8 Conclusion
8.1 Introduction
The literature review revealed that one of the approaches used to address future cities challenges is that of smart cities. Additionally, it revealed that due to the availability of smart infrastructure, open city data, and advancements in data analytics techniques, designers are making use of data-driven urban design processes to enhance evidence-based decision-making in urban design processes. While data-driven urban design has been gaining momentum in recent years, few attempts have been made to systematise data-driven urban processes by integrating design processes, data-driven cycles, and computational technologies, as well as integrative approaches across multiple urban design dimensions. Furthermore, the literature review revealed that current data-driven tools tend to restrict users, developers, and other stakeholders in terms of tool use, development, and sharing of improvements, as these tools are typically proprietary. Therefore, the aim of this research was to develop, demonstrate, and validate an integrated data-driven approach to assist in computational urban design processes. This research aim was achieved by addressing the five research objectives outlined in Chapter 7.
The first research objective was to understand the current data-driven urban design processes in practice and their enabling computational smart technologies. The second research objective was to systematise an approach for data-driven urban design processes. The third research objective was to develop a computational design toolkit prototype to assist computational urban design processes. The fourth research objective was to demonstrate the computational toolkit prototype through a scenario. Finally, the fifth and last research objective was to validate the integrated data-driven urban design approach among designers. These objectives were achieved, as presented in Chapter 7, across Sections 7.2, 7.3, and 7.4.
This chapter is divided into four sections. Section 8.1 introduces the chapter. In Section 8.2, a summary of the key findings is presented. In Section 8.3, the contributions of the research are discussed. Finally, in Section 8.4, the limitations of the research and potential directions for future research are outlined.
8.2 Summary of Key Findings
The aim of this research was to develop, demonstrate, and validate an integrated data-driven approach to assisting computational urban design processes. This objective was effectively achieved and validated by designers and data-driven designers, showing the feasibility and usability of the proposed approach. This section summarises the key findings.
- Systematisation of Data-Driven Urban Design Processes
This research has identified a gap in the field of data-driven urban design processes, such that, while practitioners and scholars use these processes, there is a lack of systematic structure to explain and support them in both academia and industry. Through the analysis of semi-structured interviews and a review of the existing literature, this study has developed a conceptual framework that integrates cognitive steps of design processes, data-driven cycles, and computational technologies, addressing this research gap. The proposed systematic structure has the potential to enhance evidence-based decision-making and provide guidance for urban design professionals. Furthermore, the research has demonstrated the feasibility and usability of the proposed approach through the development of a computational toolkit prototype, contributing to the development of new data-driven tools. The validation process involving traditional designers and data-driven designers has shown that a cohesive systematic approach implemented as a computational tool in a visual programming environment is promising, supporting the proposed approach’s face validity and feasibility. The key finding of this research reveals that while traditional designers and data-driven designers follow similar steps during the design process, there are crucial differences in how they integrate cognitive design steps with data procedures. In the traditional design process, documentation serves as the final step, creating a graphical narrative of the final design outcomes. However, in the iterative nature of data-driven processes, documentation takes on the potential re-formulator role of Function-Behaviour-Structure, adding new layers of potential reformulations and perspectives of exploration in the solution space of the design process. This study contributes to the advancement of knowledge by providing a systematic approach to assist data-driven urban design processes.
In summary, the systematic approach integrates the cognitive steps of design processes, data-driven cycles, and computational technologies. Additionally, beyond the conceptual framework, computational toolkit prototype implementation can be used and tested as a computational environment to assist “real-world” urban design projects in both academia and industry.
- Integration of a Data-Design Environment for Data-Driven Urban Design Processes
Urban design processes are complex and require designers to deal with multiple dimensions and layers of information from different datasets and software solutions. This research demonstrates that integrating design and data tools in a single design environment improves designers’ experiences in terms of interoperability and data accessibility, thereby enhancing the overall workflow of data-driven urban design processes. The comprehensiveness and cohesiveness of the integrated approach contribute to enhancing design workflows, and generating seamless collaborations between traditional and data-driven designers is facilitated through the development of interactive dashboards. Therefore, the use of interactive dashboards serves as a common ground for collaboration among stakeholders with different design methods.
The outcomes of this research highlighted that comprehensiveness and cohesiveness across multiple computational technologies, and their integration in a single visual programming environment, contribute to enhancing design workflows. Additionally, seamless collaboration between traditional and data-driven designers was enhanced through the development of interactive dashboards in the set of visualisation tools, creating a common ground that fosters discussion and collaboration among all involved stakeholders.
- Holistic Data-Driven Urban Design Toolmaking and Future Smart Cities
The increasing trend of designers becoming digital toolmakers, creating tools to design and plan for future (smart) cities, has raised awareness about the social and political implications of the use and development of these computational tools. This study has developed a computational toolkit prototype using exclusively Free and Open-Source software, publicly available on GitHub under a Free and Open-Source GPL-3 license. The proposed approach and computational implementation aim to foster collaboration among users and developers to further improve the toolkit in accordance with data-driven urban design community demands, in line with the principles of free and open-source development. This holistic approach contributes conceptually to the research objectives and computationally to the development of a usable prototype. Despite the prevalence of proprietary tools among data-driven designers, the findings indicate their willingness to use and develop open environments, recognising the technical and social benefits of free and open-source tools. This awareness regarding the holistic development of data-driven design computational tools has the potential to be expanded beyond the field of urban design into other areas of the Architecture, Engineering, and Construction (AEC) field, as well as other fields involving the development of socio-political technological artefacts. This could contribute to raising awareness of the implications of technology use by citizens, who can then become active agents of technological developments in future smart cities scenarios.
8.3 Contributions of the Research
This study has developed, demonstrated, and validated an integrated approach to assisting computational data-driven urban design processes. It provides a conceptual understanding and a practical approach to address data-driven urban design processes both in practice and in academia. By achieving the research aim and objectives, the contributions of this study include:
- Data-Driven Urban Design
One of the main contributions of this study is that it enhances our understanding of and provides a systematic structure for data-driven urban design. Although data-driven design has been explored by practitioners and scholars in recent years, it has often been conducted in an unstructured manner, as few studies have focused on providing a systematic view of the subject. Therefore, the developed approach advances the theory of data-driven urban design by providing a systematic view that integrates design processes, data-driven cycles, and computational technologies. Additionally, the developed computational toolkit prototype implementation provides a practical environment in which the approach can be tested and applied to other urban design scenarios, in both academia and industry. Although the scenario demonstrated during this study in Chapter 5 is specific to a neighbourhood in Australia, the developed approach and implementation is generalisable to other urban design contexts worldwide, as all the implemented computational libraries and services are based on free and open-source technologies and services that have good global coverage and are increasingly expanding due to the many volunteers and collaborators who support Free and Open-Source initiatives.
The findings from this study suggest that interoperability between multiple datasets is crucial for successful data-driven urban design workflows. Therefore, this research introduces an approach that enables more cohesive and comprehensive integration between data and analysis tools for data-driven urban design, offering data-driven urban designers an enhanced workflow that can lead to more effective and straightforward evidence-based decision-making.
Finally, this study provides another contribution to the field of data-driven urban design by shedding light on the similarities and differences between traditional and data-driven urban design methods. The research outcomes show that both traditional and data-driven designers follow similar steps during the design process. However, while in traditional design documentation is the final step of the design process, in data-driven design, the documentation process is part of the reformulation of the design process, expanding the potential design solutions through the creation of new perspectives via exploration of the design process due to the extra layers of possible reformulations. This contribution highlights a key benefit of using a data-driven urban design process.
- Computational Design and Digital Toolmaking
This study demonstrates that the use of visual programming tools has been accelerating the toolmaking process among designers. However, it also highlights that designers often use proprietary tools during their tool creation without considering the socio-political implications of such toolmaking. Therefore, this research presents a more holistic understanding of the toolmaking process, recognising these tools as socio-political artefacts that should be open for co-creation and empowerment of citizens in order to advance our understanding and vision of future smart cities, bridging the gap between human-centric and technologically-driven approaches.
In a more practical sense, this study also provides a new computational toolkit that expands the capabilities of Free and Open-Source visual programming languages, which can be used in the field of computational design from a more general perspective. Although the focus of this research is on urban design, the developed tools can be used in different design fields, as each tool was designed using a modular approach that allows for broader applications. For example, the “Read CSV” file tool can be used to read CSV files with any content, and the linear regression model creation tool can be applied to any problem that fits the model.
Another contribution of this study to the field of computational design is that, since the computational toolkit prototype is developed as a Free and Open-Source piece of software, its source code can be studied in at least two contexts. Firstly, it can provide material for computational designers who are more comfortable with visual programming environments to start exploring the “black-boxes” of node functions, studying how the textual code behind the node works, demystifying textual programming content, and providing a deeper understanding of the tools they are using. Secondly, it can be used as a template for developing other “plugins” for Blender-Sverchok, advancing the use of Free and Open-Source tools in the field of architecture and urban design, but also in other industries that use visual programming, such as visual arts, gaming development, and industrial design, among others.
8.4 Limitations and Future Research Directions
The research has limitations imposed by the restricted time and resources, as well as the research method employed. Despite these limitations, the study has attempted to minimise their impact and optimise the findings within its scope and timeframe. To address these limitations, future studies should consider the following issues:
- Further Explore Through Design Research
As elaborated in Chapter 3, the focus of this research is to develop a data-driven approach to assisting computational urban design processes. Given the defined focus of the study, it is critical to acknowledge that this research is not creating a specific approach to designing new urban forms, but rather it is providing an approach to “assist” design processes. Downton (2003) outlines that there are three types of research: for design, about design, and through design.
The first, for design, refers to an investigation conducted to assist in the creation of a successful outcome. The second, about design, aims to identify concepts that are relevant and intriguing for design, for instance, the practices and thought processes of designers during the design process. Finally, through design, the production of knowledge is undertaken through the act of designing, reflecting about the design process (Downton 2003). Although this research can be situated across the three types of knowledge production in design, since the framework of the approach produced is an artefact about design and its computational implementation can be defined as for design—through design, since its final product assists data-driven urban design processes, but its prototype development is a product of a reflective practice (Schön 2017), the knowledge cannot be dissociated from the production method, since it is embedded in the process of tool creation as well as in its purpose. The through design element may be further explored in future works, in which the knowledge production is obtained through the generation of data-driven urban design explorations using the proposed approach.
- Mix-method Analysis and Co-Design of the Design Process
The current study focuses on developing an approach to assisting data-driven computational urban design processes. The research is explored by the sixteen designers, equally divided between traditional designers and data-driven designers, during the semi-structured interviews presented in Chapter 4 and by the twelve participants, also equally divided between traditional and data-driven designers, during the validation phase presented in Chapter 6. The focus of both phases was to produce findings based on qualitative evidence to support the computational implementation of the proposed approach. However, future research may focus on analysing the cognitive steps of data-driven design processes and traditional design processes, using mixed methods from qualitative and quantitative analyses, such as protocol analysis and statistical analysis, to refine and solidify the outcomes of this study.
Another area of investigation could include understanding traditional designers and data-driven designers, and further exploring how to bridge the collaboration gap between traditional and data-driven designers, as noted during Chapter 6. This could include advancing user-friendly co-design environments involving multidisciplinary stakeholders.
- Expanding Scope and Technologies Beyond Urban Design
The current study is scoped on data-driven design related to the urban design field. However, the implications of this study may be expanded and further explored in other fields of the Architecture, Engineering, and Construction (AEC) industry in future work, since the findings discussed in Chapter 4 may potentially have implications, for instance, in the architecture and design fields if analysed from their field’s perspective. Future studies are needed to explore other fields of data-driven design in which the production of sociopolitical technological artefacts may also have implications.