COMMENTARY Toward achieving smart cities inAfrica: challenges to data use and the way forward Ernest Agyemang1 , Brilé Anderson2 , Jorge Patiño2 and Marie Tremolieres2 1Department of Geography & Resource Development, College of Humanities, University of Ghana, Accra, Ghana 2OECD Sahel and West Africa Club, Paris, France Corresponding author: Ernest Agyemang; Email: eagyemang@ug.edu.gh Received: 21 October 2023; Revised: 12 February 2024; Accepted: 14 February 2024 Keywords: African cities; data sharing; Ghana; innovation imperative; smart cities; urbanization Abstract The advent of smart and digital cities is bringing data to the forefront as a critical resource for addressing the multifaceted transitions faced byAfrican cities from rapid urbanization to the climate crisis. However, this commentary highlights the formidable considerations that must be addressed to realize the potential of data-driven urban planning and management. We argue that data should be viewed as a tool, not a panacea, drawing from our experience in modeling and mapping the accessibility of transport systems in Accra and Kumasi, Ghana. We identify five key considerations, including data choice, imperfections, resource intensity, validation, and data market dynamics, and propose three actionable points for progress: local data sharing, centralized repositories, and capacity-building. While our focus is on Kumasi and Accra, the considerations discussed are relevant to cities across the African continent. Policy Significance Statement The findings in this paper will guide policymakers in Africa and beyond to identify the key barriers besetting the full attainment of the innovation imperative agenda in making cities smart, resilient, and sustainable. The paper further highlights the necessity of policy actors undertaking proactive measures in building local capacity, establishing centralized data repositories, contributing to open source and global ones, and fostering local-level data-sharing initiatives for a more informed and resilient urban planning landscape in African cities. 1. Introduction There is a strategic commitment by policymakers acrossAfrica to embrace innovation and digitalization as outlined in the Agenda 2063 development framework (African Union Commission, 2023). Indeed, during the 32ndAfricanUnionAssembly of Heads of State andGovernment during a high-level luncheon dubbed “Digital Transformation for Africa: leaping into the African Century,” a representative of the African Union Development Agency (AUDA-NEPAD), Mr. Symerre Grey-Johnson, delivering his keynote speech, noted, “as we enter into the fourth industrial revolution, we have to ensure that our policies and guiding frameworks for infrastructure development evolve and to develop the complementary skills that will drive smart infrastructure that Africa urgently requires” (AUCommission, 2019). However, in Africa, strategies andmeasures to fully adopt the smart city concept—underpinned by the notion that “rational and ©TheAuthor(s), 2024. Published byCambridgeUniversity Press. This is anOpenAccess article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited. Data & Policy (2024), 6: e13 doi:10.1017/dap.2024.11 https://doi.org/10.1017/dap.2024.11 Published online by Cambridge University Press https://orcid.org/0000-0003-3254-4611 https://orcid.org/0009-0001-0343-8134 https://orcid.org/0000-0002-9767-7522 mailto:eagyemang@ug.edu.gh http://creativecommons.org/licenses/by/4.0 https://doi.org/10.1017/dap.2024.11 http://crossmark.crossref.org/dialog?doi=https://doi.org/10.1017/dap.2024.11&domain=pdf https://doi.org/10.1017/dap.2024.11 optimal decisions, as well as intelligent behaviors, can be attained in city environments through the use of information and communication technologies (ICTs)” (Slavova and Okwechime, 2016, p. 214)—are isolated and limited. Notable initiatives so far include Rwanda’s Smart City Masterplan (UN-Habitat, 2017), Nigeria’s Smart City Initiative, and Ethiopia’s Smart Parking Initiatives (Siba and Sow, 2017). The rest are Morocco’s City of Casablanca Digital Project, Mauritius’ Moka Smart City, and Kenya’s iHub Initiative (Maphangwa and van der Waldt, 2023). Despite the limited number of smart cities on the continent, Bandauko andArku (2023) argue that the concept holds transformative potential in the planning, management, and governance of African cities. Elsewhere, it has been found that by adopting such innovative technologies and solutions, cities may advance their unique position to tackle many of the sustainability challenges faced globally today (Cuperstein, 2022; Nilssen and Hanssen, 2022). It is recognized that cities can drive global change at an unprecedented scale by transforming land cover, hydrological systems, climate, biogeochemistry, and habitats (Espey et al., 2024). This expected global change is premised on better coordinated and innovative smart city policies that are obtained through data to foster the political and social inclusion of citizens when it comes to decision-making and service provision, which then leads to improvement in the quality of life (Dawes et al., 2016; Pereira et al., 2018; Liva et al., 2023). However, even in the wealthier and more industrialized countries of the Global North, several organizational, technological, and legal barriers beset the potential of cities to reap the full dividends accruing from the smart city agenda (Liva et al., 2023). It, therefore, comes as no surprise that in poorer countries of the Global South, formidable barriers to smart cities, from legislative (Danquah et al., 2019) to governance, technological, social, legal, and ethical issues, continue to persist (Kolandaisami, 2020). Thus, an enabling environment is still needed to facilitate the transition toward smart cities to effectively reduce social, economic, and environmental disparities emerging in cities. Otherwise, smart cities risk entrenching inequalities, as highlighted by Dos Santos and Mota (2019). Central to the realization of smart cities in Africa, in particular, is the recognition of data as an indispensable asset to empower decision-making capacities among diverse stakeholders. Open data, in particular, assume a pivotal role in addressing the swift transitions (e.g., urbanization) and intricate challenges (e.g., climate crisis) confronting African cities. Nikiforova et al. (2023) underscore its role in providing insights and solutions within compressed time frames. Despite the critical importance of open data, a counteractive trend is emerging in the African context, creating a potentially adverse environment for smart cities to reach their full potential. Case studies drawn from East Africa highlight the key challenges involved in sourcing open data on the African continent ranging from over-reliance on external donor support, governmental restrictions, and limited motivation for community engagement. For instance, national open data portals, which had been launched in Kenya and Tanzania, failed to thrive and became self-sustaining following the cessation of continued donor support (Vasileva, 2022). Earlier, Ranchod (2020) had highlighted bureaucratic management and compliance protocols, limited and frag- mented organizational capacity, and politicization of data as key constraints toward the use of open data in South African secondary cities. Ricker et al. (2020) also found that city authorities, perhaps viewing civil society organizations as risky or confrontational, have been half-heartedly involved in promoting their open data and wider data activism initiatives. Consequently, the rapid adoption of digital technologies in Africa occurs organically and often without public intervention. This phenomenon, exemplified by the “datafication of cities” (Oksman and Raunio, 2018; Bibri and Krogstie, 2020; Plantinga, 2022), raises concerns about the potential detachment of local authorities from the evolving digital landscape. Such detachment may manifest in the development of alternative digital services by external entities, disrupting or competing with locally provided public services. Additionally, the risk exists that local authorities may face constraints in accessing comprehensive data generated by users and businesses within their jurisdic- tion, hampering informed decision-making and, ultimately, the transition to smart cities. Because of this, the data landscape unfolding in African cities today is notably intricate, characterized by a multitude of sources, ranging from private to public entities. Importantly, this data ecosystem is far from being scarce, particularly concerning geospatial data. Establishing a conducive environment for smart cities entails adeptly navigating the complexities inherent in these emerging data environments to facilitate informed e13-2 Ernest Agyemang et al. https://doi.org/10.1017/dap.2024.11 Published online by Cambridge University Press https://doi.org/10.1017/dap.2024.11 decision-making. It is crucial to contemplate the methodologies for data utilization and the consequential implications of decisions to formulate standards and best practices. Surprisingly, the discourse on this aspect appears relatively sparse within the existing literature. This commentary delves deeper into issues surrounding the use of data, particularly open data, in decision-making in African cities and draws on our experience working with local actors and funders to analyze the accessibility and sustainability of transport systems inAccra andKumasi (Ghana). This project grew out of one critical urban management challenge in cities, the inability of transportation systems to keep pace with rapid urban growth. Accra and Kumasi, along with other cities across the continent, are experiencing increased disconnection, burdening residents with longer travel times, greater congestion, and rising pollution and emissions, alongwith increasing motorization rates. Therefore, our project sought to spatially map accessibility to primary schools; hospitals, pharmacies, and clinics; markets (informal/ formal); and transport hubs and green spaces, by transport modes (i.e., walking, biking, public transport— trotros and taxis—ride-hailing, motorcycles, and cars) across the city, in order to identify possibilities to improve the accessibility, sustainability, and equity of transport systemsmoving forward. For example, by providing better road space allocation, the reliability, convenience, and safety of active and shared modes will be promoted (OECD, 2021; Pajares et al., 2021; Allam et al., 2023). The reliance on active and shared transport modes among urban dwellers when accessing essential services ideally within relatively short commute distances (i.e., typically 15 or 30 minutes), also known as the “X-Minute City,” has gained considerable traction and acceptance among urban planners and policy actors as a sustainable development pathway (Allam et al., 2023; Birkenfeld et al., 2023; Vich et al., 2023; Willberg et al., 2023). We generated a primary dataset for building accessibility models through strategic partnerships involving researchers from the University of Ghana, the University of Manchester Spatial Policy and Analysis Laboratory, and the Organisation for Economic Co-operation and Development (OECD), together with the International Transport Forum (ITF). We adapted the ITF/OECD (2023) Accessibility Model to suit the Ghanaian context. Such a modeling exercise, as was undertaken in our previous study, however, requires copious amounts of spatial data from points of interest (e.g., clinics) to the transport network. Throughout our project, a concerted effort was made to leverage open-source data whenever feasible. In addition, household surveys involving 1,000 participants recruited from both Accra and Kumasi were undertaken. A comprehensive overview of the research design, including the sampling approach and the initial results generated, has been articulated elsewhere (see Agyemang et al., 2023). This commentary contributes to the literature since it stands as an applied exploration of utilizing existing data for decision-making, offering practical insights for a transition to smart cities in Africa. Notably, our review of the existing literature identified the work of Peprah et al. (2019) as a solitary contribution to our understanding of smart cities inGhana.Nevertheless, the implications derived fromour commentary extend beyond the Ghanaian context and hold relevance for smart city initiatives across the African continent. The next section expands on five key considerations that warrant heightened attention in future endeavors: selecting which source of open data to use, imperfections in nearly all datasets whether open or private, the resource intensity (in terms of time) required to use these datasets, the continued need for ground truth validation, and the lack of access to public data sources. The third section discusses the actionable points and examples to overcome these obstacles: the need for capacity-building at the local level along with several initiatives, improved data sharing among different entities, and, lastly, the need for a centralized repository for data collection. The final section concludes on our experience and underscores the imperative that, notwithstanding the abundance of available data, discerning choices and meticulous due diligence must be exercised in its utilization. Decision-makers in urban contexts must critically reflect upon these consider- ations to ensure that smart city ambitions enhance rather than deteriorate the quality of life. 2. Key considerations on using data moving forward Our experience mapping the accessibility of transport systems in Ghana highlighted different consider- ations as we move into this era of smart cities, not only for African cities—but worldwide. Some of the issues that emerged from our study are presented and discussed in the ensuing section. Data & Policy e13-3 https://doi.org/10.1017/dap.2024.11 Published online by Cambridge University Press https://doi.org/10.1017/dap.2024.11 2.1. Data source selection We were often confronted with a multitude of data sources related to the same aspect (e.g., urban boundaries, road networks, points of interest), sometimes open and others not, but often it can be unclear which data source is “best.” For instance, in the context of defining boundaries for the two cities under consideration, we found it necessary to opt for an administrative area-based boundary delineation. However, it is worth noting that alternative choices, such as functional urban areas or the delineation based on built-up urban extent, were equally feasible and, in some respects, potentially more “accurate” due to the rapid pace of urbanization. There are at least four open datasets available solely for defining urban boundaries. These include administrative data (i.e., the Database of Global Administrative Areas or locally sourced administrative boundaries datasets), Africapolis (only available for 2015) (reflecting functional urban extent) (OECD/ SWAC, 2020), the Global Human Settlement Layer (GHS) Functional Urban Areas (only available for 2015) (Schiavina et al., 2019), and theGHSUrbanCentersDatabase (available for 1975, 1990, 2000, 2015, and 2019) (Florczyk et al., 2019). Moreover, there are several different global datasets that allow the identification of built-up areas globally (that are also open source) and that could serve as a starting point to delineate the urban extension. These include the GHS built-up surface dataset (latest release 2023) (Pesaresi and Politis, 2023), the World Settlement Footprint from the German Aerospace Center (latest release 2019) (Marconcini et al., 2020), theAnnualUrban Extents from harmonized nighttime lights (latest release 2020) (Zhao et al., 2022), and the Global Urban Entities based on consistent nighttime light data (latest release 2020) (Shi et al., 2023). As illustrated in Figure 1, each dataset draws a different boundary for the reference year 2015 (since this year is available in all datasets), with stark differences between these datasets. Notably, the administrative boundary, Africapolis, and GHS Functional Urban Area cover more extension, in general, than the other datasets, which are more contained in the consolidated urban cores. The latter tend to exclude the peri-urban areas that are very likely to access services and amenities located in the urban core. All these data sources are freely available—and perhaps, there will be even more potential sources in the future. The selection of boundaries has far-reaching implications in our project and for others pursuing spatial analysis, since the model only examines accessibility within the boundaries, meaning some urban residents could be inadvertently left out. Administrative area-based boundaries tend to align more closely with local decision-making processes. Yet, accessibility fundamentally pertains to where people reside, which may not always correspond with administrative realities, and areas that are not recognized as urban can have population densities high enough to be considered in transport and mobility planning processes (Gao and O’Neill, 2020). Beyond the challenge of inconsistency in the definition of boundaries, quality problems such as inaccuracy, out-of-dateness, and incompletenessmay limit the full exploitation of linked open data (Rula et al., 2016). As we move forward and contemplate the integration of such diverse data sources into the decision-making processes of local stakeholders and researchers, it is essential to consider the far-reaching implications of these choices, particularly with respect to resource allocation and future infrastructure, to create best practices. 2.2. Data imperfections: Integration Our project requires road network information, including aspects such as road surface type, road classification, speed limits, and traffic speeds during peak and off-peak hours. The research project procured data from a private provider, specializing in the road network of Ghana, offering essential details including road classifications, speed estimations, and surface attributes. The private data provider leverages a visual training algorithm, in tandem with ground-level validation, to process satellite imagery. Upon thorough examination, gaps in the road network from the private provider were identified, particularly in the western region of Accra, supported by satellite imagery (Figure 2). This revealed a completeness issue (Vetrò et al., 2016; Torchiano et al., 2017) that required addressing beforemodeling. To resolve this, an amalgamation of datasets fromOpenStreetMap and the private provider was undertaken to create a comprehensive network. The consistency issues (Vetrò et al., 2016; Torchiano et al., 2017) between e13-4 Ernest Agyemang et al. https://doi.org/10.1017/dap.2024.11 Published online by Cambridge University Press https://doi.org/10.1017/dap.2024.11 Figure 1.Delimitation of the urban functional area in the Greater AccraMetropolitan Area by year 2015 as defined by several open geospatial datasets available in 2023. Top row, from left to right: adminis- trative boundaries of the districts in the conurbation; Africapolis (OECD/SWAC, 2020); GHS Functional Urban Area (Schiavina et al., 2019); GHS Urban Centers Database (Florczyk et al., 2019). Bottom row, from left to right: GHS built-up surface (Pesaresi and Politis, 2023); World Settlement Footprint from DLR (Marconcini et al., 2020); urban extent from harmonized nighttime light imagery (Zhao et al., 2022); urban area from theGlobal Urban Entities (Shi et al., 2023). All maps are shown at the same spatial scale. Basemap: OpenStreetMap web map service (© OpenStreetMap contributors). Source: The authors. Data & Policy e13-5 https://doi.org/10.1017/dap.2024.11 Published online by Cambridge University Press https://doi.org/10.1017/dap.2024.11 datasets led to increased data processing time and the need for specialized resources to produce a suitable road network for modeling. This amalgamation process underscores the imperative recognition that no dataset is immune to imperfections, as noted in previous studies (Batton-Hubert et al., 2019; Le Galès and Robinson, 2024). While commercially procured datasets may possess certain advantages, such as, in our case, speed estimates and classification of pavement, they are not devoid of shortcomings—that is, the absence of critical road segments being a noteworthy concern.Moreover, this confluence of data sources underscores the diverse origins of data, each with its unique strengths and weaknesses, which are critical consider- ations moving forward. 2.3. Temporal and financial resources In addition, as already alluded to above, the team made formidable investments, primarily in terms of not only time but also financial resources, to acquire specific datasets for our model. Spatial data have the potential to bring a significant change in urban management. However, it is not always readily available from public sources. This means that considerable time is needed to procure and process open data or financial resources to purchase them from private providers. Our project, initiated in September 2022, involved an extensive data processing and validation phase, which extended beyond six months, specifically focused on two urban centers. The extended timeline was primarily attributable to amultifaceted approach. Firstly, the identification of requisite data sources involved the curation of information from various open-source and private providers. Subsequently, a meticulous data processing effort was undertaken to prepare the data for use. Thirdly, the assurance of data quality— defined as the ability of data tomeet the “needs and expectations of the data user” (Nikiforova, 2020, p. 419) —was a pivotal concern throughout this endeavor, necessitating comprehensive validation procedures. Remarkably, this project was executed by a team of eight individuals, geographically dispersed across different regions of the world. This collaborative effort underscores the challenges and complexities Figure 2.Data gap of the privately sourced road network, located in the southwestern part of the Greater AccraMetropolitan Area. Basemap:Google Satellite webmap service accessed in June 2023 (©Google). The privately sourced road network is shown in yellow. The overlay of the road network dataset with the satellite base map shows that a significant portion of the urban layout in this area lacked data from the private provider. Source: The authors. e13-6 Ernest Agyemang et al. https://doi.org/10.1017/dap.2024.11 Published online by Cambridge University Press https://doi.org/10.1017/dap.2024.11 associated with data acquisition and processing for urban spatial analysis (Liu et al., 2021; Le Galès and Robinson, 2024). If spatial data are to play a more active role in enhancing urban management, it is imperative that availability and usability be prioritized, streamlining the intricate processes involved in data procurement and analysis. 2.4. Ground truth validation Throughout the duration of this project, the inherent constraints associatedwith spatial data,whether of open or private origin, have become progressively conspicuous. This underscores the pressing need for ground truth validation, underpinned by insights from local experts of various backgrounds. In the current era of smart cities, there may be a propensity to view data as an all-encompassing solution for informed decision- making. However, our experience has revealed that ground truth validation is an indispensable step for garnering support and safeguarding the precision and dependability of our analytical outcomes. This validation process offers the opportunity to fine-tune and rectify the data. When disparities emerge between the collected data and ground truth observations, necessary adjustments can be made to align the data with real-world conditions, consequently enhancing data quality, as has been found elsewhere (Yan et al., 2018). Furthermore, our findings highlight the enhanced trustworthiness of validated data. Users, stakeholders, and decision-makers exhibit a greater inclination to place their trust and reliance on data that have undergone the rigor of ground truth validation, ultimately resulting in more enlightened and assured decision-making practices. 2.5. Data market: Absence of public sector It is worth noting that none of the data used in the project came from public entities, which is not by choice. We struggled to even access population density data from the National Statistical Office, even though we were working alongside local authorities. The Ghana Statistical Service (GSS) has been mandated under the Statistical Service Act, 2019 (Act 1003), to act “as the central statistics producing and coordinating institution for the National Statistical System and to strengthen the production of quality, relevant, accurate and timely statistical information for the purpose of national development.”On its website, the GSS boasts of being “consistent in performing its legal functions to serve the needs of individuals, organizations, development partners and the government without hindrance.” The practice is simply for a requester, as either individuals or organizations, to fill out a Data Request Agreement form for submission online. For our research, we initially requested enumeration area (EA) level census data obtained during the 2021 Population and Housing Census (PHC) conducted on Sunday, June 27, 2021. In addition, we requested geographic information system (GIS) shapefiles for selected districts in Accra and Kumasi. Our request was submitted in December 2022. The first email response that we received was that the GSS does not release data at such micro-level but typically shares data aggregated at the locality, regional, or district level. We complied and requested a locality-level dataset instead. However, the second and final email response we received was that such data, even at the locality level, were not readily available, some 18months after the census was conducted. Even though this project feeds into ongoing work by the local administrations in Accra and Kumasi, we were unable to access the data needed from the public administration. There might be valid reasons for this, such as privacy and data protection concerns. Public entities often handle sensitive data, and sharing data could raise concerns about privacy violations. Public entities might be subject to a complex web of regulations and laws that govern data sharing.Government agenciesmay have bureaucratic structures and procedures that can slow down or hinder data-sharing efforts. The answer could also be simpler—lack of resources or interdepartmental conflicts. Many public entities may lack the necessary resources, in terms of both technology and staff, to effectively manage and share data. Likewise, different government agencies or departments may have conflicting interests or objectives. As mentioned earlier, fragmented governance and half-hearted engage- ments of city authorities are seen as major barriers to the development of smart cities and the development of open data ecosystemsworldwide (Ranchod, 2020; Ricker et al., 2020; Kitchin andMoore-Cherry, 2021; Vasileva, 2022). Data & Policy e13-7 https://doi.org/10.1017/dap.2024.11 Published online by Cambridge University Press https://doi.org/10.1017/dap.2024.11 We, therefore, were compelled to utilize data from the privately operated community or emanating from the open-source community. It cannot be overemphasized that whatever the reason was that led to our inability to procure the data from the GSS, if these barriers and hurdles are not resolved, Accra and Kumasi, and other cities in Africa, risk being left behind and missing out on reaping the full potentials associated with the “innovation imperative” agenda, as has been cautioned elsewhere in Europe (Liva et al., 2023). 3. Policy implications: Points of action As we navigate the era of enhanced data access and management, the policy implications in addressing how data are used and shared in public administrations, at both local and national levels, are pivotal. This consideration is intertwined with the evolving discourse of decentralization in many African countries. Cities play a critical role in addressing climate change and rapid urbanization, and better access to and sharing of data among international organizations, government, non-profit or private sectors, and the general public are imperative to enable them to fulfill this role effectively. Against this backdrop, we propose three policy points of action to help address some of these considerations. First, investing in capacity-building at the local level to enhance data literacy and analytical skills among urban planning professionals will ensure the implications of data choices and imperfections are reasonably well understood. Training programs and educational initiatives can empower local governments to harness the full potential of data, while being some of the best actors for ground truth validation of the data. Specifically, for Ghana, we think collaborations like the Ghana-India Kofi Annan Centre of Excellence in ICT (AITI-KACE), an ongoing strategic partnership to enhance digital skills acquisition and inclusivity in the digital economy, particularly among vulnerable groups, are a step in the right direction. For the rest of the African continent, specializations, such as those in the Executive Masters in City Management at the African Local Governments Academy (UCLG Africa), are dedicated to equipping professionals with the skills to utilize freely accessible spatial data in land-based planning and financing. Similarly, initiatives like Digital Earth Africa offer courses to local actors, enabling them to effectively employ open geospatial data. Empowering individuals with these capabilities enhances their ability to assess and evaluate the implications of data-driven choices, reducing reliance on external expertise for territorial development guidance. Second, local data-sharing initiatives foster collaboration amongAfrican stakeholders, includ- ing government agencies, researchers, and businesses and lead to more comprehensive and context-specific datasets that better serve the needs of smart cities. This also helps varying actors understand the implications and limitations of using various sources. Promoting the open exchange of data at the grassroots level is imperative. Encouraging contributions to open-sourcemapping platforms, such as OpenStreetMap, can ensure the ongoing maintenance and updating of critical geospatial datasets, including road networks, buildings, and service and amenity locations within cities. Neglect- ing such initiatives could result in missed opportunities for comprehensive planning and resilience in the face of urbanization and climate challenges. Additionally, failing to engage in these efforts may inadvertently enable the private sector to dominate the data landscape, potentially undermining public interests. To this end, we commend the Open Cities Africa initiative, which has been developing comprehensive points of action for eleven cities in sub-Saharan Africa (including Accra). Its core mission is to foster collaboration among local government entities, civil society organizations, and private sector stake- holders. The overarching objective is to establish the critical information infrastructure needed to address the intricate urban resilience challenges characteristic of the twenty-first century. This initiative represents a unique partnership forged between the Global Facility for Disaster Reduction and Recovery (GFDRR), the World Bank, municipal authorities throughout the African continent, and a consortium of regional scientific and technological organizations, development partners, and technology enterprises. Another noteworthy local data sharing and collaboration project is that of the MapMakoko project in Lagos e13-8 Ernest Agyemang et al. https://doi.org/10.1017/dap.2024.11 Published online by Cambridge University Press https://doi.org/10.1017/dap.2024.11 (Nigeria). This initiative brought on board Code for Africa, Humanitarian OpenStreetMap, the Pulitzer Center on Crisis Reporting, africanDRONE, Makoko Dream, CcHUB’s American Corner, and the Guardian Nigeria. They implemented a bottom-up mapping approach combining data from satellite images and drones along with community participation to map waterways and the location of critical infrastructure within the settlement. More than 80 points of interest were added to the map, illustrating how innovative technologies (such as drones and satellite imagery) often provide the data that underpin and advance the aims of smart cities. Technologies, such as the installation of sensors and telecommu- nication systems, or even artificial intelligence, as well as machine learning technologies, can make it feasible to collect near real-time data; providing a deeper understanding of how cities evolve, adapt, and respond to various conditions (Allam and Dhunny, 2019; Gartoumi and Tékouabou, 2023). Third,African policy actorsmust prioritize the creation of centralized repositories whether at the local, regional, or national levels, to ensure equitable access and diffusion of data resources. These repositories can facilitate data discovery, sharing, and maintenance, reducing the dependency on third- party data providers, in addition to offering an opportunity for policy actors to create a shared vision, which helps to foster smart city data ecosystems (Gupta et al., 2020). Yet as is often the case, as already noted by Slavova and Okwechime (2016, p. 224), “visionary leadership is critical in avoiding this pitfall, as well as developing mechanisms for shared governance, empowerment and co-creation.” 4. Conclusion In a nutshell, this commentary sought to explore the pivotal considerations and challenges that surround the use of data in urban planning and management. Our work has highlighted five critical considerations as cities start to increasingly use data in urbanmanagement. (1) Data selection: The choice ofwhich data to use will become increasingly complicated as data become progressively available with greater digital- ization, but this selection can lead to varying conclusions and has profound implications for investment and infrastructure development. Decisions about data sources and types should be madewith utmost care. (2) Data imperfections: Every dataset, no matter how seemingly technical or robust, whether open source or private, will contain limitations. These imperfections may include errors, biases, and gaps. It is crucial for data users to be aware of these and account for them in their analyses and decision-making processes. (3) Resource intensity: Our project’s experience underscores the considerable time and resource intensity associated with collecting, managing, curating, and analyzing the data. (4) Ground truth validation: Data ideally should be validated against ground truth, to the greatest extent possible. The absence of ground truth validation can introduce uncertainties and potential inaccuracies into decision-making processes, emphasizing the need for robust validationmethods. (5) Datamarket dynamics:We observed a prevalence of third-party entities dominating the data market, while comprehensive access to publicly provided data is conspicuously absent. This dearth of accessible data poses a significant impediment to local govern- ment’s capacity to formulate informed policies and initiatives. While our focus has been on the experiences gained in Kumasi and Accra, it is essential to recognize that the challenges and considerations discussed in this commentary are endemic to cities throughout the African continent. Data-driven urban planning holds immense promise, especially given the pivotal role that cities play in tackling the climate crisis. Yet, realizing this potential requires a comprehensive understanding of the complexities and a concerted effort to overcome the identified challenges. By adopting a strategic approach that includes local data sharing, centralized repositories, and capacity-building, African cities can harness the power of data to tackle their urban challenges effectively. Author contribution. B.A. conceptualized the study. B.A., E.A., J.P., and M.T. designed the methodology. J.P. curated the data. J.P. visualized the data. B.A., E.A., and J.P. wrote the original draft. All authors approved the final submitted draft. Funding statement. This research was supported by funding from the Swiss Agency of Development Cooperation. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Competing interest. The authors declare none. Data & Policy e13-9 https://doi.org/10.1017/dap.2024.11 Published online by Cambridge University Press https://doi.org/10.1017/dap.2024.11 References African Union Commission (2019) Smart infrastructure accelerates regional integration, say experts at African Union STC on Energy, Transport and Tourism. Directorate of Information and Communication, Addis Ababa. Available at https://au.int/en/ pressreleases/20190416/smart-infrastructure-accelerates-regional-integration-say-experts-african. African Union Commission (2023) Our Aspirations for the Africa We Want. Available at https://au.int/agenda2063/aspirations. Agyemang E, Anderson B, Owusu K, Quinchia JP and Trémolières M (2023) Ensuring sustainability via accessible transport systems for all in Accra and Kumasi. In E3S Web of Conferences, Vol. 418. EDP Sciences, p. 02001. https://doi.org/10.1051/ e3sconf/202341802001. Allam Z and Dhunny ZA (2019) On big data, artificial intelligence and smart cities. Cities 89, 80–91. https://doi.org/10.1016/j. cities.2019.01.032. AllamZ,MorenoC,ChabaudDandPratlongF (2023) Proximity-based planning and the “15-Minute City”: A sustainablemodel for the City of the future. In The Palgrave Handbook of Global Sustainability. Cham: Springer International Publishing, pp. 1523–1542. https://doi.org/10.1007/978-3-030-38948-2_178-1. Bandauko E and Arku RN (2023) A critical analysis of ‘smart cities’ as an urban development strategy in Africa. International Planning Studies 28(1), 69–86. https://doi.org/10.1080/13563475.2022.2137112. Batton-HubertM,Desjardin E and Pinet F (2019)Geographic Data Imperfection 1: From Theory to Applications. Hoboken, NJ: John Wiley & Sons. Bibri SE and Krogstie J (2020) The emerging data–driven smart city and its innovative applied solutions for sustainability: The cases of London and Barcelona. Energy Informatics 3, 1–42. https://doi.org/10.1186/s42162-020-00108-6. Birkenfeld C,Victoriano-Habit R,Alousi-Jones M, Soliz A and El-Geneidy A (2023)Who is living a local lifestyle? Towards a better understanding of the 15-minute-city and 30-minute-city concepts from a behavioural perspective in Montréal, Canada. Journal of Urban Mobility 3, 100048. https://doi.org/10.1016/j.urbmob.2023.100048. Cuperstein I (2022) Cities, at the forefront of sustainable transformation. Field actions science reports. The Journal of Field Actions, Special Issue 24, 48–53. https://journals.openedition.org/factsreports/6920. Danquah JA,Marful AB and Duah D. (2019) Exploring barrier-free as a catalyst to smart cities initiatives in Sub Saharan Africa. In Proceedings of 24th International Conference on Urban Planning, Regional Development and Information Society. REAL CORP. https://core.ac.uk/download/pdf/199447703.pdf Dawes S, Vidiasova L and Parkhimovich O (2016) Planning and designing open government data programs: An ecosystem approach. Government Information Quarterly 33(1), 15–27. https://doi.org/10.1016/j.giq.2016.01.003. Dos Santos ML and Mota M (2019) Toward sustainable and smart cities in Africa: A review and challenges. In Bioclimatic Architecture in Warm Climates: A Guide for Best Practices in Africa. Springer. pp. 299–309. Available at https://link.springer. com/chapter/10.1007/978-3-030-12036-8_10. Espey J, Keith M, Parnell S, Schwanen T and Seto KC (2024) Designing policy for Earth’s urban future. Science 383(6681), 364–367. https://doi.org/10.1126/science.adi6636. FlorczykA,Corbane C, SchiavinaM, Pesaresi M,Maffenini L,Melchiorri M, Politis P, Sabo F, Freire S,Ehrlich D,Kemper T, Tommasi P, Airaghi D and Zanchetta L (2019) GHS Urban Centre Database 2015, multitemporal and multidimensional attributes, R2019A. European Commission, Joint Research Centre (JRC). Available at https://data.jrc.ec.europa.eu/dataset/ 53473144-b88c-44bc-b4a3-4583ed1f547e. Gao J and O’Neill BC (2020) Mapping global urban land for the 21st century with data-driven simulations and shared socioeconomic pathways. Nature Communications 11(1), 2302. https://doi.org/10.1038/s41467-020-15788-7. Gartoumi KI and Tékouabou SCK (2023) Smart-BIM for smart cities: Issues and challenges. In E3S Web of Conferences. Vol. 418. EDP Sciences, p. 03004. Available at https://www.e3s-conferences.org/articles/e3sconf/pdf/2023/55/e3sconf_acc2023_ 03004.pdf. Gupta A, Panagiotopoulos P and Bowen F (2020) An orchestration approach to smart city data ecosystems. Technological Forecasting and Social Change 153, 119929. https://doi.org/10.1016/j.techfore.2020.119929. International Transport Forum/OECD (2023) Measuring Accessibility: Methods and Issues. Available at https://www.itf- oecd.org/measuring-accessibility. Kitchin R and Moore-Cherry N (2021) Fragmented governance, the urban data ecosystem and smart city-regions: The case of Metropolitan Boston. Regional Studies 55(12), 1913–1923. https://doi.org/10.1080/00343404.2020.1735627. KolandaisamiG (2020)Drivers andBarriers to theAdoption of the Smart City Paradigm inDevelopingCountries: A SouthAfrican Perspective. Doctoral dissertation, University of Pretoria). Available at https://repository.up.ac.za/bitstream/handle/2263/79592/ Kolandaisami_Drivers_2020.pdf?sequence=1. Le Galès P and Robinson J (2024) Introduction: Comparative global urban studies in the making: Welcome to the world of imperfect and innovative urban comparisons. In The Routledge Handbook of Comparative Global Urban Studies. Routledge, pp. 1–26. Liu X, Fatoyinbo TE, Thomas NM, Guan WW, Zhan Y, Mondal P, Lagomasino D, Simard M, Trettin CC, Deo R and Barenblitt A (2021) Large-scale high-resolution coastal mangrove forests mapping across West Africa with machine learning ensemble and satellite big data. Frontiers in Earth Science 8, 560933. https://doi.org/10.3389/feart.2020.560933. Liva G,Micheli M, Schade S, Kotsev A, Gori M and Codagnone C (2023) City data ecosystems between theory and practice: A qualitative exploratory study in seven European cities. Data & Policy 5, e17. https://doi.org/10.1017/dap.2023.13. e13-10 Ernest Agyemang et al. https://doi.org/10.1017/dap.2024.11 Published online by Cambridge University Press https://au.int/en/pressreleases/20190416/smart-infrastructure-accelerates-regional-integration-say-experts-african https://au.int/en/pressreleases/20190416/smart-infrastructure-accelerates-regional-integration-say-experts-african https://au.int/agenda2063/aspirations https://doi.org/10.1051/e3sconf/202341802001 https://doi.org/10.1051/e3sconf/202341802001 https://doi.org/10.1016/j.cities.2019.01.032 https://doi.org/10.1016/j.cities.2019.01.032 https://doi.org/10.1007/978-3-030-38948-2_178-1 https://doi.org/10.1080/13563475.2022.2137112 https://doi.org/10.1186/s42162-020-00108-6 https://doi.org/10.1016/j.urbmob.2023.100048 https://journals.openedition.org/factsreports/6920 https://core.ac.uk/download/pdf/199447703.pdf https://doi.org/10.1016/j.giq.2016.01.003 https://link.springer.com/chapter/10.1007/978-3-030-12036-8_10 https://link.springer.com/chapter/10.1007/978-3-030-12036-8_10 https://doi.org/10.1126/science.adi6636 https://data.jrc.ec.europa.eu/dataset/53473144-b88c-44bc-b4a3-4583ed1f547e https://data.jrc.ec.europa.eu/dataset/53473144-b88c-44bc-b4a3-4583ed1f547e https://doi.org/10.1038/s41467-020-15788-7 https://www.e3s-conferences.org/articles/e3sconf/pdf/2023/55/e3sconf_acc2023_03004.pdf https://www.e3s-conferences.org/articles/e3sconf/pdf/2023/55/e3sconf_acc2023_03004.pdf https://doi.org/10.1016/j.techfore.2020.119929 https://www.itf-oecd.org/measuring-accessibility https://www.itf-oecd.org/measuring-accessibility https://doi.org/10.1080/00343404.2020.1735627 https://repository.up.ac.za/bitstream/handle/2263/79592/Kolandaisami_Drivers_2020.pdf?sequence=1 https://repository.up.ac.za/bitstream/handle/2263/79592/Kolandaisami_Drivers_2020.pdf?sequence=1 https://doi.org/10.3389/feart.2020.560933 https://doi.org/10.1017/dap.2023.13 https://doi.org/10.1017/dap.2024.11 Maphangwa M and van der Waldt G (2023) Towards a smart city model for local government: The case of South African metropolitan municipalities. Administratio Publica 31, 61–84. Available at https://journals.co.za/doi/epdf/10.10520/ejc-admin pub_v31_n3_a6. Marconcini M,Metz-Marconcini A, Üreyen S, et al. (2020) Outlining where humans live, the world settlement footprint 2015. Scientific Data 7(242). https://doi.org/10.1038/s41597-020-00580-5. Nikiforova A (2020) Definition and evaluation of data quality: User-oriented data object-driven approach to data quality assessment. Baltic Journal of Modern Computing 8(3), 391–432. https://doi.org/10.22364/bjmc.2020.8.3.02. NikiforovaA, FloresMAA and LytrasMD (2023) The role of open data in transforming the society to society 5.0: A resource or a tool for SDG-compliant smart living?. In Smart Cities and Digital Transformation: Empowering Communities, Limitless Innovation, Sustainable Development and the Next Generation. Emerald Publishing Limited, pp. 219–252. Available at https://arxiv.org/ftp/arxiv/papers/2206/2206.11784.pdf. Nilssen M and Hanssen GS (2022) Institutional innovation for more involving urban transformations: Comparing Danish and Dutch experiences. Cities 131, 103845. https://doi.org/10.1016/j.cities.2022.103845. OECD (2021) Transport Strategies for Net-Zero Systems by Design. Paris: OECD Publishing. https://doi.org/10.1787/0a20f779-en. OECD/SWAC (2020) Africapolis database. Available at www.africapolis.org (accessed 5 October 2023). Oksman Vand RaunioM (2018) Citizen-centric smart city planning for Africa: A qualitative case study of early stage co-creation of a Namibian smart community. In Twelfth International Conference on Digital Society and e-governments. IARIA, pp. 30–35. Available at https://api.semanticscholar.org/CorpusID:70340824. Pajares E,Büttner B, Jehle U,Nichols A andWulfhorst G (2021) Accessibility by proximity: Addressing the lack of interactive accessibility instruments for active mobility. Journal of Transport Geography 93, 103080. https://doi.org/10.1016/j.jtran- geo.2021.103080. Peprah C, Amponsah O and Oduro C (2019) A system view of smart mobility and its implications for Ghanaian cities. Sustainable Cities and Society 44 739–747. https://doi.org/10.1016/j.scs.2018.10.025. Pereira G, Eibl G, Stylianou C, Martínez G, Neophytou H and Parycek P (2018) The role of smart technologies to support citizen engagement and decision making: The SmartGov case. International Journal of Electronic Government Research 14(4), 1–17. https://doi.org/10.1145/3184558.3191517. Pesaresi M and Politis P (2023) GHS-BUILT-S R2023A - GHS built-up surface grid, derived from Sentinel2 composite and Landsat, multitemporal (1975-2030) European Commission, Joint Research Centre (JRC). Available at http://data.europa.eu/ 89h/9f06f36f-4b11-47ec-abb0-4f8b7b1d72ea. Plantinga P (2022) Digital discretion and public administration in Africa: Implications for the use of artificial intelligence. Information Development. https://doi.org/10.1177/026666692211175. Ranchod R (2020) The data-technology nexus in south African secondary cities: The challenges to smart governance. Urban Studies 57(16), 3281–3298. https://doi.org/10.1177/0042098019896974. Ricker B, Cinnamon J and Dierwechter Y (2020) When open data and data activism meet: An analysis of civic participation in Cape Town, South Africa. The Canadian Geographer/Le Géographies canadiennes 64(3), 359–373. https://doi.org/10.1111/ cag.12608. Rula A,Maurino A and Batini C (2016) Data quality issues in linked open data. In Data and Information Quality. Data-Centric Systems and Applications. Cham: Springer. https://doi.org/10.1007/978-3-319-24106-7_4. SchiavinaM,Moreno-Monroy A,Maffenini L and Veneri P (2019) GHS-FUA R2019A –GHS functional urban areas. Derived from GHS-UCDB R2019A, (2015), R2019A. European Commission, Joint Research Centre (JRC). Available at https://data. jrc.ec.europa.eu/dataset/347f0337-f2da-4592-87b3-e25975ec2c95. Shi K, Wu Y, Liu S, Chen Z, Huang C and Cui Y (2023) Mapping and evaluating globalurban entities(2000–2020): A novel perspective to delineate urban entities based on consistent night time light data. GIScience & Remote Sensing 60(1), 2161199. https://doi.org/10.1080/15481603.2022.2161199. Siba E and Sow M (2017) Smart City Initiatives in Africa. The Brookings Institution. Available at https://www.brookings.edu/ articles/smart-city-initiatives-in-africa/. Slavova M and Okwechime E (2016) African smart cities strategies for agenda 2063. Africa Journal of Management 2(2), 210–229. https://doi.org/10.1080/23322373.2016.1175266. TorchianoM,Vetrò A and Iuliano F (2017) Preserving the benefits of Open Government Data by measuring and improving their quality: An empirical study. In 2017 IEEE 41st Annual Computer Software and Applications Conference (COMPSAC), Vol. 1. IEEE Xplore, pp. 144–153. https://ieeexplore.ieee.org/document/8029602 UN-Habitat (2017) Smart City Rwanda Master Plan Version 2.0 Available at https://unhabitat.org/sites/default/files/documents/ 2019-05/rwanda_smart_city-master_plan.pdf. Vasileva R (2022) Exploring the Use of Open Data to Support Citizen-Led Initiatives for Sustainable Urban Development in East Africa (Doctoral dissertation, University of Nottingham). Available at https://eprints.nottingham.ac.uk/67631/1/FINAL_Roza% 20Vasileva_PhD_thesis_Jan%2022%202022.pdf. Vetrò A, Canova L, TorchianoM,Minotas CO, Iemma R andMorando F (2016) Open data quality measurement framework: Definition and application to open government data. Government Information Quarterly 33(2), 325–337. https://doi. org/10.1016/j.giq.2016.02.001. Data & Policy e13-11 https://doi.org/10.1017/dap.2024.11 Published online by Cambridge University Press https://journals.co.za/doi/epdf/10.10520/ejc-adminpub_v31_n3_a6 https://journals.co.za/doi/epdf/10.10520/ejc-adminpub_v31_n3_a6 https://doi.org/10.1038/s41597-020-00580-5 https://doi.org/10.22364/bjmc.2020.8.3.02 https://arxiv.org/ftp/arxiv/papers/2206/2206.11784.pdf https://doi.org/10.1016/j.cities.2022.103845 https://doi.org/10.1787/0a20f779-en http://www.africapolis.org https://api.semanticscholar.org/CorpusID:70340824 https://doi.org/10.1016/j.jtrangeo.2021.103080 https://doi.org/10.1016/j.jtrangeo.2021.103080 https://doi.org/10.1016/j.scs.2018.10.025 https://doi.org/10.1145/3184558.3191517 http://data.europa.eu/89h/9f06f36f-4b11-47ec-abb0-4f8b7b1d72ea http://data.europa.eu/89h/9f06f36f-4b11-47ec-abb0-4f8b7b1d72ea https://doi.org/10.1177/026666692211175 https://doi.org/10.1177/0042098019896974 https://doi.org/10.1111/cag.12608 https://doi.org/10.1111/cag.12608 https://doi.org/10.1007/978-3-319-24106-7_4 https://data.jrc.ec.europa.eu/dataset/347f0337-f2da-4592-87b3-e25975ec2c95 https://data.jrc.ec.europa.eu/dataset/347f0337-f2da-4592-87b3-e25975ec2c95 https://doi.org/10.1080/15481603.2022.2161199 https://www.brookings.edu/articles/smart-city-initiatives-in-africa/ https://www.brookings.edu/articles/smart-city-initiatives-in-africa/ https://doi.org/10.1080/23322373.2016.1175266 https://ieeexplore.ieee.org/document/8029602 https://unhabitat.org/sites/default/files/documents/2019-05/rwanda_smart_city-master_plan.pdf https://unhabitat.org/sites/default/files/documents/2019-05/rwanda_smart_city-master_plan.pdf https://eprints.nottingham.ac.uk/67631/1/FINAL_Roza%20Vasileva_PhD_thesis_Jan%2022%202022.pdf https://eprints.nottingham.ac.uk/67631/1/FINAL_Roza%20Vasileva_PhD_thesis_Jan%2022%202022.pdf https://doi.org/10.1016/j.giq.2016.02.001 https://doi.org/10.1016/j.giq.2016.02.001 https://doi.org/10.1017/dap.2024.11 VichG,Gómez-Varo I andMarquetO (2023)Measuring the 15-minute city in Barcelona. A geospatial three-method comparison. In Resilient and Sustainable Cities: Research, Policy and Practice, pp. 39–60. https://doi.org/10.1016/B978-0-323-91718- 6.00004-9. Willberg E, Fink C and Toivonen T (2023) The 15-minute city for all?–Measuring individual and temporal variations in walking accessibility. Journal of Transport Geography 106, 103521. https://doi.org/10.1016/j.jtrangeo.2022.103521. Yan D,Wang C, Feng X and Dong B (2018) Validation and ground truths. InWagner A, O’BrienWand Dong B (eds.), Exploring Occupant Behavior in Buildings. Cham: Springer. https://doi.org/10.1007/978-3-319-61464-9_9. Zhao M, Cheng C, Zhou Y, Li X, Shen S and Song C (2022) A global dataset of annual urban extents (1992–2020) from harmonized nighttime lights. Earth System Science Data 14(2), 517–534. https://doi.org/10.5194/essd-14-517-2022. Cite this article:Agyemang E, Anderson B, Patiño J and TremolieresM (2024). Toward achieving smart cities inAfrica: challenges to data use and the way forward. Data & Policy, 6: e13. doi:10.1017/dap.2024.11 e13-12 Ernest Agyemang et al. https://doi.org/10.1017/dap.2024.11 Published online by Cambridge University Press https://doi.org/10.1016/B978-0-323-91718-6.00004-9 https://doi.org/10.1016/B978-0-323-91718-6.00004-9 https://doi.org/10.1016/j.jtrangeo.2022.103521 https://doi.org/10.1007/978-3-319-61464-9_9 https://doi.org/10.5194/essd-14-517-2022 https://doi.org/10.1017/dap.2024.11 https://doi.org/10.1017/dap.2024.11 Toward achieving smart cities in Africa: challenges to data use and the way forward Policy Significance Statement 1. Introduction 2. Key considerations on using data moving forward 2.1. Data source selection 2.2. Data imperfections: Integration 2.3. Temporal and financial resources 2.4. Ground truth validation 2.5. Data market: Absence of public sector 3. Policy implications: Points of action 4. Conclusion Author contribution Funding statement Competing interest References