Abstract: A novel approach to automata-based modeling for spatial systems is described: geographic automata and Geographic Automata Systems. We detail a framework that takes advantage of the formalism of automata theory and GI Science to unite cellular automata and multi-agent systems techniques, and provides a spatial approach to bottom-up modeling of complex geographic systems that are comprised of infrastructure and human objects. The suitability of the framework is also discussed with reference to existing cellular automata and multi-agent systems models used in urban studies. Practical implementation of the framework is illustrated with reference to an object-based urban simulation environment and implementation of a popular socio-spatial segregation model.

Automata have the capacity to process information from their surroundings and to alter their characteristics accordingly. They are flexible and efficient abstractions that enable the construction of detailed, complex, dynamic models. The weakness of the CA approach is the inability […] in dealing with mobile objects such as pedestrians, migrating households, or relocating firms.

[…] the states S that are attributed to social-science agents are usually designed to represent socioeconomic characteristics, and agent transition rules T commonly correspond to human-like behaviors.

Until recently, the mainstream of MAS research in geography involved populating regular CA with agents of one or several kinds, which could migrate between cells, or simply reinterpreting CA as agent-based models, by attributing anthropomorphic state variables to cells.

a Geographic Automata System, G, consists of seven components:

G = (K; S, T_s ; L, M_L ; N, R_N)

where

• K denotes a set of types of automata.
• S is the set of states, and it consists of subsets S_k, characteristic of automata of each type k \in K.
• T_S represents a set of state transition rules.
• L denotes the geo-referencing conventions that dictate the location of automata in the system. It governs how geographic automata are registered in space.
• M_L denotes the movement rules for automata, governing changes in location.
• N represents the neighbors of the automata.
• R_N represents the rules that govern changes of automata relations to the other automata.

State and location transitions depend on automata themselves and on input given by the states of neighbors. The transitions occurs in the three sets S, L and N.

An automata can be referenced in the space directly (the automata has the geometry) and indirectly (the automata points to another automata). Two indirectly located automata can be considered as neighbors, when the automata they point to are neighbors.

TerraME: The concept of an agent having a Trajectory represents an indirect reference in the space.

Abstract: The concept of Geographic Automata System (GAS) formalizes an object-based view of city structure and functioning; OBEUS software implements this view on the operational level. The paper presents the GAS paradigm and latest user-friendly version of OBEUS, the latter based on .NET technology and developed according to OODBMS logic. OBEUS boosts further development of GAS theory, especially regarding the treatment of time in models describing collectives of multiple interacting autonomous urban objects. We claim that all high-resolution urban Cellular Automata and Multi-Agent models of which we are aware can be described in GAS terms and represented as OBEUS applications. GAS and OBEUS can thus serve as a universal, transferable framework for object-based urban simulation.