Dr. Dobb's Journal October 2000
Among the highlights at the 2000 Congress on Evolutionary Computation (http://pcgipseca.cee.hw.ac.uk/cec2000/main.html) was a neural net system that taught itself to play checkers. Unlike IBM's Deep Blue chess-playing supercomputer, which was programmed to play chess, the neural net started only with the basic rules of checkers and had to teach itself to play. The neural net first spawned hundreds of checkers-playing programs, each slightly different from the others. These programs then played each other, and like the process of natural selection or survival of the fittest, losers were killed off and winning versions survived. After hundreds of iterations, the system -- developed by David Fogel, a computer scientist at Natural Selection (http://www.natural-selection.com/people/dbf.html) and Kumar Chellapilla, a graduate student in electrical engineering at the University of California at San Diego (http://vision .ucsd.edu/~kchellap/main.html) -- produced an expert checkers player.
The Smart Shirt, developed by SensaTex (http://www.sensatex.com/) and Georgia Tech Research (http://www.gatech .edu/), is a T-shirt with optical and conductive fibers woven into the fabric. When wearing the shirt, it can monitor your vital functions, such as heart rate, respiration, temperature, and blood pressure. The Smart Shirt project started in 1996 at the Georgia Institute of Technology with sponsorship from the U.S. Navy. As a commercial product, it will be used to monitor vital functions of astronauts, the chronically ill, elderly persons living alone, athletes, infants, and others requiring constant surveillance for medical reasons. SensaTex is in the process of gaining FDA approval and hopes that the Smart Shirt will be commercially available next year.
A team from McMaster University (Hamilton, Ontario; http://www.mcmaster.ca/) captured the IEEE's inaugural Computer Society International Design Competition. CSIDC 2000 (http://www.computer.org/ CSIDC) challenged undergraduate computer engineering students to tackle real-world problems. This year's challenge was to develop an information appliance that could improve public health. Each of the 50 teams selected from 10 geographic regions received a high-performance embedded computer system on which to build their projects. The winning team from McMaster designed a system called "Heart Mate" for monitoring heart rhythms and other heart functions. The students on the winning team were Rachita Kohli, Ajay Arora, Wai-Yin Shum, Christopher Lambacher, and Joshua Capogna. Their faculty sponsor was Markad Kamath. National Taiwan University won second-place for its Family Health Guard project, and the Technical University of Poznan in Poland came in third with its Health Care Information Appliance.
NASA's Marshall Space Flight Center is providing a web site (http://liftoff.msfc .nasa.gov/) for tracking the International Space Station. The site relies on a Java applet called J-Pass, developed by Patrick Meyer, a data systems engineer at the Marshall Center. J-Pass provides you with optimal visibility times for your locations; detailed sky charts can be printed for outdoor use. The screen tracking display system is provided by the North American Aerospace Defense Command (NORAD). On July 25, 2000, the site showed the rendezvous of the Russian service module Zvezda and the Space Station, which is orbiting at more than 200 miles above the Earth, circling the planet approximately 16 times per day at a speed of 17,500 miles per hour. The web site also includes an optional automated mailing list which will notify you by e-mail of upcoming satellite passes.
Want to really test the security of your organization? Okay, invite Sandia National Laboratories' Red Team to hack your system. So far the team has successfully intruded 35 out of 35 information systems that it has been invited to attack. Customers that have invited Sandia's Information Design Assurance Red Team (IDART) include several large corporations and government agencies, according to team leader Ruth Duggan. An invading Red Team usually consists of three to eight intruders charged with pilfering information, corrupting data, changing operational procedures, or causing denial of service. Generally, customers are warned ahead of time and are even given details of how the attack will occur. And even so, the attacks have so far all been successful. So much for the state of network security. For more information, see http://www.sandia.gov/idart.
Computer scientists at Bell Labs, the R&D arm of Lucent Technologies, and the California Institute of Technology have developed what they call a breakthrough digital geometry compression algorithm. Geometry in this sense refers to geometric representations of objects with enough detailed information about the size and shape so that the objects can be displayed, measured, and manipulated. Digital geometric data is typically acquired by 3D laser scanning and represents objects using dense meshes of millions or even billions of triangles.
The researchers, led by Wim Sweldens of Bell Labs' Mathematical Sciences Research Center and Peter Schroeder of Caltech's Computer Science Department (currently on leave at Bell Labs) report that their technique for geometry compression is 12 times more efficient than the method standardized in MPEG4 and six times more efficient than the best previously published method.
Several aspects of the Bell Labs/Caltech approach set it apart from other research in digital geometry processing. One is the team's original use of wavelet transformation. Improvements in digital geometry compression, which are measured in terms of the number of bits per vertex needed to describe a mesh of triangles within a given margin of error, can be exploited in the same ways as gains in other kinds of compression. Application designers will be able to trade off bits or bandwidth for the quality of 3D representations. Tested against other approaches, the Bell Labs/Caltech method proved to be superior across the board and especially effective in enabling high-quality reproduction with relatively few bits. For more information, see "Progressive Geometry Compression" and "Normal Meshes" at http://cm.bell-labs.com/ who/wim/papers/compression and http:// cm.bell-labs.com/who/wim/papers/ normalmesh.