalities include the Enhanced Position Location Reporting System from Raytheon, and ITT Industries' Soldier Level Integrated Communications Environment (SLICE), which can coordinate voice communications between soldiers and map out their positions. MeshNetworks produces communicators that the Army demonstrated to police, fire, and emergency personnel earlier this month, and has licensed its systems to Viasys, which connects traffic signals, signs, sensors, and cameras into "intelligent transportation networks." The automotive telematics sector is expected to be one of the earliest users of mesh network technology, which could be embedded into onboard computers capable of downloading MP3 music or software updates; a mesh network of cars based in Germany is planned as part of a joint venture between Mitsubishi and Deutsche Telekom. Issues over data security and battery power consumption must be resolved in order to for mesh networking to catch on, and the technology's value can only rise with enough subscribers.
http://www.worldtechnews.com/p/9f/9014774408ed.html?id=107066f

The Association for Computing Machinery (ACM) has signed an agreement with Intel Corporation to help raise the visibility of the A.M. Turing Award. Intel's support will enable ACM to increase the cash award from $25,000 to $100,000. Widely recognized as the Nobel Prize of computing, ACM's annual Turing Award honors individuals for contributions of lasting and major technical importance to the computing field. "With the generous support of Intel, we can realize our goal of significantly increasing the award's cash value to better reflect the importance and prestige with which the Turing Award is held worldwide," said John R. White, ACM's CEO. Adds Intel CTO Pat Gelsinger: "With the Turing Award we recognize the scientists and engineers whose innovations have driven our industry forward, and at the same time we set a high standard for the next generation of science and technology pioneers who will deliver the ideas and inventions that fuel the global economy."
http://www.acm.org/announcements/funding.html

Arjun Heimsath and Hany Farid of Dartmouth College have devised a method to create three-dimensional digital elevation models of geographical regions using two-dimensional digital photos. Computer scientist Heimsath teamed up with Farid, whose specialty is earth sciences, to develop an approach in which at least three images of a digital region are captured with a digital camera at different vantage points, and then between 50 and 100 corresponding points are manually selected in each picture. The computer then uses a mathematical algorithm to automatically calculate the elevation map, Farid explains. "What was nice about the work, and what's representative of Dartmouth, is that I'm taking tools from the mathematics and computer vision community, and applying them to a real-world problem that Arjun works on," he says. This method saves researchers the trouble of carting heavy, cumbersome, and expensive equipment throughout regions, and allows mapping of areas that may otherwise be physically inaccessible. However, the researchers note that vegetation can affect the accuracy of the virtual landscape, while taking the photos from a good point of view is essential. Two students will travel to New Zealand next year to test Heimsath and Farid's method. The researchers detailed their work in the November 2002 issue of the Mathematical Geology Journal, and plan to make a presentation at the American Geophysical Union's annual conference in December.
http://www.sciencedaily.com/releases/2002/12/021206075955.htm

Accenture Technology Labs researchers such as Luke Hughes believe that business will be profoundly changed by the advent of pervasive computers and networked sensors, to the degree that there will be an online copy for every physical object and process that can be delivered in close to real time. "Reality online means that once you put the Internet in the physical world, you have the opportunity to suck the physical world back into the Internet and create a virtual copy of the physical world, or a digital double," he explains. Virtual doubles of business processes could enable managers to test strategies to accelerate time to market, increase productivity, or reduce waste, according to International Data (IDC) analyst Anna Danilenko. The same technology could allow doctors to diagnose patients before prescribing therapies, and help military strategists anticipate battlefield tactics. Integral to the success of online reality will be the development of "smart dust" and radio frequency identification (RFID) tags. Smart dust consists of incredibly minuscule sensors designed to be distributed over an area so they can relay data to a central command post via a wireless network. RFIDs are supporting the spread of silent commerce, in which information about products bearing such tags can be downloaded by external devices; Accenture Labs manager Paul Mackinaw says RFIDs will be used in the future to track product life cycles in order to optimize for changing business conditions, carry out troubleshooting, and order replacement parts, among other things.

At this week's Materials Research Society meeting, researchers discussed various initiatives to develop "electrotextiles," in which electronic devices and fibers are woven into fabrics. International Fashion Machines (IFM) has developed a denim jacket festooned with symbols and letters that generate musical sounds when pressed, while spokesperson Maggie Orth demonstrated a wrap with thermochromic ink patterns that change color in response to heat. Such technology could be used to make chameleon apparel for the military. Meanwhile, ILC Dover, NASA, and SOFTswitch have teamed up to produce soft keypads for spacesuits that would allow astronauts to direct robotic vehicles; ILC's David Cadogan says similar technology could be incorporated into diving suits or safety clothing. Electrotextile projects that the Defense Advanced Research Projects Agency (DARPA) is developing include a sheet equipped with acoustic sensors to detect sounds, and parafoils whose shape and texture can adapt to airflow to optimize steering. John Muth of North Carolina State University in Raleigh adds that electrotextiles could be used to monitor the flow of traffic or pedestrians by being positioned just under a road or pavement surface. For such technology to be viable, circuitry must be able to withstand getting pounded and tossed around in a washing machine, and continue functioning even when threads break.
http://www.nature.com/nsu/nsu_pf/021202/021202-11.html

One of the topics of discussion at this week's International Electron Devices meeting is a replacement for flash memory, which scientists expect to hit a manufacturing threshold later this decade. Chip manufacturers are gearing up for the transition from 130-nm chips to 90-nm, and the industry will move to a 65-nm manufacturing process in 2006. A replacement technology will be necessary once the 45-nm level is reached, according to Craig Sander of Advanced Micro Devices (AMD), the No. 2 flash memory manufacturer in the world. One alternative AMD is looking at is Quantum Well technology, in which data is stored in 5-nm wires, and Sander says the company will present a paper on it at this week's conference. The technology would be able to squeeze more performance at lower power, and Sander notes that its production cost would roughly equal that of today's flash memory. Flash memory, which retains data even when the power supply is removed, has become a basic component of cell phones, handhelds, digital cameras, and music players, and is showing up in cars, TVs, set-top boxes, and network equipment. Meanwhile, leading flash memory producer Intel has said it is investigating polymer ferroelectric RAM (PFRAM), magnetic RAM, and Ovonics Unified Memory as possible flash replacements. Texas Instruments will present a paper on its ferroelectric RAM research, and Motorola will discuss its work with magnetic RAM and SONOS technology.
http://news.com.com/2100-1040-976795.html

The speed and performance of computers--be they servers, laptops, or desktops--increases with the addition of transistors, but this generates heat that is becoming more and more difficult to manage, and raising the electricity demands--and costs--of keeping machines cool. In fact, a few years ago Intel's Pat Gelsinger warned that some day Pentiums could generate as much heat per square inch as a nuclear reactor. Cooling the system with water is one solution engineers are investigating, but other efforts seek to create chips that produce less heat. Intel's Banias chip, slated to debut in 2003, is programmed to deactivate idle transistors, but the company's Wilfred Pinfold says this can slow performance. He adds that Intel has tested chip elements that barely generate heat while running at 10 billion cycles per second. Key to this breakthrough are new techniques of chip manufacture that produce smaller, less power-consumptive transistors, and Pinfold says the technology should be commercialized by 2010. Meanwhile, Transmeta's Crusoe chip architecture features fewer transistors than Intel's Pentium, which results in less heat output and obviates the need for fans. However, there is a performance trade-off, because all PC code is required to be translated into Crusoe code before it runs.
http://digitalmass.boston.com/news/globe_tech/upgrade/2002/1209.html

The U.S. Defense Department's 2003 budget earmarked approximately $2.78 billion for the Defense Advanced Research Projects Agency (DARPA), according to the agency's Jan Walker; DARPA will in turn dedicate $124 million to micro-electromechanical systems (MEMS), microsystems, and nanotechnology research. DARPA invests in high-risk, high-payoff projects by distributing grants to industrial and academic researchers, which spares the organization from supporting infrastructure, notes Kei Koizumi of the American Association for the Advancement of Science (AAAS). The 2003 budget, which is 23 percent higher than 2002, will allow DARPA to continue ongoing research initiatives in MEMS and microsystems, as well as introduce new projects such as the $78 million Biologically Based Materials and Devices Program. However, funding for MEMS and microsystem research has been declining--only $28 million is expected for next year, compared to approximately $41 million this year. This may not necessarily hinder progress in these areas, as the technologies will be featured in other programs. A greater nanotech drive is indicated by the 2003 budget, including research and development of nanomechanical arrays and nanoscale biomaterials. The lion's share of federal R&D funding comes from the Defense Department, which finances roughly 40 percent of engineering research, the AAAS estimates. The 2003 budget included 4 percent more funding than DARPA originally suggested.
http://www.smalltimes.com/document_display.cfm?document_id=5159

With upcoming computer chips promising to generate greater amounts of heat--perhaps as high as 100 watts per square centimeter--and proportionally increase the risk of circuit malfunction, engineers must develop more efficient cooling methods. Mike Rightley of Sandia National Laboratories says conventional methods of spreading heat below the computer by fans works until the 100 watts/square centimeter threshold is exceeded and circuits start to melt. He has therefore devised a replacement for heat sinks in the form of a liquid-filled, self-powered "smart" heat pipe that diverts heat to the side edge of the computer, where it is dispersed by fans or air fins. The liquid--usually methanol--is changed into vapor by chip heat, which is released at a pre-selected site, after which it turns back into liquid and returns to its starting point to absorb more heat. The current desktop schematic puts chips next to the heat sink several inches high and wide, with the fan nearby; this architecture complicates engineers' attempts to stack chips in order to boost computational capacity and shrink computer size. Rightley explains that the optimum wick path to each heat source is achieved through the laws of fluid mechanics. He reports that the technology is being licensed to a startup with a major client in the civilian laptop sector, while the military could also use it as a thermal management solution for wearable computers. Rightley's work, which will be detailed in an upcoming issue of Microelectronics Journal, is funded by DARPA under its Heat Removal by Thermal Integrated Circuits (HERETIC) program.
http://www.newswise.com/articles/2002/12/HEATPIPE.SNL.html

Ideo cofounder Mike Nuttall has organized his company around integrating design issues with usability issues when developing products, and he has been instrumental in creating the first Microsoft mouse, as well as the Palm V, the Cisco IP Phone, and a rubber-gripped toothbrush by Oral B. Nuttall says the key to Ideo's success has been understanding a successful product always offers simplicity, and balances values relating to "technology, business, and people." He also says that most people want products to perform at the minimum levels very well, while technologists developing these products are inclined toward pushing for more and more. For instance, most people want cell phones solely in order to make phone calls, though the technology has greater possibilities. Nuttall says that if he would guess about the future, he predicts the next mass adoption will be wireless use in homes and offices with broadband functionality, though he also notes that technology experts believed broadband adoption would be more widespread in 2002 than it is in the United States. Nuttall believes U.S. broadband adoption will happen at some point, and says that high prices and legal issues concerning the distribution of online movies, music, and other content are stalling this next Internet phase. He praises Steve Jobs for being a pioneer of innovative design and still an industry leader of great design at Apple. Nuttall finds it hard to predict just when technologies such as flexible displays and handwriting and speech recognition will take off, and is less confident about the anticipated proliferation of wearable computers, although he says miniaturization will take off as might wearable peripherals such as hands-free wireless accessories and displays built into glasses. Nuttall says silver is now the dominant color for technology, although users will seek variety, while beige will remain common in offices.
http://news.com.com/1200-1120-975450.html

CIOs are increasingly turning to simplification of their business application software as a way to cut costs; they believe that trimming their list of suppliers and software packages means less spending on system integration, notes Gartner. However, this streamlining is complicated by the many classes of business software that came from the last economic boom--products that later turned out to be incompatible or hard for both employees and clients to understand. Day & Zimmerman CIO Anthony Bosco believes that today's focus should be on flexible software that is more conducive to integration, rather than specific software that may lose its applicability when business conditions change. As a result of supplier consolidation, specialty vendors are being discarded in favor of long-lasting heavyweights such as Microsoft, Oracle, SAP, Siebel, and PeopleSoft. Meanwhile, software companies are refocusing their energies on fulfilling customer needs rather than trying to develop products that outclass their rivals' offerings. Many mid-market customers shy away from the intricate and costly software offered by large companies, but this has not stopped major players from trying to tap into the market. Other companies are customizing products for niche markets. It is doubtful, however, that big business software firms will threaten companies that sell security or business intelligence software anytime soon.
Click Here to View Full Article

Researchers are more optimistic about performing useful computations using quantum computers in the immediate future as a result of new breakthroughs by Australian and Austrian researchers. At a conference at the University of New South Wales in Sydney in November, Robert Clark showed a silicon chip with a readout mechanism, and the device appeared to keep its quantum properties isolated from disturbance from the outside world. Clark said, "This was thought to be impossible just a few years ago." He expects to be able to perform processing tasks using the technology by 2007. Many observers believe that silicon chips using quantum bits (qubits) are the best hope for quantum information processing. Rainer Blatt of Innsbruck University in Austria also used the meeting to perform the first computation on a machine in a quantum state, using a single trapped calcium ion to execute the Deutsch-Josza algorithm quantum procedure to determine whether an imaginary coin was the same or different on each side. Researchers now believe they may not have to wait decades for a useful quantum processor. Imperial College quantum information researcher Peter Knight says, "There's been tremendously rapid progress in the last year. I was hugely impressed at how things have developed."

Dire warnings about the potential perils of nanotechnology, coupled with unrealistic promises of the technology's economic impact, are clouding important issues and could impede the field's progress. Some enthusiastic nanotech advocates claim that the nanotech business will be worth $1 trillion in 10 years, a figure that has attracted interest from investors, who later learn that some nanotech companies are focusing on more traditional microsystems, or are dressing up research-grant proposals as business plans. Proponents argue that the current definition of nanotech must be widened to include any system with nano-sized particles, but the many products falling into this category lack a common denominator. Nevertheless, the United States and other countries have jumped wholeheartedly into the nanotech research fray: Some $700 million has been apportioned for the 2003 budget of America's National Nanotechnology Initiative, while Europe has allocated a nanotech research budget of $1.2 billion for 2003-04. However, such enthusiasm is tempered by the reality of current nanotech breakthroughs; an IBM team recently reported that they were able to assemble individual molecules into a logic gate, but only through a painstaking--and painstakingly slow--process that could only be accomplished in a vacuum and at extremely low temperatures. Meanwhile, research groups and environmental organizations are testing nanomaterials for health risks, or demanding a moratorium on nanomaterial production until their environmental effects are measured. A research team at NASA's Johnson Space Center has determined that carbon-based nanotubes can be harmful if inhaled in large amounts, though their current applications pose little actual risk.
http://www.economist.com/science/displaystory.cfm?story_id=1477445

Internet Protocol (IP), upon which Internet traffic routing traditionally depends, is limited: Too much traffic can slow it down, while the task of handling multiple types of traffic and their specific network demands can lead to bottlenecks. Multi-Protocol Label Switching (MPLS) is designed to enhance IP by being able to accommodate traffic increases and handle diverse kinds of traffic, including less delay-tolerant types such as video and voice. MPLS can cut delays considerably because it relies on label-switching, whose function resembles the way ZIP codes are used by the U.S. postal system--it uses a label to forward multiple data packets to one destination without processing the entire IP address. However, for MPLS to be successful, the fundamental network must have sufficient bandwidth. The protocol is proving essential for Traffic Engineering initiatives, in which carriers measure and control clients' traffic to fulfill their quality-of-service requirements while staying within network performance parameters. MPLS is also increasing the reliability, scalability, and cost-effectiveness of Virtual Private Network deployments, enabling private network services to be built atop a public infrastructure. The adoption of MPLS is proceeding at a slow pace because of a lack of a proven track record. Service providers could keep risks low and revenues high by investing in an MPLS migration that involves a gradual transition to telephony-grade IP/MPLS network infrastructures.

Efforts to develop cyborg technology such as smart chip implants and neural-machine interfaces have yielded breakthroughs that some say could one day help disabled people conquer paralysis and move prosthetic limbs, allow consumers to control machines by thoughts, and even provide a vehicle for telepathic communication. About 20 years ago, scientists discovered that extraction algorithms and a computer could be used to determine the signal emitted by hundreds of neurons in tandem with a physical movement. Andrew Schwartz of the University of Pittsburgh's School of Medicine says that this revelation established the feasibility of neural prosthetics and other applications. Notable achievements since then include several American university experiments that enabled monkeys to interface with and manipulate electronic systems via brain implants. This year, John Chapin of the State University of New York's Downstate Medical Center was successfully able to control the movements of rats using an implant that tapped into the pleasure centers of their brains. Kevin Warwick, author of "I, Cyborg," is a living experiment of rudimentary neural interface technology: Electrodes surgically implanted into his median nerve are designed to receive nerve impulses from his wife through a computer connection. Neural Signals CEO Philip Kennedy believes that people should give serious consideration to such research. Meanwhile, various research teams, including one led by Schwartz and another led by Brown University neuroscientist John Donoghue, are racing to be the first to successfully give humans brain implants.
http://www.discover.com/nov_02/featbionic.html