Planning and Design Implementation
Investigating Smart City Barriers: Contribution of Experts based on a Delphi Analysis
2022 Volume 10 Issue 2 Pages 179-199
Details
2022 Volume 10 Issue 2 Pages 179-199
The lack of cooperation between stakeholders and coordination of departments, the absence of systems interoperability, and the resistance to change by policymakers represent some of the challenges that make the implementation of the Smart City a complicated task. Nevertheless, the literature is lacking a broader comprehension of the Smart City barriers to enable policymakers to design and execute their strategies. Therefore, this paper aims to find the (critical) barriers that have been associated with having a significant influence on the success of a Smart City’s implementation and that are within the control of policymakers, based on the empirical knowledge and experience of experts. A study of the temporal aspect in which they can be overcome is also performed to allow better planning. Moreover, 50 barriers were aggregated and organized in eight distinct areas from a thematic analysis of an initial list of 114 barriers collected from the literature. Nine experts participated in the Delphi Analysis, which demonstrated that although Smart Cities are usually associated with technological and financial constraints, the 15 most critical barriers are mainly from the areas of Governance, Project, and Organization. The method combined a two-round survey with a focus group, using quantitative and qualitative approaches.
In the 1990s, the Smart City concept was for the very first time associated with information and communication technologies (ICTs), forecasting that technology would be at the centre of city management (Aurigi, 2006; van Bastelaer, 1998; Gibson, Kozmetsky, & Smilor, 1992; Graham & Aurigi, 1997; Mahizhnan, 1999; Tan, 1999). Later, it was linked to a new paradigm of solving the problems caused by urbanization and globalization (Angelidou, 2015; Chourabi et al., 2012). The initial techno-centric approach led by technological companies, where ICTs were seen as the solution to every problem - Smart City 1.0 - evolved after the criticism that the focus should not be on the technical aspect (Hollands, 2008). Moreover, the concept started to thrive from a human and social capital perspective (Caragliu, Del Bo, & Nijkamp, 2011). The technology was then seen as the means to meet citizens' needs, improving the quality of life and the city’s sustainability and not as an end in itself - Smart City 2.0 (Lin, Shen, & Teng, 2021; Mirzaei & Zangiabadi, 2021). Recently, the citizens' role moved towards being contributors to the city strategy - Smart City 3.0 -, empowering them to be part of the design and thinking process (Cohen, 2015; Gutiérrez et al., 2016).
Nevertheless, cities’ strategic planning depends on their specific contexts, problems, and opportunities (Correia, Marques, & Teixeira, 2021; Tödtling & Trippl, 2005). Moreover, to design an effective action plan, it is necessary to understand what is fundamental to implement and the timing to carry out the implementation process. That reflection is ultimately connected to the prioritization and schedule of the actions to tackle specific challenges based on long and short-term circumstances. Moreover, the consideration of existing barriers is crucial to planning a successful Smart City strategy. The literature is scarce about the meaning of what a Smart City barrier is. Generically, authors have associated it to the obstacles and difficulties hindering and delaying a successful Smart City strategy and related projects (Chourabi et al., 2012; Mosannenzadeh, Di Nucci, & Vettorato, 2017; Rana et al., 2019; Tan & Taeihagh, 2020). Nevertheless, a Smart City barrier can be defined as a challenge posed to urban development that if not overcome by policymakers (or relevant stakeholders), means a partial or total impediment to the implementation of a Smart City strategy.
The struggles that some face to implement a Smart City approach may be different from that of others. There are countries where the topic has still not brought meaningful change and impacted urban policy practices (Varró & Szalai, 2021). Some authors point out that the implementation of Smart Cities has not been possible due to the unsatisfactory level of prosperity of the residents, the difficult financial condition of cities, and unfavourable demographic trends, whereas for others to the cultural challenge for community participation (Das, 2017; Jonek-Kowalska & Wolniak, 2021). The heterogeneity within the country between cities is most times associated with the capacity of larger cities to access financing programs (Correia, Teixeira, & Marques, 2021; Smékalová & Kučera, 2020). Thus, scaling pilot projects is widely perceived as a major concern (van Winden & van den Buuse, 2017).
Furthermore, barriers do not have the same impact and priority (Mosannenzadeh et al., 2017; Rana et al., 2019). Thus, it is vital to study their spatiotemporal relevance to understand how can they be surpassed to create an urban environment that combats societal segregation and marginalization, and promotes the inclusion and quality of life of inhabitants (Correia & Feio, 2020; van Gils & Bailey, 2021; Medved, Kim, & Ursic, 2020; Wolf et al., 2019).
In the literature, several studies that have performed narrative and systematic literature reviews on the topic can be found. Chourabi et al. (2012) aggregated and identified the challenges of Smart Cities. From a raw dataset of 212 barriers of empirical research of 43 communities implementing energy projects, Mosannenzadeh et al., (2017) ranked 35 barriers in 9 categories based on a quantitative approach. Rana et al. (2019) evaluated the barriers of Smart City development in the Indian context, categorizing 31 barriers in six key areas. Tan and Taeihagh (2020) listed the barriers associated with the development of Smart Cities in developing countries from a systematic literature review. Additionally, other authors have also pointed out some of the Smart City barriers (Alawadhi et al., 2012; Chatterjee & Kar, 2015; Correia, Teixeira, & Marques, 2020; Ebrahim & Irani, 2005; Gil-Garcia, Pardo, & Nam, 2015; Goyal, Sahni, & Garg, 2018; Mosannenzadeh & Vettorato, 2014; Nam & Pardo, 2011; Neirotti et al., 2014; Tan & Taeihagh, 2020). However, the consideration of other variables beyond the importance level to represent a significant contribution for policymakers is lacking, such as the impact a specific barrier may have to oppose the development of a Smart City, the capacity of decision-makers to overcome it, and the timespan in which they can perform it. Starting from the universality that all barriers are equal and have the same weight, it is vital to perform a more profound study to understand their specific attributes.
Therefore, this paper aims to find the barriers that have a critical influence on the success of Smart City implementation, and which policymakers can have a decisive role in overcoming them. Moreover, the research question of this paper is: “What are the critical barriers to the implementation of a Smart City strategy?” For a reasoned analysis, this work considers evidence through literature and empirical experiences (Painuly, 2001). The methodology combines quantitative (questionnaires to assign the weights) and qualitative approaches (discussion over a focus group). It gets the opinion and discussion of Portuguese Smart City experts by performing a Delphi Analysis combining a two-round survey with a focus group. This method promotes joint learning and convergence of opinions. The final answer of each participant is made based on the confrontation and perception of contrary points of view.
The methodology is described in the next section and is divided into three main steps: (A) identification and organization, (B) evaluation, and (C) interpretation. The results are demonstrated and discussed in section 4. At the end, conclusions are highlighted about the most critical barriers for Smart City implementation.
To find the most critical barriers to implementing a Smart City strategy, the methodology was divided into three phases: (A) identification and organization – investigation and selection from scientific works of the initial barriers and performance of thematic analysis to organize within the areas of scope, (B) evaluation – identification of the experts and development of a two-round Delphi survey to get the feedback of external stakeholders, concerning the defined evaluation criteria, to each barrier, and (C) interpretation – analysis and graphical representation of the final results and translation of their practical meaning. This way it was possible to design a systematic and participatory process combined with moments of discussion and confrontation of perspectives mobilizing the experts to a possible convergence in a common vision. Figure 1 summarizes the methodology followed in this study divided into the previously mentioned three phases.
Identification and OrganizationFor the initial collection of the barriers, a search was conducted on Scopus using the following keywords: “smart cit*” and “challenge” and (“barrier” or “pitfall”). From the 370 results, 200 abstracts were read, and from those, nearly 60 deserved closer attention and deeper analysis. The initial barriers of the present work were collected from narrative and systematic literature reviews as previously mentioned in the Introduction section. The goal was to find papers that had already performed studies to find the Smart City barriers.
Furthermore, from the careful analysis of the barriers, a qualitative approach to finding patterns within and across the data to find the final barriers list was performed. In this case, inductive thematic analysis was carried out where codes and themes were developed and associated by the authors (Rice & Ezzy, 1999).
Figure 1. Methodology
Figure 2. Smart City barrier shortlist
Moreover, similar barriers were aggregated into the same codes, and the redundant eliminated. Thus, on the one hand, those that did not contain a self-explanatory understanding, or the focus was subjective or discrepant were excluded. On the other hand, those whose understanding could be included in a more generic meaning without losing information were aggregated. The final shortlist constituted 50 barriers distributed across eight areas, namely Governance, Project, Organization, Economy, Socio-Cultural, Legal, Technology, and Environment, and is presented in Figure 2.
EvaluationAfter identifying the barriers and performing thematic analysis, a Delphi Analysis approach was followed. This topic’s complexity justified the reflection and discussion of experts with different backgrounds and experiences. Moreover, the Delphi Analysis allows each participant to give their opinion while looking for the convergence of the group throughout the iterative discussion rounds. In this case, this method was chosen because the overall goal was to provide policymakers with a conscious and agreed vision of Smart City critical barriers based on the perspectives of a heterogeneous group of experts.
The Delphi Analysis summarizes the opinion on emerging concepts, and on the development studies that lack empirical data and intends to objectively obtain experts' responses (Gordon & Pease, 2006). This method usually considers several and consecutive response rounds (Rodríguez-Mañas et al., 2013). In each round, the answers are collected. The disagreements are analysed and highlighted to ground the following round in order to converge the group's answers (Marques et al., 2008).
Given the subjectivity of the topic, a set of experts from various areas and with different years of experience was put together. This group was intended to represent the commonly interested parties on Smart City strategies. Therefore, the group included elements from academia (universities), private companies, and municipalities.
The experts' identification was based on the search of recognized elements based on iterative comparison of people's opinions in the field. On the one hand, participants were sought from research, planning, regulation, development of solutions, and decision-making roles (planning) and, on the other hand, from urbanism, mobility, waste, and environment sectors (applications).
Moreover, the group of experts was composed of nine elements (Table 1). The group had professionals with a range of 3 years to 30 years of experience in the area. The average element’s experience was approximately 12 years.
Table 1. Expert Identification
| Identification | Gender | Sector | Entity | Role | Experience |
|---|---|---|---|---|---|
| Element 1 | Male | ICT | University | Professor | 25 years |
| Element 2 | Male | ICT & Electronics | R&I Consultant | President | 10 years |
| Element 3 | Female | EC Policy & Strategy | Non-profit Foundation | R&I Department Director | 10 years |
| Element 4 | Male | Urban Management | Municipality | Technician | 5 years |
| Element 5 | Male | Mobility & Tourism | Private Company | CEO | 9 years |
| Element 6 | Female | Climate Action and Circular Economy | Municipality | Department Director | 6 years |
| Element 7 | Male | ICT | ICT Cluster | Executive Board Member | 30 years |
| Element 8 | Male | Environment and Waste Management | Municipality | Department Director | 8 years |
| Element 9 | Male | Urban Cleaning and Waste Management | Private Company | CEO | 3 years |
In terms of the evaluation of the barriers, three main criteria were established to assess them: (i) the impact of the barrier in Smart City strategy implementation; (ii) the capacity of the policymakers to overcome the barrier; and (iii) the extent of space-time in which the barrier can be overcome. Figure 3 presents the criteria and the associated classification spectrum. The digital survey of the first Delphi round was individually sent to all participants three days before the second round.
Figure 3. Criteria definition, classification, and ranking (Likert Scale)
First, the experts were asked to classify each of the 50 barriers identified in the literature review based on a Likert scale. After the first round of responses, the results were analysed.
According to each criterion (Impact, Endogeneity, and Space-time), a cluster analysis was performed on the participants’ responses to group the barriers whose evaluation matched the most.
The most discordant barrier spectrum of responses (higher standard deviation from the average response) was identified. This allowed uncovering the barriers, which had a more varied range of responses from the experts.
After, a focus group was held where the discussion between the experts was promoted. The focus group was set up to discuss the divergent barriers to ascertain whether the respondents maintained their initial response on the second Delphi round. The focus group followed the approach defended by Morgan (1998) and Stewart, Shamdasani, and Rook (2007), characterized by promoting an open and flexible discussion with a collective understanding uncovered by individual interviews, allowing the researcher's direct interaction with the experts. The second survey round followed this approach.
The exercise discipline was taken into significant consideration. The rigid fulfilling of the times and the assertive moderation promoted the objectivity of each intervention. The focus group lasted one hour divided through three consecutive moments, each divided into two distinct parts: discussion and survey. The chronological plan is detailed in Figure 4.
Figure 4. Focus group chronological plan
For each of the divergent barriers (with higher standard deviation), one element gave their opinion about why the response was in a particular direction to initiate the exercise. That triggered following participants' interventions with complimentary or contrary perspectives. After the time stipulated for each discussion moment, the elements had a tool (provided in advance) to place their new answer. Each element could see their first Delphi round answer and the average group’s response, as demonstrated in Figure 5. This way, each participant was aware of its position before the group and could re-evaluate the given answer on the first round from the discussion with the other focus group's elements. Thus, the subjective analysis was combined with the objective.
Figure 5. Second Round survey scheme tool
The results obtained in the two rounds of the questionnaire were analysed. Moreover, following the first round of the Delphi Analysis and the quantitative study of the barriers' standard deviations, a cluster analysis was also performed to study the group's relationships. These clusters clarify the experts' positions for a better and most focused discussion in the second round's focus group.
InterpretationTo contradict the critique made by Webb and Kevern (2001) about the lack of quotes and descriptions from the participants' interventions observed in the focus groups, the discussion's main points are described and transcribed in this subsection to highlight the contributions and the contradictions. Furthermore, the dominance of specific participants' opinions, the relationships among the experts, and the produced consensus is also analysed.
Following the qualitative description of each discussion moment, the verified changes from the experts' second-round answers concerning the first questionnaire’s responses are also presented.
Ultimately the results are graphically represented according to Figure 6. A general view of the path and the considered thresholds along the journey to find the most critical Smart City barriers is also presented.
Figure 6. Representation of the classification axes
The average results from the participants of the first Delphi round’s questionnaire are shown in Table 2 according to each criterion (C1 – Impact; C2 – Endogeneity; C3 – Space-time). Based on the results, the most discordant barriers were identified for the second round of the Delphi analysis. The less consensual barriers are represented in bold in Table 2.
Table 2. First Round survey average results
| Areas | Barriers | C1 | C2 | C3 |
|---|---|---|---|---|
| Governance | 1. Policymakers’ attitude | 4,22 | 3,89 | 2,44 |
| 2. Unclear vision / lack of strategy | 4,33 | 3,67 | 2,67 | |
| 3. Lack of Smart City-oriented politics and policy agenda | 3,67 | 3,78 | 3,33 | |
| 4. Lack of private-public partnerships (lack of stakeholders’ involvement) | 4,00 | 3,22 | 3,56 | |
| 5. Lack of understanding of the Smart City concept | 3,33 | 4,00 | 4,11 | |
| 6. Lack of long-term commitment | 4,11 | 4,22 | 3,22 | |
| 7. Low acceptance of new projects and technologies | 3,67 | 4,11 | 3,33 | |
| 8. Centralization of decision making | 3,56 | 4,00 | 2,33 | |
| 9. Limited influence over basic services | 3,44 | 3,56 | 3,44 | |
| 10. Political instability | 3,56 | 2,56 | 2,78 | |
| Project | 11. Inadequate project size | 3,67 | 3,56 | 3,11 |
| 12. Lack of a project leader | 4,00 | 4,00 | 3,56 | |
| 13. Lack of performance measurement tools | 4,00 | 3,44 | 3,00 | |
| 14. Poor data availability and analytics | 4,00 | 3,44 | 2,89 | |
| 15. Deficient or unreal planning | 4,33 | 3,44 | 2,67 | |
| 16. Lack of execution capacity skills | 4,33 | 3,44 | 2,22 | |
| Organization | 17. Lack of alignment of strategic goals and projects objectives | 4,11 | 4,11 | 2,78 |
| 18. Multiple or conflicting goals | 4,33 | 4,22 | 3,00 | |
| 19. Resistance to change | 4,11 | 3,67 | 2,67 | |
| 20. Lack of dedicated Smart City team | 4,00 | 4,11 | 3,11 | |
| 21. Lack of cooperation and coordination between departments | 4,33 | 3,89 | 3,00 | |
| 22. Lack of IT/technological knowledge | 3,89 | 3,67 | 3,00 | |
| Economy | 23. High cost of IT professionals and consultancies | 3,44 | 2,44 | 2,11 |
| 24. High cost of solutions | 3,67 | 2,33 | 2,67 | |
| 25. Cost of solutions’ installation, operation and maintenance | 3,89 | 2,44 | 2,00 | |
| 26. Cost of training | 2,38 | 2,56 | 3,33 | |
| 27. Lack of funding resources and financing opportunities | 3,50 | 3,33 | 2,44 | |
| 28. Lack of local competitiveness | 2,88 | 2,67 | 2,33 | |
| 29. Global economy volatility (risk and uncertainty) | 2,89 | 2,33 | 2,13 | |
| Socio-Cultural | 30. Lack of citizens' inclusion | 3,89 | 3,56 | 2,67 |
| 31. Lack of citizens’ accessibility to technology | 3,67 | 3,33 | 2,33 | |
| 32. Protection of heritage | 3,67 | 3,11 | 3,11 | |
| 33. Lack of trust between governed and government | 3,44 | 3,33 | 3,33 | |
| 34. Low Smart City awareness level of community | 3,33 | 3,11 | 3,11 | |
| 35. Unbalanced geographical development | 3,11 | 3,00 | 2,11 | |
| 36. Citizens' inequality | 3,67 | 2,67 | 1,78 | |
| 37. Social inertia | 3,11 | 2,89 | 2,00 | |
| 38. Lack of sharing culture | 3,56 | 2,67 | 1,75 | |
| Legal | 39. GDPR (privacy policy of personal data) | 2,89 | 2,67 | 2,89 |
| 40. Lack of regulatory norms | 3,22 | 3,11 | 3,22 | |
| 41. Complicated and long public procurement processes | 4,11 | 2,33 | 2,56 | |
| 42. Lack of transparency on public procurement | 3,56 | 3,33 | 2,89 | |
| 43. Long and complex procedures for authorizations/licenses | 3,56 | 2,67 | 2,67 | |
| Technology | 44. Lack of IT infrastructure | 3,78 | 3,00 | 2,22 |
| 45. Lack of integration capacity across systems | 3,67 | 3,00 | 2,33 | |
| 46. Lack of a unique data platform | 3,89 | 2,89 | 2,67 | |
| 47. System failures | 4,00 | 2,78 | 3,00 | |
| 48. Shortage of proven and tested solutions | 3,22 | 2,11 | 3,22 | |
| Environment | 49. Disregard for the environment and natural resources from policymakers | 3,67 | 4,22 | 2,78 |
| 50. Citizens’ lacking ecological view in consuming behaviour | 3,89 | 3,11 | 2,44 |
The most divergent barriers from the first Delphi round’s answers were harmonized through cluster analysis. The clusters are represented in Figure 7. Cluster number 4 for the Impact, cluster number 4 for Endogeneity and cluster number 2 for Space-time aggregated the barriers where the answers were more discordant (yellow). These groups were obtained based on a standard deviation threshold of 1.3 from the group’s average result. It corresponds to a Z-score value of 0.0968, which means that 9.68% of the total information was considered for further analysis in the second Delphi round.
The cluster analysis was also performed to find the participants' relationships concerning their answers to the first-round questionnaire, as shown in Figure 8.
Through this analysis, it is possible to notice that Elements 3 and 7 appear most often outside the clusters created by the remaining elements. This means these two participants had different opinions from the originated clusters. Thus, their views and response pattern did not match that of any other participant.
Figure 7. Barrier cluster analysis
Figure 8. Participant cluster analysis
It can be explained by their accumulated years of experience in the subject and the fact that both elements organize and manage large projects on a European and global scale with the involvement of multiple stakeholders and legislation. As for the others, it was noted that they do not follow a specific pattern across all criteria.
Analysis of the ResultsThe following sub-sections reflect the elements’ response changes observed in each criterion after the second Delphi round. The tables represent the changes on the Likert scale of each participant for the divergent barriers previously identified. The answers have converged to some key participants’ opinions. In terms of regulatory issues, the arguments of Element 3 were decisive to the answer changes of the audience. Element 7 demonstrated their expertise mostly in data privacy aspects. Elements 1 and 2 were followed carefully in terms of their feedback regarding citizens’ inclusion and financing. Element 8 demonstrated their moderated arguments with regard to the collaboration of public entities with private companies.
Barriers’ impactBy analysing Table 3, it is possible to verify that the responses to barriers 39 (“GDPR privacy policy of personal data”) and 40 (“Lack of regulatory norms”) were amended by more than half of the elements. This may be justified by the statement of Element 7, an ICT cluster board member, who said that "GDPR is not in itself a barrier but an opportunity” and shared his experience of the existence of companies that “want to come to Europe because of the respectfulness of privacy and ethics".
Element 3, a researcher of the European Commission’s innovation policy and strategy, noticed the lack of existing rules and regulations that “delay technologies’ market entrance” and may have a “strong impact on the implementation of a Smart City” by not keeping up with innovation by being “restricted to an older solution”. On the contrary, Element 7 stated that the lack of legislation may be an opportunity and not a barrier because it “gives space for innovation”, encouraging other players to join and help later regulatory processes.
The argument of Element 8, who is a municipality's department director, was about the fact that “often the use of technology is forced without having a real context for its use by citizens”, and that the lack of inclusion has a significant impact on the implementation of a Smart City which may explain the remaining answer changes. To which Element 1, a university professor, agreed by stating that “it is the most critical Smart City barrier” and, therefore, it is crucial to understand their literacy and to promote education because the other way around means that they do not become an active participant.
At the business level, Element 9 said that the most significant difficulty is still “opening up mentalities” and combating “the disregard for environmental issues”. Element 7 stressed that a Smart City should not only focus on environmental issues, waste management and mobility. Therefore, the “culture, vocational training, economic development, and the governance aspect” must be considered also. To which Element 5 added that “the denial of the lack of transparency and environmental considerations is the denial of a true Smart City”.
Table 3. Second questionnaire round answer changes (Likert scale) on the barriers’ impact by each participant
| Barrier | Element | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | |
| 29. Global economy volatility (risk and uncertainty) | 5 -> 4 | 1 -> 3 | |||||||
| 30. Lack of citizens' inclusion | 3 -> 4 | 1 -> 3 | |||||||
| 39. GDPR (privacy policy of personal data) | 5 -> 3 | 4 -> 3 | 5 -> 4 | 3 -> 4 | 1 -> 2 | ||||
| 40. Lack of regulatory norms | 4 -> 3 | 3 -> 5 | 1 -> 2 | 3 -> 2 | 1 -> 4 | ||||
| 42. Lack of transparency on public procurement | 4 -> 3 | 2 -> 3 | 1 -> 3 | ||||||
| 49. Disregard for the environment and natural resources from policymakers | 3 -> 2 | 2 -> 4 | |||||||
Element 2, which deals directly with financing opportunities, stated that in the face of extreme events there is a “reconditioning of the budget to other areas (namely health and social), thereby preventing Smart City strategies from moving forward”. Moreover, this can be witnessed by the lack of public investment in Smart Cities during 2020 (because of the Covid-19 pandemic). On the other hand, Element 7 mentioned that it could be seen as a way, with less budget, to think of more optimized forms of participation and progress, because despite the risks and economic volatility, “the digital era represents an opportunity for other cities lagging in the process without dedicated infrastructure”.
In Table 3, the cells represent the changes respondents have made in the second round from their first-round answers. As an example, the first cell (“5->4”) shows that Element 1 evaluated the barrier “Global economy volatility (risk and uncertainty)” with strong impact (level 5) and downgraded his evaluation to level 4 in the second round.
Barriers’ Endogeneity LevelTable 4 helps to understand the changes made by respondents according to the second criteria – Barrier’s Endogeneity Level. Barrier 41 (“Complicated and long public procurement processes”) was the one with more changes, all in the same orientation. This can be justified by the arguments of Element 5 and Element 8. The first stressed that "legislation is imposed on decision-makers”, and therefore, they can do little in that respect. The same happens to the “delay in public procurement". The second added that "in procurement processes that involve technology there is a huge interpretation to what can be done” which lead to very different proposals and make the evaluation process harder.
As for barrier 42 (“Lack of transparency on public procurement”), Element 9, as a Private Company CEO, stated that because of solutions subjectivity, it is usual that “procurement processes have their specifications conditioned on specific requirements or products”, because procurement was previously carried out, and policymakers know the specific solution they want to acquire. However, the changes verified in Table 4 may have been influenced by the perspective of Element 3. The expert on European Commission and regulation matters stated that specifications must be "transparent", where those who define the requirements have the opportunity to engage with the community, acquire knowledge through formal and informal methodologies, see and test several alternatives and define the procurement specifications, which are open to everybody. Element 7 added that it is the decision-maker’s responsibility to improve requirements' definition and bridge the lack of municipalities' infrastructure and technical teams by promoting an "open discussion with the community and other stakeholders (Universities, Clusters, and Associations) in a constant sharing of knowledge".
The cells of Table 4 represent the changes that Elements (columns) have performed to each barrier’s evaluation (lines) in the second round, from their first-round answers.
Table 4. Second questionnaire round answer changes (Likert scale) on the barriers’ endogeneity level by each participant
| Barrier | Element | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | |
| 39. GDPR (privacy policy of personal data) | 5 -> 4 | ||||||||
| 41. Complicated and long public procurement processes | 2 -> 4 | 2 -> 3 | 2 -> 4 | ||||||
| 42. Lack of transparency on public procurement | 5 -> 4 | 1 -> 3 | |||||||
The analysis of Table 5 shows that the discussion around barrier 42 (“Lack of transparency on public procurement”) triggered the change of responses by most elements. Element 8, with a municipality's background, expressed that "there are no proven and properly documented case studies" with accurate data of the solutions' impact on citizens' lives", meaning that the lack of transparency may have delayed the process and that "can be solved in the short term". Contrary to this view, Element 2, from a regulatory point of view, stated that "it is possible to quantify, however, the measurements have to consider the context analysis and experience of each person" and that "will not be achieved with a Likert scale of 1 to 10". In the same direction, Element 5, from the private sector, added that it is necessary to ensure that the systems implemented can collect data in real-time “to allow the monitoring and quantification of their benefit".
Additionally, the barriers with more changes were 8 (“Centralization of decision making”), 15 (“Deficient or unreal planning”), 30 (“Lack of citizens' inclusion”), and 39 (“GDPR privacy policy of personal data”). These may be explained by the arguments presented by Element 2, who stated that "centralization of decision-making will only be solved in the long term", and by Element 4, who gave the example of its municipality’s initial difficulty in realizing “the importance of citizen participation”. However, as soon as some solutions were implemented, they noticed “that the engagement tripled and public service itself was forced to be improved”, which is revealed to be in the hands of policymakers. Element 7, also stated that "a major cultural change is needed at various management levels on plans and objectives definition”, mapping them in relation to the projects that are being implemented, and that “will not be solved in the short term”; and Element 2 explained that "Smart Cities are still related to short-term agendas" and that implies that "there is no follow-up after the conclusion of the projects financing period”, resulting in the lack of long-term commitment. To the last observation, Element 7 interfered to agree and say that it still happens because of the inadequate size of projects for communication and marketing purposes, concluding that "it will take time to change the mindset”.
On the topic of regulation, Element 8 argued that "adequate legislation allows solutions to enter the market by less traditional means" and reduces the need for public investment, which can be solved in the short term. On the contrary, Element 3 shared that in their experience regulation is typically "carried out over the long-term and requires the creation of a joint proposal, or the perception of an existing trend by a sovereign body".
Table 5 demonstrates the changes that respondents have made in the second round. As an example, for barrier 6, “Lack of long-term commitment”, Elements 4, 7 and 8 changed their initial response.
Table 5. Second questionnaire round answer changes (Likert scale) on the barriers’ to be overcome in space-time by each participant
| Barrier | Element | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | |
| 6. Lack of long-term commitment | 4 -> 3 | 3 -> 2 | 3 -> 4 | ||||||
| 7. Low acceptance of new projects and technologies | 4 -> 3 | 4 -> 3 | |||||||
| 8. Centralization of decision making | 4 -> 3 | 4 -> 2 | 2 -> 1 | 2 -> 3 | |||||
| 11. Inadequate project size | 4 -> 2 | 4 -> 3 | |||||||
| 14. Poor data availability and analytics | |||||||||
| 15. Deficient or unreal planning | 1 -> 2 | 4 -> 3 | 3 -> 2 | 4 -> 3 | |||||
| 17. Lack of alignment of strategic goals and projects objectives | 2 -> 3 | 4 -> 3 | 4 -> 3 | ||||||
| 26. Cost of training | 5 -> 3 | 5 -> 4 | 2 -> 3 | ||||||
| 30. Lack of citizens' inclusion | 4 -> 3 | 2 -> 1 | 4 -> 3 | 2 -> 4 | |||||
| 32. Protection of heritage | 5 -> 4 | ||||||||
| 34. Low Smart City awareness level of community | 3 -> 2 | 4 -> 3 | |||||||
| 39. GDPR (privacy policy of personal data) | 1 -> 2 | 4 -> 3 | 4 -> 3 | 5 -> 4 | |||||
| 40. Lack of regulatory norms | 5 -> 2 | 5 -> 3 | 4 -> 3 | ||||||
| 42. Lack of transparency on public procurement | 5 -> 4 | 2 -> 3 | 5 -> 4 | 4 -> 3 | 1 -> 2 | ||||
| 48. Shortage of proven and tested solutions | 5 -> 4 | 5 -> 4 | 1 -> 3 | ||||||
Following the analysis of the changes made by each of the experts, it is possible to verify that Element 9 was the one that made more changes, having changed 66.7% of their initial answers. This may be because it is the element with fewer years of experience. Moreover, the arguments and respect for the other elements may have influenced them. Besides, it can also be noted that the two female elements were the only ones who did not make any changes to their first-round responses. On average, the remaining elements changed 40.48% of their responses.
During the focus group discussion, the various positions that supported the response to the first round of the questionnaire were noticeable. Element 7 proved in practice to be the most divergent element of the group. Additionally, the different opinions and contrary positions within the group were noticeable. Private company members had a more assertive with stronger critical opinion, and public body parties were more concerned about guaranteeing an open and inclusive mindset. Furthermore, during the Focus Group discussion, the relation between several barriers was perceptible. The qualitative relationships are presented in Table 6.
Table 6. Qualitative Barriers’ Relationships from Content Analysis
| Barriers | Evidence | |
|---|---|---|
| C1 – Impact |
30. Lack of citizens' Inclusion 31. Lack of citizens’ accessibility to technology 34. Low Smart City awareness level of community |
“Often we force technology without having a real context of its use by citizens" (Element 8) “If citizens lack literacy and appropriate education they will not be able to understand and be active the initiatives” (Element 1) |
|
1. Policymakers’ attitude 49. Disregard for the environment and natural resources from policymakers |
“Decision-makers' disregard for environmental matters makes difficult the progress on a large scale" (Element 9) "The denial of environmental considerations is smart city's own denial" (Element 5) |
|
|
27. Lack of funding resources and financing opportunities 29. Global economy volatility (risk and uncertainty) |
"In the face of extreme events, as happened with Covid-19 in 2020, there is a budget reallocation to other areas (namely health and social), disallowing strategies’ progress" (Element 2) | |
| C2 – Endogeneity |
39. GDPR (privacy policy of personal data) 38. Lack of sharing culture |
"The cultural context and the sharing culture are essential in a Smart City and are upstream of the GDPR itself" (Element 1) |
|
42. Lack of transparency on public procurement 1. Policymakers’ attitude 4. Lack of private-public partnerships (lack of stakeholders’ involvement) |
“Those who define the requirements should engage with the community, acquire knowledge through formal and informal methodologies, see and test several alternatives, and know the existing solutions in order to define specifications open to everybody" (Element 1) “It is up to the decision-makers to obtain a better definition of requirements, bridging the lack of infrastructure and team’s technical knowledge in municipalities opening the discussion to the community and involving other stakeholders (Universities, Clusters and Associations) in a constant knowledge sharing" (Element 7) |
|
| C3 – Space -time |
27. Lack of funding resources and financing opportunities 40. Lack of regulatory norms |
"Regulation requires the design of a proposal or the perception of an existing trend by a superior governmental body to push for the discussion and subsequently mobilize financial incentives" (Element 3) |
|
14. Poor data availability and analytics 21. Lack of cooperation and coordination between departments 1. Policymakers’ attitude |
"The change of mindset and the cooperation between departments are mostly related to the policymakers’ attitude. The Smart City strategy will only be effective when it is thought in holistic terms rather than vertically or departmentally." (Element 4) | |
|
6. Lack of long-term commitment 11. Inadequate project size |
"Smart Cities are still related to short-term agendas" (Element 2) "The lack of long-term commitment is due to project’s dimension usually for communication and marketing purposes, rather than smaller and easier to manage, with effective results, captivating citizens, and where the follow-up is not automatically dependent on funding issues” (Element 7) |
|
|
32. Protection of heritage 43. Long and complex procedures for authorizations/licenses |
"There is great difficulty in implementing a Smart City strategy in historic cities, because of the existing bureaucracy which turns it undesirable for innovation." (Element 3) |
These relationships help to understand that although some barriers may have less impact than others, they can be related and indirectly influence more significant barriers. Therefore, that should also be considered by policymakers.
Nevertheless, the results of the second Delphi round are specified in Table 7. These shall replace those from Table 2.
Table 7. Second Round Survey Average Results
| Areas | Barriers | C1 | C2 | C3 |
|---|---|---|---|---|
| Governance | 6. Lack of long-term commitment | 3,11 | ||
| 7. Low acceptance of new projects and technologies | 3,11 | |||
| 8. Centralization of decision making | 2,00 | |||
| Project | 11. Inadequate project size | 2,78 | ||
| 14. Poor data availability and analytics | 2,89 | |||
| 15. Deficient or unreal planning | 2,44 | |||
| Organization | 17. Lack of alignment of strategic goals and projects objectives | 2,67 | ||
| Economy | 26. Cost of training | 3,11 | ||
| 29. Global economy volatility (risk and uncertainty) | 3,33 | |||
| Socio-Cultural | 30. Lack of citizens' inclusion | 4,22 | 2,56 | |
| 32. Protection of heritage | 3,00 | |||
| 34. Low Smart City awareness level of community | 2,89 | |||
| Legal | 39. GDPR (privacy policy of personal data) | 2,67 | 2,56 | 2,67 |
| 40. Lack of regulatory norms | 3,56 | 2,56 | ||
| 41. Complicated and long public procurement processes | 2,89 | |||
| 42. Lack of transparency on public procurement | 3,78 | 3,44 | 2,78 | |
| 48. Shortage of proven and tested solutions | 3,22 | |||
| Environment | 49. Disregard for the environment and natural resources from policymakers | 3,78 |
The final results are represented in Figure 9, which positions the barriers according to the average result in each criterion. The vertical axis positions the barriers according to their endogeneity, that is, whether their resolution is within policymakers’ control (endogenous) or not (exogenous), while the horizontal axis reveals the opinion of experts considering the time-spacing in which the barriers can be overcome (short or long-term). The diameter of each circle provides information about their level of impact. The smaller circles have less impact than the larger ones.
Considering that this study aims to provide policymakers with information about the barriers they must acknowledge and prioritize, the attention shall be oriented to the barriers classified as endogenous. Thus, to find the Smart City critical barriers a three step-process followed.
First, the barriers were sorted by their Endogeneity result and divided into Endogenous (if the result was higher than 3) or Exogeneous (lower or equal to 3). After that, the barriers were sorted by their level of impact (from highest to lowest).
Second, because all Endogenous barriers had an impact superior to 3, it was considered a threshold of 4 for the impact to separate those with significant impact. Therefore, the barriers with an impact higher than 4 were considered as “Strong Impact”.
Finally, the critical barriers controllable by policymakers with significant impact were divided and sorted (from lowest to highest) for the space-time in which they can be overcome.
As previously mentioned, the action plan, which shall be the foundation of a Smart City strategy, must contemplate the short-term and long-term actions. The path with the defined thresholds described above to find the critical barriers is represented in Figure 10.
Figure 9. Distribution of the barriers according to their average level for each criterion
From the literature review, it was noticed that this topic is still unexplored, which can be explained by the lack of measurable results and conclusions from finished Smart City projects. Moreover, by recurring to Smart City experts to evaluate the barriers according to their impact, endogeneity level and space-time, this paper provides an overview to policymakers of how they ought to prioritize their decisions and ground an action plan.
The 15 critical barriers, represented on the top right corner of Figure 10, are mainly from the areas of Governance, Project, and Organization. Thus, the endogenous barriers, with high impact, ordered by time-priority were: (i) poor data availability and analytics; (ii) unclear vision/ lack of strategy; (iii) lack of alignment of strategic goals and projects definition; (iv) resistance to change; (v) lack of citizens’ inclusion; (vi) deficient or unreal planning; (vii) policymakers’ attitude; (viii) lack of execution capacity skills; (ix) lack of a project leader; (x) lack of public-private partnerships; (xi) lack of long-term commitment; (xii) lack of dedicated Smart City team; (xiii) multiple or conflicting goals; (xiv) lack of cooperation and coordination between departments; and (xv) lack of performance measurement tools. In summary, these barriers leverage the idea that a successful Smart City implementation depends on adequate organization, tools to evaluate and analyse data, and a skilled and dedicated team.
Besides, there is only one barrier (“30. Lack of citizens’ inclusion”) from the Socio-cultural area, which highlights the importance of including citizens. Thus, it provides the literature with significant insights about the need to define frameworks and participatory methodologies to enable the implementation of a Smart City strategy. This fact is aligned with the evolution of the Smart Cities concept. Smart Cities started to be techno-centric and moved to focus on sustainability and citizens' quality of life, breaking silos and promoting interoperability among solutions, allowing the city to have a real-time and integrated perspective. The placement of the end-user at the centre of decision-making is increasingly present and moving them forward. Nowadays, cities are increasingly created with and for the citizen. This reveals the alignment of the experts’ comprehension with the evolution of the Smart City concept and provides the literature with significant insights about the need to include social literature in the Smart City environment.
Figure 10. Representation of the journey with the respective thresholds for each criterion to find the critical Smart City barriers.
On the contrary, there are not any barriers from the technology area. Moreover, it is the only area that is not represented among the endogenous barriers. This leads to the comprehension that although Smart Cities’ implementation is at most times associated with technological challenges, those are not within the control of policymakers. Despite this, from the technological barriers, there is only one with a strong impact (47. “System failures”).
Additionally, there are not any financial-wise barriers amongst the endogenous with high impact. This sets apart the thinking that Smart City strategies’ successful implementation is dependent on economic and technological capacities pointed to by other studies’ results (Mosannenzadeh et al., 2017; Rana et al., 2019).
In the overall picture, there are only two exogenous barriers with significant impact, which leads to the conclusion that the success of Smart City implementation is within reach and is primarily dependent on policymakers’ actions.
In summary, to implement a successful Smart City approach, policymakers should acknowledge different data sources to have useful information from which to base their strategy for the territory, and perform continuous evaluation. After defining a long-term vision with the involvement of their citizens, they should appoint a project leader to organize it through smaller projects oriented toward the same strategic goals. A team with complementary skills shall be set up for each project, not disregarding the cooperation between departments and public-private partnerships to search for outside valuable skills and know-how.
One of the major gaps in the Smart Cities field is that policymakers do not have the basic know-how to plan a successful strategy, which is perpetuated by the lack of empirical knowledge sharing by those who have already experienced the process. Thus, the design of an effective strategic plan is dependent on the correct prioritization of actions to respond to identified challenges.
This paper started with an extensive review to find the critical barriers to Smart City implementation. This way, policymakers could have vital information to define an action plan to overcome each identified barrier based on its impact, endogeneity, and space-time.
Moreover, 15 critical barriers mainly from the areas of Governance, Project, and Organization were obtained. Besides these, there was only one additional barrier (“Lack of citizens’ inclusion”) from the Socio-cultural area, highlighting the necessity to include citizens. Technology and Economy are two of the areas that were not represented. Therefore, although Smart Cities are often associated with technological and financial challenges, based on this study's findings, those are not within the control of policymakers nor are relevant to the success of Smart City implementation. Furthermore, there are only two exogenous barriers classified with a significant impact which means that policymakers and their internal influence have a vital role.
In conclusion, the success of Smart City implementation depends on policymakers’ actions and their capacity to envision a plan with concrete objectives, build relationships, and set a skilled and dedicated team.
The Delphi questionnaire only counted two rounds, in quick succession, to attempt to prevent external phenomenon from disturbing or influencing the participants' second responses. The fact that there is no unlimited time to give experts the chance to reflect on the topic may present a limitation. However, this was combated from the analysis of their first answers, which allowed a more concrete and assertive discussion in the second round. Although the exercise format was all digital, although it favours the actors' attendance, it may also represent a limitation because of the lack of physical contact allowing for greater openness.
Key stakeholders are often left aside, and studies’ conclusions are based on theoretical modules without a broader comprehension of empirical evidence. Smart Cities have evolved to include citizens and other stakeholders in decision-making. Nevertheless, there still exists the need to develop participatory approaches to combine top-down and bottom-up perspectives.
In summary, based on experts’ perspectives, this paper provides a deeper characterization of the Smart City barriers considering their impact on the development of a Smart City, the policymakers' capacity to tackle them and within what period of time. New studies can emerge upon the results to help decision-makers in the Smart City implementation process. The findings significantly contribute to the literature the importance of including social literature in Smart Cities’ scope and the need to define frameworks and participatory methodologies to implement a Smart City strategy.
