S C I E N C E & T E C H N O L O G Y
Scientists trying to build a better robot are finding some of the answers in Mother Nature's complicated handiwork: the human body. At Hopkins, researchers are designing new models in part by mimicking joint movement and the human sense of touch.
In some cases, they're doing biology one better. "Biology has a lot of baggage carried on through evolution," says Gregory Chirikjian, associate professor in the Whiting School's Department of Mechanical Engineering. "As engineers we can engineer from scratch."
The more agile a robot, the more likely it can assist surgeons in delicate operations or aid geological exploration teams on Mars--as well as better perform the simple robotic duties of lifting things and maneuvering about.
David Stein, left, and
Gregory Chirikjian study a spherical motor created to improve
Photo by Keith Weller
Chirikjian is leading a group of Hopkins engineers who have
invented a spherical motor that can more easily rotate in
all directions, much like the human ball-in-socket joint.
Such a deve-opment could give robotic arms better range,
precision, and flexibility, as well as reduce breakdowns.
Though robotic spherical motors are not new, the Hopkins researchers have created a software-driven electromagnetic system they believe will allow greater range of motion (among other facets, they've mounted 80 permanent magnets inside a hollow sphere, which is then nestled in a "saddle" laced with electromagnets that can be turned on and off). One benefit: the elbowlike joints most robots now use could be replaced by more versatile shoulder-style joints that don't require clunky, error-ridden conventional motors.
Chirikjian, doctoral student David Stein, and Edward Scheinerman, professor of mathematical sciences, have filed for patents on components for the device, and Stein is still working on a prototype.
Elsewhere at the Whiting School, mechanical engineering professor Allison Okamura has created a Haptic Exploration Lab. Okamura and a team of student researchers are testing ways to help robots better examine objects through touch. The group is also developing virtual environments where people can "feel" surfaces using computer joysticks or mouses. (When the cursor touches a sphere on the computer screen, for example, a user experiences a sensation something like a pencil point bouncing off a rubber ball.)
Allison Okamura develops
methods to give robots a more refined sense of
Photo by Jay Van Rensselaer
Such advances could be used to enhance the feedback provided
by robot explorers in space or in underwater scientific
missions, or to aid surgeons in conducting long-distance
surgery. "You have the robots do the touching, and humans
can explore as if they were there," Okamura says. "You don't
just get a picture, you can interact."
For example: "An underwater robot could look around at artifacts on the ocean bottom when you don't want to disturb the site. But scientists who would like to touch the objects can 'feel' the texture or get the general shape," explains Okamura, who is working with Hopkins's Engineering Research Center for Computer Integrated Surgical Systems and Technology.
"Or take a remote planetary geologist," Okamura notes. "Geologists like to feel rocks, and remote planetary geologists can't do that. The robot could send information back through the sense of touch."
To better create the informational relay, researchers might need to replicate part of what goes on in the brain. So Okamura is studying the work of Ken Johnson, Hopkins professor of neuroscience, and Steven Hsiao, assistant professor of neuroscience, at the Kennedy Krieger Institute. Among other studies, Johnson and Hsiao have looked at neural patterns in monkeys to determine which areas of the brain light up when the animals detect roughness, such as Braille characters.
"We've looked at how neural activity conveys the information about the tactile stimuli to the brain," Johnson says. "It is all a very complicated problem and robotics people need to understand this in order to emulate it."
Okamura, who earned her PhD from Stanford last June, did her dissertation on robotic exploration, work that includes the design of robotic fingers that explore bumps and cracks. --Joanne Cavanaugh Simpson
Particle physics requires collisions. Much of the experimental work in the field demands huge, expensive machines that fling subatomic particles at extraordinary speed into collisions with other subatomic particles. Existing circular colliders may not be the best technology for newer experiments, and this prompts the question: What to build next?
Morris L. Swartz, Hopkins professor of physics, favors a next-generation linear collider that smashes electrons and positrons together. He is chairman of the American Linear Collider Workshop that convened at Hopkins just as this issue went to press. The workshop, third in a series of such meetings, was expected to bring together 100 particle physicists from across the United States to discuss the science that would be made possible by a new linear collider.
|A representation of an electron-positron collision in the SLAC Linear Collider at Stanford University.||
Particle physics has been approaching what might be termed a
theory vs. hardware impasse. For theoretical physicists, the
so-called Standard Model has worked well explaining the
fundamental structure of the universe, but it still leaves
some basic features unexplained, such as why some forces are
so much stronger than others. Theoreticians, says Swartz,
have been working on a new theory called "supersymmetry," or
SUSY. If SUSY is valid, then each of the Standard Model's
subatomic particles--quarks, neutrinos, leptons, photons,
the whole subatomic men-agerie--has a supercounterpart,
"partners" not yet glimpsed because current circular
colliders are not powerful enough to produce them.
Supersymmetry, says Swartz, fits very well with what else is known about the complexities of particle physics, which means there's a host of additional particles out there, more than double what are currently known, just waiting to be seen. Furthermore, the Higgs boson, the so-called "God particle" that physicists believe actually makes atoms possible, has yet to be glimpsed. Scientists think they're tantalizingly close to finding what the theorists predict. Says Swartz, "There's a lot of evidence about what ought to be lurking out there, just beyond the reach of our current colliders."
Finding what may be lurking requires much more powerful machines. The Fermi National Accelerator Laboratory in Illinois has a newly renovated circular collider called the Tevatron. And in five years, CERN, the European Organization for Nuclear Research, will bring online the Large Hadron Collider, a circular collider that smashes protons together. But circular smashers have problems, Swartz says. To significantly speed up electrons sent whizzing through its tunnel, a circular collider must tremendously increase the amount of energy it applies. A hadron collision like that of the CERN collider produces what Swartz calls "lots of uninteresting stuff," noise, so to speak, which scientists must work through to find the pertinent results.
Swartz and other workshop participants favor an electron-positron linear collider. Linear machines are smaller and can achieve higher speeds with much smaller increases in energy. Electron-positron collisions are much "cleaner." One drawback: linear colliders are more complicated--and potentially more costly--to build and maintain.
The central dilemma is that this is Big Science, which carries a hefty price tag in the hundreds of millions of dollars. Through grassroots efforts--such as the Hopkins workshop--physicists hope to influence the decision as to whether to build a collider, and if so, what kind.
Swartz says that German scientists are about to submit a proposal to build a linear collider under Hamburg, and that Japanese physicists also have an effort under way. Noting the demise in 1993 of the proposed Superconducting Supercollider, after projected costs swelled from $5 billion to $11 billion, Swartz says, "The U.S. lost a big program and physicists have focused on CERN's supercollider in Europe. A large part of our community is moving to Europe. It's a question whether we can get them to come back again." --Dale Keiger
"We have the solar system here," insists Bruce Marsh, professor of earth and planetary sciences, gesturing to the red steel cabinets and wooden crates lining the corridors of Homewood's Olin Hall. Inside are chunks, slices, and dust from every corner of the Earth and even some places beyond, each neatly labeled, cataloged, and filed away.
There are metamorphic rocks from Vermont and mottled black-and-white gabbro from Sweden; cuprite from Cornwall, England; shiny white opal from Rattlesnake Canyon, Washington; rusty iron limonite from Lancaster, Pennsylvania; clay from Lake Magadi, Kenya; not to mention a sprinkling of garnets and other gemstones here and there. The cabinets also contain the occasional fossil-- 500-million-year-old trilobites from British Columbia, for example--as well as pieces of that province's famous Burgess Shale, the priceless treasure trove of Cambrian age life forms.
Bruce Marsh with a fraction
of the collection of rocks, minerals, fossils, and space detritus
stored in Olin Hall.
Photo by Jay Van Renssalear
But the collection, which numbers in the tens of thousands,
represents a historical record as well. This becomes clear
when Marsh slides open drawers containing the extensive
collection of George Huntington Williams, founder and first
chair of the department (then called geology), who taught at
Hopkins from 1883 to 1894. Rocks from the Williams
collection are all roughly the size and shape of a bar of
soap. Geologists took great pains to "cob" their rocks in
Williams's day, explains Marsh. They hammered and beveled
the edges of a rock shard all over until what remained had
no rough edges and could be easily held by hand. Though
cobbing fell out of favor among geologists, Marsh still
requires his students to learn the craft.
Marsh's tour ends with a grand finale: an iron meteorite that was discovered in 1836 in Namibia. It's the size of a toaster, but seems to weigh as much as an anvil. At 4.5 billion years, it is about the age of the solar system. --Melissa Hendricks
Sun-born electromagnetic storms in outer space can disrupt power grids on Earth and knock out satellite communications, or bring bursts of penetrating radiation that threaten astronauts and spacecraft. Yet space weather has long been more difficult to forecast than even its earthly counterpart.
Researchers at Hopkins's Applied Physics Lab (APL) may have found a better way to predict the storms, a development that could help prevent disastrous breakdowns of global communication and other networks. It would be a big leap. Some say space weather forecasting is about where ordinary weather forecasting was a century ago.
"As we become more dependent on a high-tech infrastructure, we become more vulnerable," says Brian Anderson, an APL principal staff physicist.
|Illustration by Charles Beyl||
Anderson and his APL team are measuring magnetic fields in
the Earth's ionosphere to better monitor the invisible
electrical currents that create space weather--storms of
electrically charged particles laced with magnetic fields
coming from the sun. The scientists have been able to create
global maps of the currents, which will prove helpful in
Last summer, a major solar flare created a magnetic storm that caused power outages in New England and Wisconsin; satellite watchers lost track of 1,300 orbiting objects. At the high point of the sun's 11-year sunspot cycle (this past year has been one of "solar max") such storms can occur as often as once or twice a month, and subside to once every few months at the lowest ebb.
Anderson's research, sponsored by the National Science Foundation, uses magnetometers carried on the 66 satellites of the Iridium System communications network, a onetime Motorola-funded telecommunications project. The satellites circle the globe in 470-mile-high polar orbits. APL scientists have developed techniques to interpret the satellites' magnetic field readings, and then extract the signatures of electrical currents. To create maps, they also use radar signals bounced off the polar regions to measure variations in the electric field (that radar network, known as SuperDARN, was developed by APL scientist Raymond Greenwald).
Previously, scientists relied mostly on ground-based magnetometers that monitor storms in the polar regions partly by tracking the beautiful aurora the currents create. Yet those detectors haven't been able to keep up with the storm as it moves across the globe. "It's a little like trying to understand a hurricane," Anderson notes. "When the thing really goes gangbusters, the detectors are in the wrong place."
The new approach also builds on a number of weather satellites and devices, including ACE, the Advanced Composition Explorer--an APL-developed spacecraft orbiting one million miles from Earth in the gravitation pivot point between Earth and the sun. That craft, whose detectors measure solar wind (including speed and density), give a one-hour advance warning of the electromagnetic field clouds. What's not yet fully understood is where storms will go and how they will behave if they hit Earth.
Though real forecasting abilities are likely a few years away, APL scientists have so far shown "hot spots" of electromagnetic energy to be more concentrated than previously thought. Such "hot spots," which act much like thermals over a scorched Texas desert, can create atmospheric drag on spacecraft, among other problems.
Better predictions would potentially allow satellite operators or power companies to shut down functions to protect equipment and give airline personnel time to take precautions. --JCS
The Johns Hopkins Magazine | The Johns Hopkins University |
3003 North Charles Street |
Suite 100 | Baltimore, Maryland 21218 | Phone 410.516.7645 | Fax 410.516.5251