Partners in the street ballet: An embodied process of person-space coupling in the built environment

Authored by Alasdair Turner

Date Published: 2017

DOI: 10.1177/0265813516638185

Sponsors: No sponsors listed

Platforms: No platforms listed

Model Documentation: Other Narrative

Model Code URLs: Model code not found

Abstract

The diurnal movements of pedestrians in the built environment are sometimes typified as a `street ballet', where each actor or dancer has their own set role within a larger complex. Every individual in the ballet may have many influences on their behaviour including the physical layout of the environment, cognitive strategies to navigate it, experiential or affective preferences as well as social, economic and political factors, but ultimately each one seems to obey apparently choreographed actions. The aim of this article is to understand whether or not there is in fact an underlying choreography to the ballet, in that certain steps or moves are more likely than others, such that a ` dance' through daily life is constructed. To do so, simple automata that use active perception to inhabit the world are evolved against different tasks within the environment, representing different sets of moves that may be taken. It is shown that any evolved automaton appears to embody a mathematical person-space relationship that joins visual affordance with motor action: the convergence of a simple Markov model of visual movement. From the Markov model, a general model of embodied action in the environment is proposed, whereby memory of the dance is ingrained over evolutionary history, such that it forms building blocks for non-discursive action within the built environment and comprises a possible common phenomenological framework.

Tags

Agent-based models behavior natural movement Markov model Artificial neural networks evolutionary robotics Embodied space Proto-visual automaton Ambient optic array Animat model Affordances Artificial evolution