[Back]

A. Faller:
"Thresholds of privacy in urban process modeling";
Talk: ACSP-AESOP 4th Joint Congress, Chicago, Illinois, USA; 07-06-2008 - 07-11-2008; in: "Bridging the Divide: Celebrating the City", A. Faller (ed.); (2008), 1 pages.

Running through a constant evolutionary process cities have developed a characteristic spacial order: each space/room is made accessible by at least one other space/room and on its part also makes other spaces accessible. This feature is repeated on a cascade of different scales. With every step the degree of publicity decreases and privacy increases. The lowest level of private order, of intimacy is the bottom of this structural hierarchy. In cities, the cascade ranges from centers of highest public order like airports to single rooms or offices. This structure is fractal because the basic scheme is scale invariant; it is linked in a self-similar way on different scale levels. The accessibility functions like a filter. On each level, the scheme works as a threshold in a way that it selectively reduces accessibility. The threshold between levels of privacy can be physical, or asking for authorizations of various kinds.
In this hierarchy of spaces the process of space-use also has an underlying fractal structure. Most of the urban processes that this work is focused on are stable, i.e., returning to the state where they started from. Stability with respect to processes doesn´t mean that nothing happens, but that whatever happens, happens over and over again. The epitomes of stable processes are rhythms. The rhythms that the space-use in cities follows are, again, organized in a fractal way: longer rhythms span shorter ones being synchronized at the common beats. The fractal structure of cities is perfectly adapted to these stable rhythms. Commercial functions are more conglomerated and residential functions are more decentralized.

Modeling and simulating such hierarchies of processes needs to take the fractal spacial structure and the spatiotemporal structure into consideration. A main part of this paper deals with the issue of semantic zooming in process models. The approach that I am following is to split processes into sub processes at the thresholds of privacy according to their spatiotemporal relevance.