Vermont Environmental Consortium

And

BRATTLEBORO REFORMER

November 6, 2007
UVM’s Domenico Grasso Engineers Unity from Complexity
by
Daniel Hecht

Dr. Domenico Grasso, Dean of UVM’s College of Engineering and Mathematical Sciences (CEMS) was trained as a chemical engineer, but his approach to his discipline suggests a more encompassing range of interests and perspectives. He thinks holistically, systemically, moved by an impulse to see or create coherency in our rather chaotic world.

Seeing things whole, getting the parts to cooperate, requires combining knowledge from different disciplines. It’s precisely this outlook that the times demand.

Of medium height, dark haired, Dom is not an assertive presence in a crowd. He tends to let others speak. But his listening is undeniably authoritative — he radiates an understated but acute attentiveness.

In his various roles – he’s dean of CEMS, Vermont’s largest center for engineering education and research, he chairs the Governor’s Commission on Environmental Engineering, he’s active on many boards and commissions – Dom is a mover and shaker in “upstream” developments in Vermont. These are innovations in technology, and in thinking, that may not yet appear on the radar of the media or the public, but soon will. They’re being researched, brainstormed, built, and tested right now, and they hold promise to change the world.

When asked what engineering innovations hold the most promise for the environment, Dom is an unabashed proponent and enthusiast of complex systems theory.

The term “system” presupposes some level of organization of phenomena. “Complex” systems are indeed complicated; however, central to the concept is that they’re characterized by not by their parts, but by interplay among diverse things and actions. They are seen as having “emergent properties”: Not only is the sum greater than its parts, but that sum is evolving, learning about itself as it goes; things are continuously adapting and becoming.

Complex systems theory has many applications, such as the study and treatment of disease. But in the environmental context, its value is to let us see natural phenomena in their full complexity and interdependence. Every part influences every other in feedback loops that can’t be perceived by our senses but can be modeled by supercomputers in various ways that allow us to better understand. Weather, watersheds, ecosystems, animal populations, and human behaviors are all phenomena of such complicatedness and interactivity as to require a complex systems approach.

“The complex systems outlook allows us to see a more holistic picture of problems,” Dom says. “We can not only look at natural systems and understand them, we can look at the legal, social, and economic aspects of environmental problems, and how best to integrate solutions into society.”

Some of the most fascinating work done at CEMS’s Complex Systems Center relates to artificial intelligence, or AI. Though we take our thoughts and actions for granted, what we call intelligence is incalculably complex.

Take, for example, the phenomenon of proprioception – knowing where our body and its limbs are. We can scratch an itch on the back of our neck because built into our neural circuitry is a “model” of our own form. That model knows the length and flexibility of our arms and hands, and where that itch signal is coming from, allowing us to find the right spot. We take the capacity for granted – but it’s extremely complex.

“Resilient Machines Through Continuous Self-Modeling” is a Complex Systems Center research project by Joshua Bongard that explores this “self imaging.” The video of his “resilient” robot shows a little mechanical creature with four flexible limbs, resembling a fat starfish. Sensors in the limbs report what it does and how it’s oriented relative to the tabletop; linked by wires to a computer, it is given a simple instruction – move in a certain direction, say. The robot contorts, arches, scrabbles, in a seemingly agonized attempt to organize its parts.

It remembers which actions accomplish the objective, and in doing so it assembles a sense of itself. How many limbs does it have? Where do they bend, and how much? Which movements accomplish motion, and which do not?

Bit by bit, it assembles a model of itself that is reasonably correct, that we can see on the computer screen. It is “resilient” because, when Joshua detaches one of its limbs, it painstakingly relearns its new shape and capacities.

Imagine the complexity of our huge, majestic world of gaseous atmosphere and liquid oceans and innumerable life forms! We’ve barely begun to understand it.

Dom envisions a future in which human systems become so smart and so interconnected that they’re virtually conscious. “I think we’ll see semi-alive infrastructure,” he says levelly, as if everyone knows what this is.

Functionally, it means building sensor arrays into our buildings, roads, vehicles, and natural landscapes. Computers will correlate data from soil, air, water, and plants with data from the human things. By seeing the interaction of all the components, we’ll understand how to minimize negative environmental impact.

Perhaps complex systems theory is part of humanity’s struggle to acquire our collective proprioception, a way to envision a new model of our species -- one interdependent part of a much larger, complex, magnificent whole.

To watch Joshua Bongard’s robot in action, go to www.uvm.edu/~cems/complexsystems/, link to the Research Projects page, and click on “Resilient Machines Through Continuous Self-Modeling.”

###

Daniel Hecht is a novelist and executive director of Vermont Environmental Consortium. For more information on any Green Grapevine topic, contact vec@norwich.edu.

<< Back