Spatial cognition in indoor vs. outdoor planning: An experimental approach

The concept of spacescape has been studied by many scholars over time (Golledge, 1998; Lloyd, 2009; Lynch, 1960). Both landscapes and townscapes are knowledge-intensive entities that humans adapt for their life. Because of their dynamic complexity, spatial behaviours are hard to be simulated in mainstream AI robotics. Therefore, a question arises about the basic features, or ‘fundamentals’, of spacescapes by agents who live in and move through them. In fact, from the one hand, common sense claims that well designed space architectures make space more meaningful for humans than amorphous spaces. Conversely, the drama of social marginality in cities also depends on the abundance of landmarks and symbols, often inappropriate and alienating to poor people. Yet, a distinction emerges in literature and is worthwhile emphasizing in general. As a matter of facts, spacescape structural, ‘fundamental’ qualities may be opposed to spacescape ‘ornamental’ qualities in describing spaces (Goodman, 1951). This sort of ontological representation of space is essential for artificial intelligence and robotics, because of the inherent need of fine-tuning the characterization of space in planning automatic navigation. As there is circularity between AI and cognitive science, it is evident that developing robotic devices may in turn increase knowledge on human behaviours in space. Therefore, space imaging can be of great interest in strategic spatial planning, too, because it enhances the representation of the structural, invariant, resilient characters of the environment, for the development and management of human spaces. The present paper looks at space ontology as made by human agents. In doing this, it follows the cognitive approach used by AI robotics, with integrations coming from the expert knowledge of the planning domain. After an introduction in the first chapter, the second chapter shows two experimentations carried out in the context of space perception, imaging and navigation and discusses some results achieved. Brief conclusions and research perspectives are reported in the final chapter.