The Virtual Fishtank


The Virtual Fishtank is an innovative new museum exhibit, developed by the Boston Museum of Science, the MIT Media Lab, and NearLife Inc., with generous support from the National Science Foundation.

Museum visitors can:

  • create their own artificial fish
  • design behaviors for their fish
  • play with their fish in the giant fishtank
  • observe their fish interact with other fish
  • analyze the ecological patterns that emerge

    Using state-of-the-art 3D graphics and computer modeling, the exhibit introduces visitors to important ideas from the sciences of complexity — ideas that help explain not only ecosystems but also economic markets, immune systems, and even traffic jams.

    In particular, visitors learn how complex patterns can arise from simple rules.

    The first version of The Virtual Fishtank opened at The Computer Museum in Boston on June 13, 1998. The exhibit moved to the Boston Museum of Science in September 1999, after the two museums merged.

    The Virtual Fishtank has received press coverage in The New York Times, The Boston Globe, and The Boston Herald.

    For more information, see the Virtual Fishtank home page at the Museum of Science.

    Background

    During the 1980s, there was a ground swell of interest in the study of complex systems — systems in which complex patterns arise from interactions among simple parts. This surge of interest was due, in part, to new theoretical insights and to the availability of new computational tools for modeling complex systems. Research involving chaos, self-organization, adaptive systems, and nonlinear dynamics are all part of this broader interest in the new "sciences of complexity."

    Research into complex systems touches on some of the deepest issues in science and philosophy — order vs. chaos, randomness vs. determinacy, self-organizing vs. centrally-controlled systems. In the minds of many, the study of complexity is not just a new research branch; it is a new way of thinking about all science. Whereas scientists have long seen the world in terms of centralized controls and causes, the new sciences of complexity emphasize decentralized, self-organizing systems. The principles of complexity have the potential to change the way that not only scientists but also the general public thinks about natural, social, and technological systems.

    The Virtual Fishtank draws on scientific research in complex systems, and makes these ideas accessible to a broad public audience. In particular, the Fishtank draws on research from the new field of "artificial life," which focuses on computer modeling of animal behavior.

    The Virtual Fishtank also builds on recent research on learning and education. Constructionist theories of learning are based on the idea that people learn with particular effectiveness when they are engaged in design and construction activities. The Virtual Fishtank adopts this approach, enabling visitors to design (and not merely observe) the behaviors of fish. In line with recent research, we believe that these design activities will offer rich opportunities for learning scientific concepts.

    A number of commercial software companies have developed interesting systems simulations, both of fish and other natural systems. This software, however, generally does not allow users to create or modify (or even inspect) the models or rules underlying the behaviors. By contrast, The Virtual Fishtank gives visitors the opportunity to create the rules underlying fish behaviors, making it more likely that participants will understand the connections between underlying rules and emergent behaviors. In addition, most commercial software is designed for individual users, with little opportunity for group learning. The Virtual Fishtank provides a continuum of individual and group learning experiences. The exhibit provides a social context that encourages people to share and discuss their ideas about what they observe.

    Exhibit Activities

    The Virtual Fishtank includes a variety of different activities. In each activity, visitors make connections between the behaviors that they observe in the giant fishtank (projected on several wall-sized displays) and the rules that underlie the behaviors. The Virtual Fishtank is full of interactions: fish interact with one another, with the fishtank environment (including plants and coral reefs), and even with the visitors themselves (using sensors embedded in the exhibit).

    Visitor activities will include:

  • Build Your Own Fish. Each visitor can create a "program" for an individual fish, then "launch" the fish into the giant fishtank. Fish programs are based on simple condition-action rules. For example: if you see a shark, swim away from it; if you sense food nearby, swim towards it; if you detect a human nearby, follow it. Visitors can "personalize" their fish (for example, adding special stripes) so that they can recognize and follow their fish in the big fishtank.

  • Fish Schooling. There are several schools of fish in the fishtank. At special stations, visitors can modify the rules for individual fish within a school — and observe the changes in the overall schooling behavior. In this way, visitors can see how group behaviors can arise from simple interactions among individuals. The exhibit includes objects (such as sharks) that disrupt the schools — enabling visitors to observe how schools split apart and then gracefully reform.

  • Diving Deeper. Next to the giant fishtank, there are several stations showing examples of other complex systems. Visitors can explore the workings of ant colonies, simple market systems, and traffic patterns. The goal is to show that the mechanisms in the fishtank are not unique to fish, but can be used to understand many other phenomena in the world. A version of one of the Diving Deeper stations (called Exploring Emergence) is available on the Web. The other Diving Deeper stations are implemented in StarLogo software, which you can download for free from the Web.


    For more information, contact:
  • Mitchel Resnick, MIT Media Lab (mres@media.mit.edu)
  • Oliver Strimpel, The Computer Museum (strimpel@tcm.org)
  • Tinsley Galyean, Nearlife Inc. (tinsley@nearlife.com)