December
2004
Jack Welch with one of the several hundred antennas that will eventually complete the Allen Telescope Array. The reflective dots are used to measure the dish's surface accuracy. (David Pescovitz photo) |
The SETI Institute predicts that we'll detect an extraterrestrial transmission within twenty years. If that turns out to be true, it'll probably be the folks at UC Berkeley's Hat Creek radio observatory who will have heard the call. Right now, the Allen Telescope Array of more than three-hundred dishes is under construction at Hat Creek five hours north of San Francisco. Within a year, the first thirty dishes will be operational, forming the basis of a giant ear that listens for intelligent beings in space while simultaneously gathering data for groundbreaking astronomy research.
William "Jack" Welch, UC Berkeley professor of electrical engineering and astronomy, has been a driving force in the design and construction of the Allen Telescope Array (ATA) since the project first got off the ground five years ago as a joint effort between UC Berkeley and the SETI Institute. Named for major donor Paul Allen, co-founder of Microsoft, the array will eventually consist of 350 6.1-meter radio dishes electronically networked together into a radio telescope with unprecedented sensitivity. Precisely distributed across 2.6-acres on the Hat Creek grounds, the combined dishes will have far greater sensitivity than much more expensive 100-meter telescopes.
The SETI project scours millions of radio channels for narrow-band signals, indicative of intelligent origin. It's like listening for a station as you twist your car radio's tuning knob past all the static. Until now, SETI has used limited time from myriad radio telescopes around the world, limiting the number of stars that can be observed. However, the ATA will be dedicated to the project, speeding up the SETI search by a factor of 100. Meanwhile, the unique design of the system enables astronomers to monitor a huge range of wavelengths to observe other cosmic phenomena simultaneously with the SETI search.
Three prototype radio dishes now in place at Hat Creek Observatory in northern California. By 2007, 350 of these 6.1-meter-diameter dishes will be assembled to form the Allen Telescope Array, the largest radio array in the world. (courtesy Radio Astronomy Laboratory) |
"SETI is admittedly a long-shot," says Welch, holder of UC Berkeley's first Chair in the Search for Extraterrestrial Intelligence. "I don't have the patience to do only that, so it appeals to me to have a steady flow of other data for us to study as well."
For example, Welch and his colleagues will use the array to make a cosmological map of atomic hydrogen, the most abundant element we know of. Indeed, the visible universe may be composed of up to ninety-percent hydrogen. Determining its spatial distribution in nearby galaxies could provide insight into the evolution of the cosmos and the mysteries of dark matter.
"We'll be able to look halfway back to the beginning of the universe," Welch says. "The ability to observe that far back into time is limited right now."
Integrating the receiver directly into the antenna's pyramid-shaped feed dramatically improves the dish's sensitivity. |
To crank up the telescope's sensitivity, Welch and his colleagues devised a bit of ingenious antenna technology. In traditional pyramid-shaped antennas like those used in the ATA, the signal is picked up at the tip of the structure, called the feed, and runs down wires to the receiver. The problem, Welch explains, is that much of the signal gets lost along the way. To keep the signal as pure as possible, the Berkeley researchers shoehorned the receiver components inside the feed itself.
"It's just one new wrinkle for technology that was originally developed in the 1950s, but it enables our feed to essentially have no limitation on bandwidth," Welch says.
Right now, just three prototype dishes are being put through their paces at Hat Creek. In the next few months though, the researchers will install more than two-dozen others, nearly one dish a day. By Summer, Welch hopes this first small array will be scanning stars many light-years away. Whether ET is intelligent enough to call remains to be seen, or rather heard, but Welch is convinced that there's something out there.
"The recent discovery of planets around many nearby stars is a strong argument that our solar system isn't really unique at all," he says. "That in itself makes it almost certain that there are nearby planets with some kind of life on it."
Lego My Robot
by David Pescovitz
A Lego bot works its way around a barrier. (David Pescovitz photo)
|
A gaggle of tiny robots is making its way through an obstacle course, detecting barriers in its path and weaving back and forth along the ground. Some manage to make it from one end to the other, coming to rest at a light source glowing brightly at the edge of the course. They pause and then slowly rotate on their wheels to tackle the task again. This is not a test of new military technology or an experiment in artificial intelligence, but rather the culmination of an undergraduate course in robotics. And the contraptions traversing the terrain in this engineering building's basement were constructed from Legos.
The aim of Professor Roger Glassey's introductory robotics course is to instill students with the most fundamental skills in designing computer-controlled mechanical systems, and provide them with the discipline and stamina to solve difficult engineering problems systematically and efficiently. Basically, they should learn to suck it up when their robots crash-and-burn and return enthusiastically to the drawing board.
"I'd like to stir up the students' excitement about robotics, but also inspire them to continue their research and development until their projects actually work," says Glassey, professor emeritus of Industrial Engineering and Operations Research.
Roger Glassey with Anthony Levandowski and the BillSortBot, winner of the first Java Technology Lego Mindstorms Challenge. (courtesy Sun Microsystems) |
Teams of students in the wildly popular course are given Lego Mindstorms Robotics Invention System kits, consisting of traditional Lego bricks along with gears, motors, touch and light sensors, and a small microprocessor. The assignment is to design and construct mobile robots and program them in Sun Microsystems's popular Java computer language, co-developed by UC Berkeley alum Bill Joy.
"I played with Legos as a kid, so that's fun, but my weekends were totally taken up by coding and debugging,"says mechanical engineering student Juan Ruiz, who plans to study robotics as a graduate student. "It's worthwhile, though, once you see your robot start to move."
The students' robots were judged in timed races on a small obstacle course. Touch sensors on the robots' front ends enabled them to change course when they hit an obstacle. The goal was to reach desk lamp "beacons"detected by the robots' onboard light sensors. It's far harder than it sounds. But several of the students' creations managed to best their professor's personal robot.
"The key is having tight control over the sensors,"says Mario Garcia, a student with a double major in mechanical engineering and materials science engineering. "You need a really good control scheme to win this."
Students Mario Garcia and Juan Ruiz with their Lego creation. (David Pescovitz photo) |
And that all comes down to smart programming. While Glassey's research focused mostly on production planning and scheduling at semiconductor plants, he's no stranger to the elegant software design he teaches to his undergraduates.
"I've been interested in programming languages since I punched my first deck of program cards"in the earliest days of computing, he says.
Toward the end of his active research career, Glassey designed software to simulate semiconductor manufacturing systems. To do it, he leveraged the benefits of object-oriented programming, a paradigm similar to Java where data and functionalities are "packaged"into units that can be linked together.
"I thought it was a great way to code,"Glassey says. "I taught the robotics course once using a much simpler programming language, but when Java was released, it enabled the students to really increase the functionality of these simple robots."
Glassey has become a celebrity of sorts in the Lego Mindstorms community of garage hobbyists and student engineers. He was a pioneering user of leJOS, an open source bit of firmware that that enables the Lego Mindstorms robots to be programmed with Java. Three years ago, Glassey's student Anthony Levandowski led his classmates to victory at the first Java Technology Lego Mindstorms Challenge, sponsored by Sun Microsystems.
"I'm a great believer in competition as a motivator for research,"Glassey says.
Hari Dharan holds one of the prototype booms. The machine beside him is used to push the molding form out from the center of the rolled carbon fiber tube. (David Pescovitz photo) |
Two years from now, a group of five small satellites will ride a spacecraft "bus" into orbit where they'll gather data about Earth's magnetosphere. The satellites' 8-foot long articulated limbs, laden with various sensors, will measure the spectacular auroral eruptions behind such phenomena as the Northern Lights. It's the job of UC Berkeley mechanical engineering professor Hari Dharan and his students to ensure that the satellites' arms deploy properly and remain steady as they spin around the Earth at thousands of miles per hour.
Dharan is the director of the Berkeley Composites Labortaory, a state-of-the-art facility where he and his colleagues design, test, and fabricate new materials that will eventually be used in myriad structures, from buildings to airplanes to space vehicles. Composites, Dharan explains, combine filament and a resin bonding matrix to produce a new material with useful mechanical, electrical, or other high-performance properties. Fiberglass, for example, is one well-known composite.
Dharan's specialty is carbon fiber composites, materials in which carbon threads are woven or braided in specific directions to provide incredible strength with very low weight. These materials, Dharan says, are tailor-made for space applications, where the approximate cost of launching something into orbit is roughly $15,000 per kilogram.
"In addition to being very stiff and strong, the carbon fiber structures we build have one-fifth of the weight of steel," he says.
For three decades, Dharan has championed the use of carbon fiber composites for space vehicles. As an engineer at Ford Aerospace in the 1970s, Dharan helped construct the carbon fiber dish antenna for NASA's Voyager 1 space probe, now the most distant human-made object in space. Since then, he's spearheaded the recent shift from aluminum to carbon fiber in the construction of the International Telecommunications Satellite Organization's Intelsat communications satellites.
Graduate student Tyler Williams with the system he co-designed to test the boom. (David Pescovitz photo) |
Right now, Dharan and his students are fabricating essential components for the NASA THEMIS (Time History of Events and Macroscale Interactions) mission, at $173 million program led by the UC Berkeley Space Sciences Lab. A collaboration between NASA, four universities, and seven foreign nations, the THEMIS satellites are slated for launch in 2006. Outfitted with magnetometers, electrostatic analyzers, solid state telescopes, and other instruments, the five identical satellites will give an unprecedented "bird's-eye view" of magnetic substorms. The missions' aim is to help scientists determine how energy from solar wind is transported and explosively released, visible to us in the form of the Aurora Borealis and Aurora Australis.
Inside Dharan's fabrication laboratory at the Richmond Field Station, the researchers are constructing and testing the hinged booms that will unfurl to deploy the sensors once the 200-pound microsatellites are in orbit. This is the last step in a process that began with developing the perfect carbon fiber composite recipe for the job.
"The main challenge here is dimensional stability and stiffness," Dharan explains. "Our carbon composites are designed to have a thermal expansion that is less than 1/100 that of aluminum. Once you deploy an 8-foot long beam, you don't want it wobbling back and forth with the large temperature changes that exist in space."
The design constraints were non-trivial, Dharan says, and much stricter than those applied for terrestrial applications. For starters, the sensitive electronics onboard make iron and other ferromagnetic materials completely off limits. Meanwhile, any new material used in space must undergo an arduous approval process, a problem when the research and development cycle is short.
"Our approach is to invent new combinations of materials that have already been approved by adding ingredients and arranging the fibers in different ways," Dharan says. "After all, the fundamental research in my laboratory involves characterizing new composites, new configurations, and mechanical behavior under loading."
Once the researchers settled on a material, they constructed prototype booms in the machine shop. The carbon fiber "fabric" material is wrapped tightly around an accurate metal form and then cured in an oven into the desired shape. Then, the form is pushed out of the tube using an incredibly strong but precise device. The booms are machined, assembled, and subjected to temperature cycling and vacuum comparable to that of space. Finally, they're tested under conditions that must adequately simulate the zero gravity environment of space.
To accomplish that, graduate students Tien Tan, Tyler Williams, and Arezki Rahman designed and built the entire hinged boom assemblies. The boom deployment mechanism is tested on the floor of the laboratory atop a piece of smooth acrylic. Frictionless air bearings pistons enable the boom to glide across the acrylic panel on a cushion of air so that the researchers can measure and tune the system as if it's floating in zero gravity.
In addition to the sensor booms, Dharan's group is building part of the main body structure for each spacecraft. Once the booms, structures, and mechanisms are complete, they'll be integrated into the rest of the space vehicle at the Space Sciences Lab. Then, in October 2006, if all goes as planned, Dharan will watch as his research truly takes off again.
Benn Karne (BS '72, ME) became an engineer because he loved hot rods. Indeed, now he specializes in reconstructing vehicle accidents to determine the speed and position of the parties involved. Recently though, Karne has driven full circle. He co-produced Bonneville: Wide Open , a documentary about the unique characters who compete for land speed records.
The film was shot at Speedweek, an annual summer event at Utah 's Bonneville Salt Flats where unusual mobile machines, from powered barstools to souped-up diesel trucks, tear down the track at speeds up to 250 miles per hour.
"We wanted the film to appeal to a general audience but also to gearhead types, and I think we achieved that," says Karne, who made the movie with friend Steve Davy. "But we had to keep some of the speed secrets out. For example, if someone had a special motor detail or steering arrangement—and that was the stuff that really appealed to me as an engineer—the builders said it would put them years behind if we revealed those secrets to their competition."
Bonneville: Wide Open premiered last month at the California Independent Film Festival in Livermore where it won the Audience Award.