Automated Driving: Driving Urban Development?: An integrated research-by-design computational modelling approach in urban planning problems

Complex urban planning problems comprise socio-spatial conditions, human behaviour and external factors (e.g. technological developments). The perspective of urban planners and designers is crucial to comprehend these problems from a holistic point of view but does not possess the skills to grapple the full problem. Cities bear characteristics of complex systems in which the interaction between decision makers is an important factor in the manifestation of urban change. Quantitative and computational methods are becoming increasingly capable at evaluating this aspect of urban planning problems. Computational models, although a simplification of reality, can help to evaluate many data and therefore extend the thinking capacity. Design thinking can interpret model outcomes in the broader socio-spatial context. But integration between these engineering-based methods on the one hand and the conventional urban planning and design methods proves challenging. This research investigates how an integrated methodology can lead to better understanding of urban planning problems. Models that comprehend built environment dynamics, can be used as design evaluation tool. The uncertainty around the spatial impacts of automated driving is used as case study for a complex urban planning problem. How automated driving will affect cities depends on the development and the deployment of the technology but also on how human agents adapt their behaviour. Mobility aspects (and therefore automated driving) relate to the built environment through the concept of accessibility and by the integration of the transportation system in space. An important condition to employ computational models within a research process is to explicitly define the system one evaluates. This research narrows down to households and their location choice behaviour as important behavioural aspect in urban development. The residential location choice depends on characteristics of the households itself and on dwelling, environment and location attributes. Automated vehicles are propagated to be more efficient in terms of infrastructure demands and travel efficiency compared to conventional vehicles. Therefore, this technology relates to the decision factors of households through spatial quality and accessibility effects. A scenario approach is employed to grapple the uncertainty around automated driving. In four scenarios, different development paths of automated driving are assumed. The province of Utrecht in the Netherlands is used as case study to provide for context and data. This research employs a residential location choice model to examine the changes in moving behaviour of households in space caused by the development of automated driving. The accessibility effects are studied with a transport model and the spatial quality effects are obtained by research and design. The residential location choice model evaluates these factors. The result is a change in urban development pattern. The results are interpreted in an urban strategy where the results of previous research steps are synthesised and assessed, to provide for interpretation of the research outcome from a holistic socio-spatial perspective. Computational models ask for explicit data which complicates the compatibility between various research steps. This research shows that it is possible to implement qualitative design methods within the data collection process. Design-based research methods are easier to apply and therefore relate directly towards the context of the problem. For using a computational model, this requires often for an operational model before one can comprehend the context. This makes for a more complicated research process. Therefore, employing an integrated methodology in urban planning research proves challenging but helps to approach the urban planning problem from a holistic perspective. An integrated approach broadens both perspectives to consider and combine the most important variables. By explicitly defining the problem, core principles reveal and computational models help to explain the dynamics between these core principles. Interpretation of the results in the socio-spatial context allows to account for the values that are inherent to urban planning problems. The method is labour intensive and does not guarantee more inspiring and detailed plans or designs. But the outcomes for proposed actions relate better to the factors that are significant to the system and elaborate more accurately on the uncertainty and dynamics the urban planning problem encompass. This research, is in which this report is established, is part of the graduation process for both the master of Urbanism at the faculty of Architecture and the master of Transport & Planning at the faculty of Civil Engineering. This report is submitted for the partial fulfilment of the Urbanism exam. The Transport & Planning report can be found here: http://resolver.tudelft.nl/uuid:5e88d71c-3878-4c4e-967b-0b5727bd0d8fArchitecture, Urbanism and Building Sciences | Urbanis