6 Validation Phase

6.1 Introduction

In the previous chapter, the development of the computational toolkit prototype was described. This chapter presents the validation process that was carried out through Focus Group Interviews with traditional designers (Group 1) and data-driven designers (Group 2), as shown in Table 6.1, aiming to gather feedback and different perspectives on the subject being investigated, with the goal of achieving convergence of ideas. The validation process involved twelve participants, who were evenly divided between traditional and data-driven designers. Seven of these participants also participated in the semi-structured interviews, which were discussed in Chapter 4. This face validity validation process is carried out to ensure the quality of the proposed approach, certifying its feasibility and usefulness in assisting data-driven computational urban design processes among designers.

This chapter is structured into five sections: Section 6.1 introduces the chapter, 6.2.1 outlines the focus group structure and the scenario used to demonstrate the toolkit’s functionality to the participants. Section 6.2 presents the findings from the focus group validation. Section 6.3 presents the implications of the validation findings. Finally, Section 6.4 concludes the chapter.

6.2 Focus Group Interviews

Given that the participants from the previous semi-interviews were located worldwide, as described in Chapter 3, the focus groups were conducted in three rounds, with participants grouped together based on their availability and time zones. To maintain the dynamic of the focus group, participants were grouped in minimum groups of three. Additionally, the discussions from previous rounds were presented in subsequent rounds, with the aim of integrating comments and discussions across rounds. Recent research has shown success in conducting online focus groups in rounds, using whiteboard technology to track changes as an investigation of the social interaction restrictions due to the post-COVID-19 scenario (Menary et al. 2021).

The profile of participants is presented in Table 6.1.

Table 6.1: Focus Group Participants Profile

Participant.ID	Group	Background	Participated..in.Interviews	Years.of.Experience
TT	2	Architect and Urban Designer with experience in Urban Analytics, and Computational Design	Yes	10
FM	2	Architect and Urban Designer with experience in Urban Developments on Hotels and Resorts, and Computational Design	Yes	10
GP	2	Architect and Urban Designer with experience in BIM, and Computational Design	Yes	5
AC	2	Architect and Urban Designer with experience in Environmental Design projects, and Computational Design	Yes	15
JM	2	Architect and Urban Designer with experience in Master Plans and Data-Driven Urban Design	Yes	10
AD	2	Architect and Urban Designer with experience in BIM and Management	No	10
RT	1	Urban designer and Planner	Yes	10
LG	1	Architect, and Urban Designer	Yes	20
EN	1	Architect, and Urban Designer	No	10
AB	1	Urban designer and Planner	No	10
AS	1	Architect, and Urban Designer	No	10
ZY	1	Urban designer and Planner	No	10

6.2.1 Demonstration

To validate the usability of the approach during the focus groups, a demonstration using a real-world scenario was conducted using data collected from publicly available governmental sources and open-source services. Scenarios are limited descriptions of how a system and its environment may behave in specific situations (Benner et al. 1993). They help researchers identify potential challenges, risks, or opportunities that may arise in a particular situation and develop strategies or solutions to address these challenges. The scenario chosen was Port Adelaide, an historically complex area in Adelaide that is striving to become a “living port” through redevelopment and new urban characterisations (RenewalSA 2014). This scenario provides a suitable scale across macro and micro urban design dimensions and characteristics that can be explored using the developed approach.

The scenario for the demonstration has been described in Chapter 5. The scenario was pre-recorded and can be viewed on YouTube²⁹.

6.2.2 Questions

The questions presented in Table 6.2 were formulated based on the three approach development strategies (framework and computational toolkit prototype) described in Chapter 5 and the potential for the developed approach to be adopted by designers. The questions aim to ensure that the validation process covers the key factors driving the development of the proposed approach and to explore the potential for adoption. Table 6.2 shows the relationship between the approach to development strategies, approach to adoption, and questions.

Table 6.2: Focus Group Questions

Validation.Focus	Id	Questions
Systematisation of Data-Driven Urban Design Processes	Q1	In your opinion, how does the proposed approach (framework and computational toolkit) assist urban design processes?
Integrating Datasets, Design and Data Tools	Q2	In your opinion, how do the chosen technologies and tools support urban design processes?
	Q3	Are there other technologies and tools that would be beneficial to integrate into this approach?
	Q4	In your opinion, how does the integration of design and data tools enhance the workflows of urban design processes?
	Q5	In your opinion, how can the chosen visual programming environment be seamlessly integrated into your design process?
An Holistic Toolmaking Development of a Computational Toolkit	Q6	This approach was developed under an holistic socio-political understanding of Free and Open-Source Software (FOSS). Would you be more willing to create, contribute, and collaborate in the development of Free and Open-Source Data-Driven Computational Tools if more free software and open data were more widely available in the AEC community?
The Adoption Potential of the Approach	Q7	In your opinion, what are the advantages of and barriers to using the proposed approach?
	Q8	Would you be more willing to use data-driven computational tools to assist your process with this approach? (Answered only by traditional designers)

The questions are designed to validate the overall usability and effectiveness of the data-driven approach, and review the appropriateness of the chosen technologies, noting the presence of any missing essential features, and any potential barriers to adoption.

• Q1 aims to validate the overall coherence and systematic nature of the data-driven approach in assisting computational urban design processes. This question is aligned with the main research aim “to develop, demonstrate, and validate an integrated data-driven approach to assisting computational urban design processes” and Objective 5, as outlined in Chapter 1, and the first gap discussed in the literature review (2).

• Q2 seeks to validate the suitability of the chosen technologies in supporting data-driven urban design processes and their integration into the overall approach. This question is linked to Objectives 2 and 3 of the research.

• Q3 addresses the potential for future development by identifying any missing features in the proposed approach.

• Q4 focuses on validating the integration of design and data tools in the proposed environment, which was identified as an essential feature during the interviews (Chapter 4).

• Q5 assesses the validity of the visual programming environment interface as the medium of integration and identifies any potential barriers.

• Q6 examines the relevance of considering a socio-political view of the holistic toolmaking process among designers, which was identified as the second gap in the literature review (2). It also gauges the willingness of designers to contribute to this vision.

• Q7 as part of Design Science Research (DSR), it is important to address practical problems. Therefore, this question seeks to identify the advantages and barriers of the proposed approach that could impact its adoption in the field and assess its feasibility.

• Q8 similarly, this question investigates the potential for the proposed approach to be adopted in practice by traditional designers.

6.2.3 Findings

This subsection presents the focus group findings. These findings are presented following the validation focus and questions order, presented in Table 6.2.

6.2.3.1 Validation Focus I: Systematisation of Data-Driven Urban Design Processes

Q1, presented in Table 6.2, has the purpose of validating the overall approach as a cohesive and systematic pathway for assisting urban design processes. Regarding this question, both traditional (Group 1) and data-driven designer (Group 2) participants agreed that the developed approach (framework and computational toolkit) assists urban design processes in a structured manner, as it encourages systematic decision-making during the urban design process.

“Yes. It provides some very helpful tools for urban designers to use. In my opinion, from what I saw, the approach allows professionals to be more conscious in their decision making in a structured manner, expanding their perception of the urban context by making meta data visual. It also, by using a low code platform as its basis, makes data analysis easier for non-developers.” (TT - Group 2)

“The toolkit is designed very good, in a comprehensive and structured way. It has ability to analyse big data about urban design in an integrative way.” (RS - Group 1)

The comprehensive set of tools and technologies, structured in different categories, as well as the data-driven procedures provided in the developed toolkit, were pointed to as important features to assist with data-driven computational urban design processes. Besides that, participants highlighted the tools categories, which are based on the data-driven procedures, as an important feature to assist computational urban design processes.

“Yes, I would say so. Looks like a very extensive toolkit that in Grasshopper you would need multiple plugins and yet still wouldn’t be able to cover everything.” (JM - Group 2)

“Yes. It presents some potential tools to read, analyse and manipulate data. It accepts different formats and files, so it guarantees a high level of interoperability.” (GP - Group 2)

“It does help with the urban design process, especially during the data visualisation process. The plug-in also makes data processing and retrieving much easier. I particularly like the option of retrieving multiple images around the perimeter. It helps with visualising and understanding the surrounding area and I think that is important for future urban development. (EN - Group 1)

The overall approach was received positively by both groups. According to the answers, the categorisation of the tools regarding its purpose and steps provides structure and guidance for the designers. Also, the comprehensiveness of the toolkit in integrating different technologies, tools, and datasets contributed to the high acceptance of the approach among both groups.

6.2.3.2 Validation Focus II: Integrating Datasets, Design and Data Tools

Q2, presented in Table 6.2, has the purpose of investigating how the developed tools and their different technologies assist the urban design processes. Regarding Q2, the three elements of integration (datasets, design and data tools) that drive the development of the toolkit were mentioned by the data-driven design participants as features that the tool provides that assist with the processes in data-driven urban design.

“I see the most potential on the developed [system] is that it allows users, at the same time,[to] work with GIS files and analyse datasets. The correlation between those two could allow a data driven urban design.” (TT - Group 2)

“It integrates different processes in one unique workflow. In this way, it’s easier and faster to manipulate and generate data-driven outputs.” (GP - Group 2)

“The chosen technologies and tools support urban design processes by providing data in an integrated and accessible manner.” (FM - Group 2)

The cohesive set of tools and their integration into visual programming contributes to enhance the workflows of data-driven design. The agility and speed during the integration highlights that a diverse and correct set of tools, when well-integrated, helps to better inform decisions in different use cases in urban design. For instance, participants highlighted that visual programming tools are often more beneficial for enhancing existing workflows than using bespoke coding solutions or Geographic Information Systems (GIS). Also, location analytics is a crucial aspect of every project, but it is regularly hindered by the lack of access to high-quality data, such as the basemap information obtained from OpenStreetMap.

“Having an open-source or well supported toolkit that does all this work would be vastly superior to our own tools. The value of these tools appears to be the wide range of use cases.” (JM - Group 2)

Besides the three integration elements, traditional designers (Group 1) highlighted aspects related to visualisation and analytics characteristics of the toolkit, such as the plot functions, used to plot the relationship between data and maps.

“I think the plugin’s strength is in its data visualisation and how easy it is to call a specific dataset. Especially with the data being open-source and non licensed. How easy it is to change and generate new information is also very helpful.” (EN - Group 1)

“In my opinion, it will help demographers to simply do some statistics on the population of the case study conducting correlations and scatterplots.” (AB - Group 1)

The participants’ answers agreed that the integration of three data-driven elements used to develop the approach (datasets, design and data tools) have an important role in data-driven processes. Also, the choice of a visual programming in this aspect tends to facilitate the integration as mentioned earlier.

Q3, presented in Table 6.2, has the purpose of exploring those potential technologies and tools missing in the proposed approach that could turn the approach more cohesively to assist data-driven urban design processes. Regarding Q3, the answers from both group of participants were plural, reinforcing the complex nature of urban design problems.

“Further development on more dataset manipulation features and possibly some other ML approaches. I think there are more practical uses that AI can offer urban designer rather than subjective image segmentation.” (TT - Group 2)

“More ecological aspects added to the plugin in order to properly tackle all the aspects initially pointed out in the framework. If not included, I wonder if the toolkit could serve as an input setting for broader CFD analysis in Blender, for example.” (FM - Group 2)

“If you can integrate the qualitative analysis tool or extensions like SPSS analysis, it would be very helpful to get comprehensive analysis and results.” (RT - Group 1)

This plurality of views carried the discussion towards the extensibility and flexibility of the approach, since these characteristics tend to be inherent features in open-source developments (Tuomi 2005), suggesting that this kind of feature facilitates the adoption of new technologies and tools though the approach. Furthermore, it was highlighted that the potential of the approach of integrating with third-party plugins (which is a feature by design), such as Ladybug Tools for Blender for realising environmental analysis, provides a crucial capability of expanding the approach beyond the functionalities its original functionalities.

Q4, presented in Table 6.2, has the purpose of investigating the perspective of designers in relation to the proposed integration of data and design tools and how this integration could enhance their workflows during data-driven urban design processes.

Regarding Q4, Group 2 highlighted that the integration of data and design tools can explore more informed decisions and potentially enhance their creativity during the process. Also, Group 2 mentioned that this integration supports clear and confident communication of decisions with client and other stakeholders involved in the process.

“It brings to light less obvious correlations that could affect the urban space, by manipulation, merging, splitting and filtering different datasets, making it all visual and with geo-referenced locations. With these pieces of information in hand, designers can be more creative in their projects.” (TT - Group 2)

“My view is that designing with data leads to more informed and accurate decisions. Removing the complexity of some thought processes and being confident that the decisions we make as designers are going to make a difference. (JM - Group 2)

In Group 1, the integration of data and design tools was also understood as beneficial to the design processes, enhancing the ability to create urban narratives (telling stories) and the potential of the integration of the data and analysis in the same design environment.

“Tell[ing] stories is one of the most important impactful ways of communication and the integration of design and data tools will definitely enhance this.” (AB - Group 1)

“As the tool helps to analyse integrated data in various aspects in a unique environment, it would speed up the urban design processes.”(RT - Group 1)

Q5, presented in Table 6.2, has the purpose of understanding how Sverchok, which is the chosen visual programming environment and is not the most adopted environment among data-driven designers, as discussed in Chapters 2 and 4, could have an impact on the workflow of data-driven designers, as well as trying to explore from the traditional designers’ point of view how the use of visual programming could assist computational urban design processes.

In answering Q5, both Groups answered positively about the use of the proposed visual programming environment, Group 1 as new users of visual programming and Group 2 as new users of this visual programming environment. The discussion around this question highlights a few concerns about training in computational thinking in the field of urban design and planning.

“As an urban researcher, I am not very much familiar with the visual programming environment. But I think more urban planners and designers should familiarise themselves with these kind of programming environments because by ever-growing developments in the ICT, the line between data science and urban planning is getting blurred.” (AB - Group 1)

According to the answers to Q5, the adoption of a new visual programming environment (Sverchok) by data-driven designers, tends to not be a barrier for data-driven designers used to Rhino-Grasshopper, since these environments are similar. However, visual programming, in general, has a steep learning curve that challenges the learner through new ways of thinking about and conducting the design process. Consequently, this shift of thinking combined with learning new computational tools creates a barrier for traditional designers in adopting this type of approach.

According to the answers to Q2, Q3, Q4, and Q5, the integration of datasets, design and data tools in a visual programming environment has a positive impact in assisting data-driven urban design processes, corroborating the feasibility and usability of this strategy.

6.2.3.3 Validation Focus III: Holistic Toolmaking Development of a Computational Toolkit

Q6, presented in Table 6.2, has the purpose of understanding if the availability of Free and Open-Source (FOSS) pieces of software in the industries of Architecture, Engineering, and Construction (AEC), influences designers’ participation on FOSS projects, given that community-based development is an essential feature to drive the directions of technological development that will shape our future cities, as discussed in Chapter 2.

In response to Q6, most participants in both groups would be more willing to participate in free and open-source projects if more become available. Among the answers, the importance of open data and the value of communities sharing advancements in tools helps to direct tool development in favour of the common good.
A divergent answer acknowledges the benefits of FOSS but believes that the tool has commercialisation potential as a proprietary piece of software.

“Yes. Access to data is important, not just for research but for the community as well. I think that free information is the key for a better future in research.” (EN - Group 1)

“Definitely. One of the biggest difficulties that I face…about data-driven design is the lack of public data. If we have more accessible and open data, the results obtained will be even more accurate. The Open-Source Community makes a software develop quick. Bugs are fixed faster, and new versions are constantly bringing optimisations for the tools. Besides that, it creates a huge group of support and sharing, that sometimes doesn’t occur in pieces of software developed by private companies.” (GP - Group 2)

“Yes. Open-source is often seen as a negative thing in our field; however, it’s never considered in the bigger picture. The problem is often seen as intellectual property and who owns what. If an office creates a tool, then it is theirs and why should they share it. Whilst that is a valid concern, the tool itself is not designing, it is an aid. The actual intellectual property is what the tool helps create. In that light, it makes sense to me that unless a tool can be turned into a profitable venture and it is not abstracted from its intended purpose, then it should be free and open-source. We can take cues from the software development industry on this. Facebook, Google, etc. have created tools and invested in others that are open-source.

According to the responses to Q6, the development of Free and Open-Source pieces of software in the AEC industry has the potential to grow if more tools are developed as part of an open-source ecosystem, where data and tools are openly shared.

6.2.3.4 Validation Focus IV: The Adoption Potential of the Approach

Q7, presented in Table 6.2, has the purpose of exploring potential advantages and barriers to the proposed approach that could influence their adoption in assisting urban design processes.
The advantages and barriers discussed in Q7 with both groups are presented in Table 6.3.

Table 6.3: Advantages and Barriers in the approach adoption

Advantages	Barriers
Integration of datasets, data and design tools	Visual programming learning curve
Data analysis in visual programming	Low information and adoption of Blender-Sverchok as visual programming environment
Toolkit comprehensiveness and cohesiveness
Accessibility of urban analysis for everyone

The advantages and barriers mentioned Q7 are recurrent in other answers, reinforcing those aspects as crucial elements that can influences toolkit adoption. For instance, among the advantages and barriers, integration of datasets, data and design tools and the visual programming learning curve, respectively, were pointed out several times during the focus groups.

Q8, presented in Table 6.2, was answered only by traditional designers, and its purpose is understanding if, among traditional designers, the proposed approach would have an influence on the potential adoption of data-driven tools during their design workflows.

Answering Q8, participants from Group 1 were positive regarding the adoption of computational tools through the proposed approach. They mentioned how computational tools could help them to accelerate some processes as well as their potential in making the design process more participatory. During the discussion regarding these questions, few concerns about the digital literacy of traditional urban designers and the learning curve of visual programming tools were pointed out as potential barriers to approach adoption.

“Yes, it is very helpful to analyse the related data in data-driven computational tools, as it is very fast and quick.” (RT - Group 1)

“Yes! I am working on developing a Public Participation GIS (PPGIS). I have been using the Esri 123Survey platform, which has several deficiencies. Developing more and more open-source map-based survey tools will make participatory and communicative planning processes easier.” (AB - Group 1)

“Yes, especially with the amount of data in urban design could be: computational tools can speed up the process and leave more room for discussion and development.” (EN - Group 1)

Overall, the participants declared that the approach has potential for assisting data-driven urban design processes, highlighting the three main validation focal points defined based on the three main development strategies used to develop the proposed approach.

6.3 Implications of the Validation Findings

Overall, both groups agreed that the comprehensive set of tools developed in the computational toolkit assists data-driven urban design processes. Also, the visual programming environment was discussed as a crucial environment for the integration of data and design procedures in an interoperable design environment. Furthermore, both groups agreed about the importance of open data and open-source software initiatives in promoting the democratisation of technological development in the AEC industry to drive future cities visions. Even though both groups have similar visions about the subjects discussed above, their perspectives tended to focus on slightly different features of the computational toolkit. While Group 1 highlighted how the developed tools for visualisation could facilitate urban design narratives based on the data presented, Group 2 highlighted the integration of different technologies in the same environment.

Based on the feedback received during three rounds of focus group, the implications of the validation findings are highlighted bellow, following the structure of the validation focus.

6.3.1 Systematisation of Data-Driven Urban Design Processes

One of the most highlighted aspects of the overall approach, regarding its potential in assisting data-driven urban design processes, was the comprehensive set of tools that are systematically structured and integrated in the same visual programming environment. This systematisation of data-driven urban design processes has been explored as a demand in the literature review (Chapter 2), discussed during the interviews (Chapter 4), and appreciated as a feature during the focus groups.

Even though data-driven urban design is increasingly adopted in the industry with the availability of smart infrastructure, computational design tools, and the increasing accessibility of cities’ data through government portals and open-source services, there is a lack of systematisation about the steps that move raw data into design actions that can effectively inform evidence-based decisions in design (Kvan 2020).

The semi-structured interviews presented in Chapter 4 reveal that there is a diversity of approaches and tendencies among data-driven urban designers; however, a systematic structure can also traced between these approaches regarding design processes, data driven steps, and computational technologies, as per the proposed framework.

The overall approach (framework and computational toolkit) received positive feedback during Q1 regarding its structure and comprehensiveness. The integrated approach provides an holistic structure and tools to assist data-driven urban design processes. Both groups agreed that the structure of the proposed approach helps during decision-making for data-driven urban design processes. Therefore, the systematisation of the data-driven urban design process proposed as Objective 2 in Chapter 1 was successfully validated regarding its usability and feasibility during the face-validity process.

6.3.2 Integration of Datasets, Design and Data Tools

A core feature from the proposed toolkit is the integration of datasets, data and design tools. During the data-driven urban design process, designers operate a diverse range of datasets across different pieces of software, considering multiple dimensions of urban design (Carmona 2002, 2021; Boyko, Cooper, and Davey 2005; Batty 2013b). This process is carried out to transform raw data into information, information into insights, and insights into design actions (Deutsch 2015; Mathers 2019). An integrated workflow is crucial to support data procedures through the steps of the design process. This need for data diversity and its integration is highlighted in Chapter 4.

The integration of datasets, design and data tools implemented in the proposed approach was discussed from different perspectives across Q2, Q3, Q4, and Q5 from Table 6.2. The practical procedures of how the integration works in the approach are explored in Q2. During the discussions about design-data procedures, participants highlighted as a crucial feature the integration of multiple datasets through design processes, data-driven cycles, and computational tools. The comprehensiveness and cohesiveness mentioned during Q1 are also highlighted in Q2, explaining how the data-driven design process is enhanced due to the proposed data-design integration into a single visual programming environment.

Even though the approach presents a comprehensive set of tools and technologies, urban design problems are complex problem-solving tasks, involving consideration of multiple urban design dimensions during the process. The different possible combinations of a set of tools are limited to address a finite number of urban design problems. Therefore, the categorisation of data and design tools is designed to be flexible and modular, accepting new categories of tools and technologies that can potentially emerge from the urban design community’s demands in assisting the uncovering of topics in a set of urban design problems. Q3 is discussed in this context, trying to understand and define the missing features that the approach can explore to enhance its comprehensiveness and cohesiveness. The answers from participants were pluralistic about this subject. During the discussions, the suggestions acknowledged future potential implementation, directing the discussion towards the potential extensibility of the approach in absorbing those new implementations propositions as well as potential new ones that could emerge from the demands of a data-driven community through community-driven open-source development.

Furthermore, integration with third-party plugins, powered by the visual programming environment, was also discussed as a way of expanding the approach’s capabilities beyond its current limitations.

Creativity-enhanced, clear communication among stakeholders, and history-telling based on pieces of evidence were highlighted in Q4 as features of the integration of design and data tools empowered by the visual programming environment. In addition, the feasibility of the chosen visual programming environment was accessed in Q5, revealing minor concerns from Group 2, regarding the need to learn a new visual programming interface. However, even though Group 1 acknowledged the benefits of the visual programming environment, its steep learning curve could represent a barrier to approach adoption in this group. Consequently, the discussion evolved towards the important role of the education system in implementing into architecture and urban design courses a mandatory minimum of credits in the curriculum about computational thinking, providing a foundation for computational thinking within these fields, which could soon contribute to building a common “computational language” that all designers can speak and understand. Moreover, the role of multidisciplinary teams was discussed as an important feature in practitioners’ offices to diminish the communication gap among stakeholders.

Overall, the integration of datasets, design and data tools across Q2, Q3, Q4, and Q5 was positively accessed by participants of both groups, showing that the approach provides a feasible level of integration across datasets, design, and data tools, validating its feasibility and utility among designers, which successfully achieves Objectives 3 and 5 of this research, presented in Chapter 1, respectively: Objective 3) developing a computational toolkit prototype to assist computational urban design processes; and Objective 5) Validating the integrated data-driven urban design approach among designers.

6.3.3 Holistic Toolmaking Development in a Computational Toolkit

To comprehend the level of awareness and initiative of data-driven designers in participating in free and open-source AEC developments projects, Q6 was developed. Toolmaking, as described by Ceccato (1999) and Burry (2011), has increasingly become common practice among architects and urban designers in the computational design field, given the emergence of end-user programming empowered by visual programming tools embedded in 3D modelling pieces of software, as discussed in Chapter 2. As noted by Frazer (1995), proprietary design software packages “never quite seem to do what one wants in the way that one wants.” (Frazer 1995.p25). Even though visual programming moves designers towards the paradigm of digital toolmakers, there are limitations and restrictions related to lack of freedom of users and developers in proprietary software that will not be overcome in the AEC industry while the majority of data-driven designers rely on this type of solution alone, as discussed in Chapter 4. Furthermore, it is crucial that the tools that will shape future smart cities should be open and driven by the interests of the community, since “the city” is a product of its citizens, and designing for the city is a collective product that mutates and evolves, driven by the collective realisation of a crowd (Ratti and Claudel 2016). From a smart city’s perspective, an holistic toolmaking process should consider socio-political views of Free Software and the “right to the city” (Lefebvre 2008), reclaiming the city as a co-created space, and diminishing the gap between the human-centric approach and technological-driven approach (Kummitha and Crutzen 2017), thereby shaping future smart cities through the co-creation of “smart people” (Ameijde 2022).

In answering Q6, the majority of the participants demonstrated that they are willing to participate in free and open-source projects if more initiatives start to become available. Some benefits of community-driven developments were highlighted, such as faster updating and bug fixing, the value added to a tool from a community, and the benefit of having the public participating in driving developments. Also, the importance of open datasets was discussed as a crucial element to enable data-driven urban design processes.

Overall, participants recognise the value of free and open-source software and open data in democratising access to the tools for anyone willing to contribute. The discussion shows that as soon as more development of free and open-source data-driven tools starts to become available, the tooling process will tend become more collaborative, enhancing citizens’ sense of participation as to how the technologies are shaped, including a sense of by whom and for whom, through feedback loops of the synergy of city, citizens, and technologies towards the shaping and reshaping of future cities. Therefore, this validates the approach conceptually among designers, in terms of the concepts underlying Objectives 3 and 5.

6.3.4 The Adoption Potential of the Approach

An important validation step is understanding the potential that the approach has to be adopted by designers. Two questions address this step: Q7 and Q8. Q7 explores, from a designers’ perspective, the advantages and barriers to the proposed approach.

In response to Q7, the synthesis of the discussion among participants highlighted four main advantages, and two disadvantages, as presented in Table 6.3.

Advantages

The first advantage, “Integration of datasets, data, and design tool” is one of the strategies of development of the proposed approach and it is discussed in detail in Subsection 6.3.2 across Q2, Q3, Q4, and Q5 and this feature of the proposed approach was reinforced as an advantage in Q7, demonstrating its importance for the overall approach.

The second advantage, “Data analysis in visual programming” also highlights the potential for integration of data and design tools in a single design environment to produce, in the case discussed, analytical insights during the design process. This was discussed across questions Q2, Q4, and Q5. A facilitated integrated design environment has the potential to better support the data-cycle process from raw data to design actions, as well as providing a faster response during the evidence-based decision-making process.

The third advantage, “Toolkit comprehensiveness and cohesiveness” highlights its structure as an overall approach, but also the set of tools and technologies available to assist data-driven urban design processes. Besides its citation in Q7 as an advantage, this feature was discussed across Q1 and Q2.

The fourth advantage, “Accessibility of urban analysis for everyone” highlights the accessibility of the implementation of the approach through the visual programming language, which is simpler than running analysis in textual code. This feature was discussed during Q7.

Barriers

The first disadvantage, “Visual programming learning curve” highlights the designers’ concerns regarding the steep learning curve of the visual programming environment. Visual programming languages do not require previous experience with code, but it was discussed that the reason that the learning curve is accentuated is because it requires a different way of thinking about the process, computational thinking, that designers are not used to, in which the problem-solution formulation should be broken into small steps and be carefully made explicit. Beyond this specific problem of the proposed approach, this concern was discussed during Q5 a structural educational concern that should be addressed in the architecture and urban design curriculum.

The second disadvantage, “Low information and adoption of Blender-Sverchok as a visual programming environment” highlights that the chosen visual programming environment is not the most utilised in the industry and there is little information about its utilisation. During the discussion, this concern was presented both as a threat and an opportunity, since on the one hand learning a new tool is time consuming and demands extra effort of the user to reach a level of excellence that can be compared with the preferred tool of first choice;whilst, on the other hand, the low information and adoption of the tool shows that there is still a lot to be produced regarding information for familiarising users with the environment and there are important points to be discussed about data-driven designers’ tool choice adoption and the role of toolmaking in a city context, as discussed in Q6.

The Q8 is addressed to traditional designers (Group 1) to understand if the approach has impacted their perceptions in relation to the use of data-driven computational tools during the design process and if they will be more willing to attempt this through the proposed approach. Most participants demonstrated that they felt incentivised by the presented approach. However, as discussed during Q7, concerns about the learning curve for visual programming tools were also highlighted.

Overall, the proposed approach was received positively by both groups of participants through the three points of validation focus. Besides the positive evaluation, limitations about the visual programming environment were discussed, highlighting that there are structural concerns about designers’ educational curricula providing insufficient computational thinking education. Seeking to flatten this learning curve is beyond the scope of this research to address.

6.4 Conclusion

This chapter presented the approach validation process carried out through three rounds of focus groups. This validation was conducted to ensure the quality of the proposed approach, certifying its feasibility and usefulness among designers through a face validity process. Participants answered eight questions that were directly related to the five objectives of this research (presented in Chapter 1), and the three strategies used for approach development (presented in Chapter 4). The aim of this research, presented in Chapter 1, was achieved in this chapter, since the three validation foci and the potential of the approach to be adopted were positively validated by the designers during the focus group rounds.

• Validation Focus: The systematisation of Data-Driven Urban Design Processes. The overall approach was highlighted as structured and comprehensive, helping decision-making during data-driven urban design processes.

• Validation Focus: Integrating Datasets, Design, and Data Tools. The developed integration of datasets, design, and data tools was highlighted due to the visual programming environment being the core component of the integration, which facilitates workflow and interoperability among distinct datasets. Also, it was highlighted that the integration of the proposed approach enhances creativity during the data-driven process and provides features such as clear communication among stakeholders and history-telling creation based on evidence.

• Validation Focus: An Holistic Toolmaking Development of a Computational Toolkit. The conceptualisation of an holistic digital toolmaking development was validated among participants, since they demonstrated that they are aware of the importance of a collaborative process that enhances the sense of participation in citizens regarding the tools that are produced to shape future cities.

Overall, potential adoption of the approach was well received by both groups, who were willing to use the proposed approach. Beyond the positive evaluation, the discussion highlighted that there are several concerns regarding the educational curriculum that are beyond the scope of this research, but which can be addressed in future research. Also, potential improvements regarding the computational toolkit were pointed out during the discussions, such as further development of tools in dataset manipulation and processing, implementing new machine learning approaches beyond the current features of image classification and regression models, inclusion of environmental tools in the toolkit, and integrating qualitative data analysis tools in the approach. These suggestions will be taken into consideration progressively in future releases of the computational toolkit, since the developed computational toolkit is currently an ongoing open-source project, which will continued to be developed beyond the conclusion of this PhD.

The next chapter presents a discussion about the research findings, a response to the research gaps, and the consideration of implications of the presented research.

---

29. The demonstration is available in YouTube through this link: https://youtu.be/lTRNIa2PwhQ↩︎