WALKING, LEARNING
Although most of us never give it a thought, walking is one of the most sophisticated activities our bodies engage in. Sometimes called “controlled falling,” walking involves a complex and precise interplay among our brain, nervous system, and legs and arms — and if anything goes wrong, it can quickly turn into just plain falling. In an effort to nudge robots closer to a human-like mode of perambulating, an MIT team has designed and built a robot that can “learn” how to walk in just 20 minutes. The MIT robot, called Toddler, shares a design approach taken with machines from two other teams, from Cornell University and Holland’s Delft University. (All three robots, in turn, reflect lessons learned from toys capable of waddling passively down a ramp.) Using custom-crafted software, Toddler can learn to navigate different types of terrain by controlling the electric motors mounted in each ankle. Russ Tedrake, a postdoctoral associate in brain and cognitive sciences, says the device he helped create “is the first to learn to walk without any prior information built into its controller.” Tedrake’s advisor on the project was H. Sebastian Seung, an MIT professor of brain and cognitive sciences, and an investigator for the Howard Hughes Medical Institute.
BRAIN AND SIGHT
It’s long been known that our brains have ways of responding when their role is altered by disease or injury, and an MIT-Harvard group has now shown where and how that process works in the case of a major sight-impairing illness. The condition, macular degeneration, afflicts some 1.7 million Americans and is a leading cause of blindness. It’s a disease whose main effect is to destroy the macula — the central part of the retina — thereby gradually depriving victims of both their color vision and their ability to make out fine detail.
But MIT’s Nancy Kanwisher and Chris Baker, working with Harvard professor Eli Peli, found that over time the brain regions affected shift their role to processing so-called peripheral visual signals — information from areas of the retina that deal with tasks like night vision and “seeing out of the corner of your eye.” This reorganization may help those afflicted by macular degeneration see better with their remaining vision, says Baker, a brain and cognitive sciences postdoctoral associate. Future research will further explore the role of these brain changes. “It may be possible to develop rehabilitation strategies that exploit the increased cortical involvement that we found to partially compensate for the loss of retinal function,” notes Kanwisher, a professor of brain and cognitive sciences.
SUPERSIZED STARS
An MIT undergraduate is the lead author of a paper in a prestigious astrophysics journal reporting the discovery of three stars featuring the largest diameters of any normal stars known, more than a billion miles across. Emily Levesque, a junior physics major from Middleborough, Mass., who’s been interested in astronomy as long as she can remember, was part of a research team that studied 74 red supergiant stars in our own Milky Way galaxy. A supergiant is a massive star that’s nearing the end of its lifetime and is extremely cool compared to the sun. The researchers found that three of the supergiants have a diameter about 1,500 times that of our sun, which is roughly 865,000 miles wide — in other words, each of the gargantuan stars would engulf the five innermost planets in our solar system if it were to replace the sun. The finding drew coverage in a variety of news media, including USA Today and CNN. “The untold story of all the press coverage is that Emily took much of the data, reduced it, analyzed it, participated as a full colleague in figuring out what it all means, is the first author on the Astrophysical Journal paper, and presented the results at the American Astronomical Society meeting in San Diego,” says Philip Massey, an astronomer at Lowell Observatory in Flagstaff, Ariz., and Levesque’s supervisor on the work. Other researchers involved, besides those from the Lowell Observatory, are from the Cerro Tololo Inter-American Observatory in Chile, the Universite de Montpellier II in France, and the Geneva Observatory.
SEESAWING CLIMATE SYSTEMS
MIT climate researchers have discovered a pattern in which drought in South America’s Amazon River basin coincides with heavy rainfalls in Africa’s Congo River basin and vice versa. MIT’s Elfatih Eltahir and his co-workers first noticed the pattern in an analysis of the rainfall patterns in the two basins reflecting data back to 1997. In a later analysis of river flows dating to 1905, the pattern still held. Eltahir, a professor of civil and environmental engineering and an expert in the young science of hydroclimatology, says that for now, both the causes and effects of this fascinating interplay are unclear. He plans further research to try to get answers, though. “I’m interested in studying the physical mechanism behind this phenomenon and the extent of its global impact,” notes Eltahir, who in earlier work had identified linkages between current flows in the eastern Pacific and climatic conditions as far away as North Africa. Meanwhile, the faculty member says the findings to date should be helpful to farmers and others who have a direct interest in rainfall forecasts in the two affected regions.