Self-organization is a basic mechanism by which complex urban systems organize themselves. This mechanism emerges from individual agents’ local interactions, often with unpredictable consequences at the regional level. These emergent patterns cannot be controlled by traditional hierarchical methods, but they can be steered and encouraged towards desirable goals. Self-organization is often used as an allegory for all ‘unplanned’ activity in cities. It is important to study the actual mechanisms of self-organization in cities to link the theory of self-organization to planning praxis. This work builds on ongoing work exploring novel complex planning tools and methods. Here I explore the key features of open dynamic systems identified in the literature as indicators of self-organizing capacity. I study their applicability in urban spatial planning, and propose three measurable characteristics for estimating the self-organization potential of urban activities. Flow reflects generic accessibility, and is measured using space syntax. Internal order refers to autonomously organizing entities, in this case the clustering tendencies of activities. Enriching rests upon increasing complexity and is measured as changes in degrees of entropy over time. The results indicate that (1) the study area meets the criteria for self-organization, and (2) these characteristics can be applied to discover nodes of higher potential for self-organization in a city.

New developments in Artificial Intelligence (AI) and digital ubiquity bring revelations of emerging smart cities. However, urban designers are particularly reluctant to become digital and use software that automatically generates cities. Instead, they return to traditional design skills such as creating scale models, sketching, notations and drafting. There is an increasing advocacy for design to a human scale, placemaking and liveable cities. This viewpoint asks questions about the application of AI and generative algorithms in digitizing urban design practices. It reflects on the possibilities of conjoining urban morphology and design theory into City Information Modelling (CIM) as a new digital tool for urban designers and reveals challenges in the ongoing development of new CIM software. Urban designers work within intricate design worlds with toolboxes that consist of customized design elements and symbologies. The design worlds consist of elements, rules and patterns and they act as holding environments for their unique diagrammatic design knowledge. CIM and AI should understand design worlds with customized toolboxes and provide help to automate repetitive behaviour patterns while designing.

No abstract

As the key aspects of theories of complex systems have been established, the premises for academic research on planning and on planning praxis still necessitates the development of novel planning tools and approaches to address inevitable urban self-organizing transformations. We have accepted that cities emerge from bottom up. However, planning methods simulating this emergence are still limited. Progress has been made in recent decades and many systemic, evolutionary, and computing based planning approaches have been proposed. The work here builds on these premises. Network theoretical, computational, and democracy discourses have proposed proxy or liquid approaches as for genuinely democratic forms of decision-making. More importantly, they enable information organization from bottom up in a digital platform. This process actually follows the very principles of self-organization of information in information or cognitive sciences: entropy decreases as the ''bits'' of information self-organize into coherent classes. These principles are also applicable in bottom-up planning. Hence, and to bring this discourse closer to the planning realm, I compared the conceptualized structures of Liquid Democracy, SIRN cognitive model and prior self-organizing planning proposals in a bottom-up planning experiment in Pispala neighborhood, Tampere, Finland. I evaluated its capacity for self-organization of information and hypothesized that the case provides a frame for a new self-organizing planning method. Based on this evaluation a structure for a digitalized Liquid Planning procedure is suggested and discussed.

This paper compares four different approaches to urban morphology: historico-geographical, process typological, space syntax, and spatial analytical. It explores in particular the use of four fundamental concepts proposed in these approaches: morphological region, typological process, spatial configuration, and cell. The four concepts are applied in a traditional gateway area of the city of Porto, Portugal. The area includes considerable variety of urban form. The main purpose is to understand how to combine and co-ordinate these approaches so as to improve the description, explanation and prescription of urban form.