Flock is a new exhibit at the Exploratorium. Through Flock, visitors can interact with a mathematical model of swallow flocking behaviour.
Swarming simulations model the observed behavior of animals that are flocking, herding or schooling. The original observations and algorithms were created by computer scientist Craig Reynolds in the mid-1980s. The Flock exhibit enables visitors to interact with the three parameters that produce the behaviour:
• cohesion: the desire of each animal to move toward the center of its closest neighbors
• alignment: the desire to align speed and velocity with nearest neighbors, and
• separation: personal space around each animal.
Imagine a sphere with each animal at the center. Each animal is only reacting to the other animals in that sphere. In this way, a larger (beautiful, swooping) pattern of intelligence emerges. Changing the sizes of the spheres for cohesion, alignment, and separation produce the varying behaviours.
Parallels to human existence are obvious, but untested. These may be explored in a future exhibit.
Visitors adjust knobs on a physical interface to change the cohesion (grouping on the label), alignment, and separation to see how the model changes. Goals for the exhibit can be grouped into levels of engagement, not necessarily in linear order:
• Recognizing that each knob controls a flocking rule, and that this in turn affects the behaviour of the flock
• Strategizing ways to test the effects of the rules on the flocking model (the primary way — setting the knobs to middle position and turning one knob at a time — is suggested by the label instructions)
• Interpreting the effect of the parameter changes on the behaviour of the flock, asking productive questions
• Understanding that this is a mathematical model
• Contemplating the relationship between the rules and behaviour in their own way, for example, wondering if flocks of birds and flocks of fish follow slightly different rules and to what effect, whether there is a relationship between groups of people and flocking, what else we can model mathematically, and what a model does and doesn't capture.
Exploratorium exhibit development has a forty–year history of iterative, formative development through testing exhibits on the floor with visitors. Originating with interative exhibits that were non–computational, the Exploratorium process parallels the user–centered design process found in the human–computer interaction field.