Professor Uses Nature as Inspiration for Engineering
Posted: February 19, 2008 at 1:00 am, Last Updated: November 30, -0001 at 12:00 am
Kenneth De Jong
Creative Services photo
By looking to nature and more specifically to how species adapt over time, Mason computer scientist Kenneth De Jong believes we can improve any number of things about the world we live in, including the personal computers many of us use every day. To illustrate this, he uses the example of an ant colony.
“[The ants] are able to work together as a complex adaptive system and create complicated things without blueprints or zoning boards. To accomplish this, a lot of adaptation and learning goes on during the lifetime of individual ants and across generations,” De Jong says.
Could we build better computers if we took inspiration from nature? This is one of the questions De Jong, who heads the Evolutionary Computation Laboratory in the Volgenau School of Information Technology and Engineering, is tackling.
“What I’ve been really interested in is whether you can make computers more adaptive and more flexible and more robust over time so that they do the adapting, not us.”
De Jong’s vision isn’t that far from reality. This kind of adaptive technology is currently used in robotics, an area of research in which Volgenau School faculty members are involved.
“We are looking at robots whose behavior evolves, changes or adapts over time by making the engineering more like nature,” De Jong says.
“Robots have a long association with industry and the military,” he says. “Robotic vehicles can be used in dangerous or unpleasant situations, like a nuclear waste site or the sewer system in New York City. A robot could help long-distance truckers navigate or provide an alert mechanism, because a robot doesn’t get sleepy or distracted.”
And in space exploration, De Jong says, “Since we really don’t know what is going to happen on Mars, it is more effective to have the robot learn and adapt as it goes than to constantly reprogram it from Earth based on what it encounters.”
One of his computational projects at the Krasnow Institute for Advanced Study, where De Jong is associate director and oversees the Adaptive Systems Lab, focused on host-pathogen interactions, using what researchers call “in silico science.”
“What they really mean by in silico science is performing experiments on computer models of the real thing you are trying to study,” he says. “My interests have always been interdisciplinary, so we spend a lot of time not just doing hard-core computer science, but also looking at ways the computer can be used for other disciplines.”
Working with the Potomac Institute for Policy Studies, De Jong and his team looked at the effects of a pathogen, inhalation anthrax, on the human body. Because a number of victims of the 2001 inhalation anthrax incident were treated at nearby Inova Fairfax Hospital, Inova has some rare data that documents exactly what happened to the affected postal workers over the weeks and months following their exposure.
“We used that data, together with other information, to build what I believe is the first attempt at a computer model of the effects of inhalation anthrax on the human body,” says De Jong.
While De Jong’s model is a prototype, he can see applications for this kind of computer modeling in areas such as cancer research and drug design.
“We now have the capability to build computer models with enough fidelity that they can be used in science,” he says. “This is not to replace the wet lab or field work, but to add another dimension to the research.”
De Jong is seen as a pioneer in machine learning. He is the founding editor in chief of the journal Evolutionary Computation and sits on the executive committee of the Special Interest Group in Genetic and Evolutionary Computation of the Association for Computing Machinery.
He also received an Institute of Electrical and Electronics Engineers Pioneer Award in the field of evolutionary computation and was recognized by the Volgenau School in 2007 with its outstanding research award.
De Jong will give a presentation on his research on Friday, Feb. 22, at 3 p.m. in Research I, Room 163, on the Fairfax Campus. The lecture, “”The Evolution of Evolutionary Computation,” is open to all.