research-article

Open access

Governing Smart City IoT Interventions: A Complex Adaptive Systems Perspective

Authors:

Vinay Reddy Venumuddala,

Amit Prakash,

Bidisha ChaudhuriAuthors Info & Claims

Digital Government: Research and Practice, Volume 5, Issue 3

Article No.: 23, Pages 1 - 24

https://doi.org/10.1145/3660644

Published: 13 September 2024 Publication History

PDF eReader

Abstract

Urban infrastructure systems (UIS) encompass physical infrastructure and human stakeholders, delivering essential services through distributed flow networks. These systems exhibit properties of Complex Adaptive Systems (CAS), such as self-organization and adaptivity, which makes them resilient to perturbations from the outside. Smart city strategies advocate for external technology interventions to improve the efficiency and sustainability of UIS. Internet of Things (IoT) is a key enabler for such strategies that helps to gather data from various infrastructure components and optimize flow networks in UIS. Although beneficial, such interventions often give rise to ethical concerns, underscoring the need for appropriate governance measures. In this study, we adopt a CAS perspective to analyse the ethical concerns of a proposed smart city IoT platform in the cities of Karnataka, India. Based on our analysis, we present a framework to aid city authorities in formulating appropriate strategies to govern Smart City IoT interventions in general.

1 Introduction

Urban infrastructure systems (UIS) encompass physical infrastructure and human stakeholders, enabling the delivery of vital urban services like water, energy, transportation, waste management, and more to citizens through distributed flow networks [77].

Extant literature widely acknowledges that UIS displays characteristics typical of complex adaptive systems (CAS) [4, 57]. Self-organisation and adaptivity are some of the defining properties of any complex adaptive system. Self-organization signifies that agents or subsystems within such a system can act autonomously and organise from the bottom up without the need for any central control. Adaptivity, as the term suggests, refers to the system's capacity both at the level of the agents and as a whole, to autonomously adjust to changes in the external environment. These properties of CAS make them resilient, robust, fault-tolerant, and fail-safe in response to changes imposed by the external environment [33]. UIS are also increasingly seen as complex adaptive systems “structured from the bottom up” [4], as evident in examples such as daily traffic patterns, housing segregation, and industrial agglomeration, which are emergent from the spontaneous self-organisation of consumer groups linked to infrastructure systems, such as transportation and buildings [57].

Agents, interactions, and the environment are three components that are considered as crucial to understand the structure of any CAS [54]. Agents are the “basic entities of actions,” which include humans, organisations, and technical objects; interactions define the relations between agents in terms of the nature of resources or information exchanged; environment defines the medium where these agents interact. Suppliers and their operations personnel, consumers, and governments are the major agents whose behaviour and interactions are of interest in understanding any UIS [77].

Smart city interventions leverage data-driven technologies to enhance the efficiency and sustainability of individual UIS and also envisage applications that cut across UIS [47]. Internet of Things (IoT) is a key enabler for smart city applications. Any IoT intervention is made up of physical devices (sensors and actuators), communication networks, platforms hosting application-enabling services, and applications [38]. IoT platforms simplify the development of smart city applications by abstracting the complexities of data generation from diverse infrastructure systems.

Smart city IoT interventions in UIS primarily focus on optimizing the respective flow networks [47, 77]. Suppliers use IoT devices to collect real-time data from various social and technical components of the UIS, solve network optimization challenges, and facilitate automated control of such components. The transition of UIS toward smart UIS therefore envisages transformation of their respective flow networks to become more dynamic and reduce manual operations. Apart from suppliers, consumers are expected to benefit from improved information about their consumption patterns and costs, and governments can benefit from real-time data from various UIS to shape policies to promote sustainability in infrastructure systems [6, 77].

Although smart city interventions are expected to fulfil efficiency and sustainability goals, they are also known to pose potential ethical concerns that need appropriate governance measures. These interventions may often ignore the persisting socio-economic, spatial, and digital divides within the city and between its citizens and raise concerns about justice and fairness in decision-making [36, 42]. When logics of centralised and automated control exercised through these interventions fail to account for the rights of citizens toward privacy, safety, and security, they pose threat to citizen trust in UIS [18, 43, 44]. The data-driven nature of such interventions may also affect the autonomy of operations personnel and expose them to precarity of data work potentially affecting their dignity of labour [7, 29, 79].

Our study is motivated by the need to have effective governance measures to tackle ethical concerns associated with smart city IoT interventions in UIS. To this end, we analyse an empirical case of a Smart City IoT platform proposed for deployment across all the designated smart cities in the state of Karnataka, India, and how it proposes to integrate IoT interventions of two urban infrastructure systems—solid waste management and transportation management. We rely on semi-structured interviews of relevant stakeholders from the government, industry, and civil society, and review of official smart city project documents from three designated smart cities of Karnataka to build this case.

We adopt a CAS perspective to explore the impact of these interventions on the role of different agents within the concerned urban infrastructure systems, and on the interactions between such agents. Using this perspective, we also foreground some of the potential ethical concerns that are important to resolve from a governance point of view. We restrict ourselves to an a priori chosen set of concerns related to justice, fairness, trust, and dignity. Drawing from this analysis, we then construct a structured set of inquiries intended to assist city authorities in formulating appropriate governance strategies for Smart City IoT interventions.

Our study represents an early endeavour to utilize a complexity lens in the development of an ethics-centred governance framework. Emerging literature underscores the limitations of traditional ethics frameworks in aligning with the challenges of today's rapidly evolving socio-technical landscape. Some limitations include overlooking the contextual nature of ethical issues, prioritizing metrics, or adopting a narrow, bottom-up focus on specific contexts at the expense of broader values, thus risking short-sightedness [32, 41, 75]. With emerging technologies introducing a plethora of new ethical concerns, governance frameworks rooted in the complexity worldview can be more effective than traditional approaches in navigating contemporary challenges [75].

2 Theoretical Framework

2.1 Complex Adaptive Systems

Any CAS is recognized by three core components: agents, interactions, and environment [54]. Agents include individuals, groups, technology components, or even the subsystems made up of interaction between these entities. Agents are described by their attributes and behaviour rules, which together determine how they interact with other agents and with the environment. Interactions are described by the connections, or channels for interaction, and the flows of physical or informational resources between agents [35, 54]. Environment is the medium where agents interact. It is defined by a structure, for instance, in terms of the distribution of resources across different sites in the environment. This structure influences the actions of different agents, and the ongoing interactions of agents can also modify it [21, 54].

2.1.1 Properties of CAS.

Complex adaptive systems often evolve through self-organisation and adaptation rather than central control [71]. Self-organisation implies that the system autonomously arranges its agents and interactions into a global organisation to maximise its fitness within the environment without the need for an external designer or controller [33]. This global organisation showcases emergent properties that cannot be reduced to the properties of interacting agents but might constrain the latter's behaviour [40, 76]. The responsibility of maintaining such an organisation is distributed among agents so that it is difficult for any one agent to deviate from or unduly influence it [33].

A self-organising system becomes relatively self-sufficient and closed from the influence of external environment, and maintains its overall organisation despite minor perturbations. More generally, such a system settles down into self-sufficient subsystems, whose interactions then settle into self-sufficient subsystems at a higher level, and so on; eventually forming nested hierarchies of systems within systems [31, 33]. Therefore, each agent within a complex adaptive system can also be seen as a subsystem within the focal system functioning autonomously to maintain its own local organisation [10].

Adaptivity is the property of a system to achieve a fit within its environment. Fitness here refers to the property of the system to grow or maintain its organisation without disintegrating under the given environmental conditions [33]. The properties of self-organisation and adaptivity exhibited by complex systems are closely related. Self-organisation implies adaptivity, because agents within self-organising systems mutually adapt with each other. The property of adaptivity becomes distinguishable from self-organisation when we choose a specific boundary to differentiate between the system (or subsystem) and its environment [33].

Self-organisation and adaptivity make CAS robust or resilient than consciously designed systems as the intrinsic uncertainties of the environment facing agents, forces them to have sufficient redundancy and diversity [34]. Redundancy refers to the property that often more than one agent fulfils the same or a similar function, while diversity refers to differences in attributes of agents interacting within the environment [60]. Since agents in self-organising systems share the responsibility of maintaining the overall organisation, redundant agents help to restore such organisation when other similar agents fail to function. Diversity, however, ensures that the system and its agents carry a sufficient variety of actions and capacities so that at least some of them will be able to neutralise perturbations from the environment while maintaining or growing the organisation. Those agents capable of responding to the environment first adapt and trigger the adaptation of other interacting agents, and so on, until the system restores its original organisation or grows into a new stable configuration. When the perturbation is significant, it leads to a drastic disruption of existing organisation forcing the system to self-organise from the start [33].

These above discussed properties make CAS robust or resilient to perturbations, and capable of restoring themselves or failing gracefully in the event of faults. However, any interventions that may cause drastic disruptions may deplete key resources of the system and take it to an impoverished state from which it could become impossible to revert [11, 56].

2.2 Urban Infrastructure Systems as CAS

UIS consist of physical infrastructure and people associated with systems related to water supply and distribution, energy management, transportation management, building infrastructure management, waste management, and so on. They facilitate the provision of respective essential services such as water, electricity, solid waste disposal, mobility, and so on, to the citizens through the corresponding flow or distribution networks [16]. These networks are made up of social and technical components that support the flow of various resources—like electricity, water, waste, and so on—or people such as in the case of transport [77].

Suppliers and their operations personnel, consumers, and governments are the major human or organisational stakeholders who determine the future of any UIS. Suppliers could be purely government agencies, public-private partnerships, or purely private entities. They are responsible for the operation and management of flow networks of an infrastructure system and for the provision of associated services to the citizens. They could either operate the infrastructure systems exclusively or multiple suppliers could compete in providing the associated services. Some of the main objectives of suppliers include improvement of the flow network efficiency and ensure the long-term sustainability of respective infrastructure systems [77].

Operations personnel of the suppliers are vital to conduct the day-to-day operations of any infrastructure system, address consumer issues related to service delivery, and provide timely feedback to the system administrators [68]. Most existing infrastructure systems are centralised, with ownership and decision-making authority regarding both the system's management and the delivery of its services vested in central entities, such as utilities in the case of traditional electricity distribution [1].

Consumers often bear the cost of services delivered by the suppliers, and therefore seek to minimise costs while maximising their utility. They differ in various aspects. For example, they may differ by type such as residential, commercial or industrial consumers; by service consumption characteristics such as their capacity to avail services that in turn may depend on their income or revenue; or by their responsiveness to changes in supplier policies such as pricing of services [27]. Supplier policies are one among the factors that influence the state of the environment where consumers act.

When both suppliers and consumers work for their self-interest, it often impacts the long-term sustainability of infrastructure systems especially if seen in terms of serving the society's future interests. The role of governments is vital here as they need to institute relevant policy measures to change the state of the environment for suppliers and consumers so that they channelize their actions toward sustainable and socially desirable outcomes [77].

UIS are traditionally viewed as being “centrally organised from the top-down,” but increasingly they are being seen as CAS systems that are “structured from the bottom-up” [4]. Nel et al. [57] discuss some examples highlighting the CAS properties of UIS. Examples like daily traffic patterns, housing segregation, and industrial agglomeration demonstrate emergent behaviours arising from the spontaneous self-organisation of consumer groups linked to respective UIS, such as transportation and buildings. Urban adaptation is also evident in cities as seen in the form of shifts in land and building usage with change in policies, shift in mobility preferences of consumers with the introduction of novel transportation methods and networks, and so on. Diversity is evident, for example, the diverse forms that UIS pertaining to land and building usage take, to cater to a diversity of population groups and urban morphologies. Redundancy is evident, for example, in transportation networks that offer alternative routes when roads are blocked or during congested periods when individuals can opt for public transport. Both diversity and redundancy strengthen the self-organising and adaptive capacities of UIS [56, 57].

2.2.1 Smart Cities: IoT-based Transition of UIS.

Under smart city initiatives, a widely adopted strategy is to utilise data-driven technologies to improve the efficiency of individual infrastructure systems and also interconnect multiple infrastructure systems to leverage applications at their intersection [47, 63]. For example, applications to improve the efficiency of electric mobility charging infrastructure in a city require interconnection of both the transportation and energy infrastructure systems [47].

The IoT is a fundamental driver of smart UIS. According to the IEEE standard 2413, any IoT intervention is made up of (a) physical layer: devices including sensors and actuators that augment the functions, properties, and information exchange capabilities of various infrastructure components, (b) network layer: communication networks that interconnect the devices of physical layer, (c) platform layer: application enabling services facilitating the discovery, composition, and orchestration of IoT devices and facilitating data management and analytics, and (d) application layer: applications built on top of the IoT platform. IoT platforms enable development of smart city applications by decoupling developers from the intricacies of data generating processes across different infrastructure systems [38].

An IoT-driven transition primarily targets the optimization or efficiency improvement of flow networks in UIS. Suppliers associated with the respective UIS rely on IoT devices to gather real-time information from different components and facilitate their remote, automated, and real-time control. This facility allows them to optimize the respective flow networks in UIS and transform them into dynamic and active networks that minimize the need for manual interventions by operations personnel [77]. Nevertheless, effectiveness of such a transition hinges upon the responses of different consumers to the dynamic supplier policies enabled by these interventions, and also the consumer capabilities in terms of making additional investments at their level such as smart metres, data management systems, and so on [1, 77]. Table 1 provides an example of an IoT-driven transition from a traditional electricity distribution grid to a smart grid.

Table 1.

Smart Electricity Infrastructure
For example, in the context of an IoT-driven transition from a traditional electricity distribution grid to a smart grid, utilities aim to improve the efficiency of their distribution networks by relying on various IoT components. Such components may include smart metres, home energy management systems, data collection platforms, and utility level energy management systems [48]. Smart metres at households provide energy consumption data to utilities for billing and offer insights into power quality, voltage, and load profiles. Home energy management systems (HEMS) are IoT platforms that help in gathering and analysing information about household level energy consumption. They empower consumers to effectively manage the trade-offs between costs of purchasing energy services from the utility vs their own captive generation (like installing rooftop solar), if any. The cost of smart metres and HEMS are generally borne by the consumers. HEMS can also allow utilities to directly control specific consumer loads and use pricing strategies to adjust their consumption to meet the overall efficiency of the distribution network (e.g., reducing peak power consumption expectations from the distribution grid). Design of applications to fulfil such control strategies are facilitated by the utility level data collection platforms and energy management systems that integrate data from edge components and enable advanced analytics. In addition to enhancing the efficiency and sustainability of energy infrastructure amid growing population demands, such IoT interventions also eliminate the need for operations personnel tasked with manual metre reading, outage reporting, and restoration [48].

Smart Electricity Infrastructure

For example, in the context of an IoT-driven transition from a traditional electricity distribution grid to a smart grid, utilities aim to improve the efficiency of their distribution networks by relying on various IoT components. Such components may include smart metres, home energy management systems, data collection platforms, and utility level energy management systems [48]. Smart metres at households provide energy consumption data to utilities for billing and offer insights into power quality, voltage, and load profiles. Home energy management systems (HEMS) are IoT platforms that help in gathering and analysing information about household level energy consumption. They empower consumers to effectively manage the trade-offs between costs of purchasing energy services from the utility vs their own captive generation (like installing rooftop solar), if any. The cost of smart metres and HEMS are generally borne by the consumers.
HEMS can also allow utilities to directly control specific consumer loads and use pricing strategies to adjust their consumption to meet the overall efficiency of the distribution network (e.g., reducing peak power consumption expectations from the distribution grid). Design of applications to fulfil such control strategies are facilitated by the utility level data collection platforms and energy management systems that integrate data from edge components and enable advanced analytics. In addition to enhancing the efficiency and sustainability of energy infrastructure amid growing population demands, such IoT interventions also eliminate the need for operations personnel tasked with manual metre reading, outage reporting, and restoration [48].

Table 1. Example of Smart Electricity Grid as an IoT-driven Transition

In essence, some of the stated advantages of popular smart city IoT interventions in UIS are as follows. Consumers now get better information about their own consumption patterns and the associated costs. Based on the observed cost structure, they can choose between multiple suppliers in purchasing services, fulfil such services on their own (e.g., installation of roof-top solar for energy, or own personal vehicles for transport), or avail services from other prosumers (consumers who also produce) in the local markets [1]. Suppliers can gather real-time information about consumption and optimise their distribution networks to efficiently match demand and supply of services. Governments can leverage this data to craft enforcement policies that encourage both suppliers and consumers to transition toward more sustainable infrastructure systems [77].

In this study, we adopt a CAS perspective to identify the potential impact from a Smart City IoT platform proposed for deployment in all the seven designated smart cities in the state of Karnataka, India. These cities include Bengaluru, Belgavi, Davanagere, Hubli-Dharwad, Mangaluru, Shivamogga, and Tumakuru. While this intervention, in each city, proposes to integrate multiple UIS, we focus on two—solid waste management and transportation. We explore the impact of this intervention on the role of different agents within the concerned UIS, and on the interactions between them. Through this perspective, we also foreground some of its potential ethical concerns as expressed by the interview respondents, which are important to resolve from a governance point of view. We restrict ourselves to an a priori chosen set of concerns related to justice, fairness, trust, and dignity. Drawing from this analysis, we then attempt to construct a structured set of inquiries intended to assist city authorities in formulating appropriate governance strategies for smart city IoT interventions. In the following section, we provide a brief background to highlight the significance of ethics-based governance of smart city IoT interventions.

2.3 Ethics-based Governance of Smart UIS

The words governance and government have common root in the Greek verb “kubernan,” which means to steer a ship. Governance refers to the “manner in which power is exercised in the management of a country's economic and social resources for development” [3]. Government, however, commonly refers to the state that exercises such power [62]. The concept of governance has evolved from this notion of government or the state as being the sole controller to the management of public affairs by all the actors in a society [9].

In the context of governance, the state itself is now seen as a collection of “interorganizational networks made up of governmental and societal actors with no sovereign actor able to steer or regulate” [65]. In this view, the government, which refers to the ruling body of the state with its executive and legislative apparatus, is a principal actor in governance. Additionally, it is also seen as the creator of an environment that enables other actors such as the market and the civil society to participate in the process of governance [9].

Stoker [69] provides two perspectives on governance—managerial and systemic. Focus of the former is on the emerging new processes of governing primarily for the government. Such processes, for example, can be seen in terms of entrepreneurial governments giving primacy to competition, markets, customers, and outcomes [37, 64]. In the systemic view the focus is “less on the changing tools of government and more on the emergence of a system of self-governing networks” [69]. In this view governance refers to the set of regulations emerging from an interactive relationship between governmental and non-governmental participants that work better than governments acting alone to effect desired changes [9, 50, 69].

2.3.1 Governance of IoT Interventions.

An understanding about the governance of IoT interventions builds from an understanding about governance of the Internet and the governance of Information Technology (IT) systems. De Bossey [14] defines internet governance as “the development and application by governments, the private sector, and civil society, in their respective roles, of shared principles, norms, rules, decision-making procedures, and programs that shape the evolution and use of the internet.” Similarly, governance in IT systems indicates the outlining of appropriate roles and responsibilities for various stakeholders for effective utilisation of IT infrastructure meeting a set of high-level objectives [15].

IoT interventions often reside in system-of-systems environments. For instance, in the case of smart cities, IoT platforms enable integration of multiple infrastructure systems to create a system-of-systems environment. This makes governance of IoT interventions much more challenging relative to the internet or IT, as different components in an IoT intervention may be owned, managed, or operated by different entities with diverse incentives, responsibilities, and obligations [67]. IoT governance therefore calls for a greater involvement of other governing bodies, such as the private sector and civil society, in addition to the government, to devise effective regulatory measures [78]. As per Singh et al. [67], these regulatory measures can be broadly thought along two dimensions. First is the legal dimension, which focuses on the appropriate laws that can regulate any general IoT interventions. Second is the technical dimension where technical means (for example, enforcing IoT architectures that complies with ethics-by-design principles) can assist IoT interventions to better align with the legal and regulatory requirements.

2.3.2 Ethics Informed Governance Objectives.

Governance refers to the regulatory actions emanating from an interactive relationship between governmental and non-governmental participants to achieve desired objectives from a given system [9, 50, 69]. Desired objectives from the governance of smart city IoT interventions, for instance, could be efficiency and sustainability of urban infrastructure systems as we introduced earlier. However, in addition to efficiency and sustainability, ethical compliance is also a crucial objective in the context of governing IoT-based interventions. The significance of governance measures to regulate IoT interventions against ethical concerns such as trust, privacy, safety, and so on, are widely acknowledged in the extant literature [39, 67]. For the purpose of our study, we stick to four ethical dimensions - justice, trust, fairness, and dignity. Below, we briefly define and contextualise the said ethical dimensions in relation to the governance of smart (or IoT-enhanced) UIS.

Justice: In the context of smart UIS, questions of justice arise due to the impact of IoT interventions on the society's social structure. From the consumers’ perspective, the rapid proliferation of data-centric solutions can widen the digital divide, particularly for marginalised groups based on factors like gender, geography, income, and more. For instance, in the context of transportation, research highlights the challenges faced by lower-income groups who may have limited access to smart public transportation systems due to constraints such as the inability to afford smartphones and internet expenses [30]. Furthermore, data-driven interventions might improve infrastructure efficiency, but they often also replace political negotiation mechanisms and thereby reinforce inequalities [61]. To address potential inequalities, it is crucial to consider individual needs and rights, including access to services, representation, the freedom to make technology choices based on one's capabilities, and the preservation of informational privacy within the new UIS [72].

In the context of justice, the egalitarian norms should be applied, not only to consumers but also to the suppliers. Inequity within the latter often manifests as a result of the central role played by data in the success of IoT interventions. While intended use-cases should guide data collection and analysis, the availability of data tools has led to short-term, profit-driven use-cases [70]. This commodification of data raises concerns about equity, with large corporations gaining a competitive advantage due to their access to crucial data, financial resources, and technological capabilities [17]. Tight budgets often lead city authorities to rely on large private companies, potentially limiting the scope for local enterprises to offer tailored solutions [8, 66]. This situation risks reducing opportunities for small businesses that traditionally address both expressed and unexpressed local needs [51, 61].

In essence, justice means equitable access for consumers to the services offered by an UIS regardless of socio-economic or digital disparities, and equal opportunities for the diversity of suppliers, big and small, in urban infrastructure development.

Trust: In the context of IoT interventions trust is critical between consumers and IoT devices/applications [39]. Consumers’ trust therefore becomes critical even in the context of smart UIS. Privacy, safety, and security mechanisms are vital for building this trust [13, 19]. Privacy concerns arise due to the vast amounts of personal information collected by interconnected IoT components, potentially limiting consumers’ control over data sharing and processing [2, 19]. Like in traditional software solutions, safety and security are also essential in IoT solutions requiring continual upgrades and investments in terms of necessary human capacities, technologies, and workflows. Nevertheless, they are frequently disregarded due to cost and scalability concerns, as investing in these aspects are seen to increase expenses and hinder interoperability between components, constraining their widespread adoption [2].

Given the scale of IoT interventions in terms of their vast number of components, interconnections, and actors involved, data privacy and security breaches can also lead to physical safety and security concerns for the consumers [67]. Identifying responsible parties for breaches although becomes complex in the case of IoT interventions they become even more essential. Ensuring accountability w.r.t. the IoT intervention is crucial to regain end-users’ trust [13]. This involves mechanisms for reviewing, determining, and establishing liability in case of failures. These mechanisms necessitate transparency and explainability in the decisions driven by IoT interventions, benefiting consumers by making IoT outcomes understandable and empowering them with greater control over their data within the system [59].

In essence, trust involves consumers’ confidence in the delivery of services by UIS, considering factors such as privacy, safety, security, and accountability.

Fairness: Fairness in data-driven technological systems extends beyond assessing outcomes, and encompasses examining the intentions and convictions that underlie a system [22]. When transitioning from less-automated alternatives to new, automation-intensive IoT solutions, a fair assessment must consider whether the new UIS addresses biases or exacerbates them at scale. Biases in terms of the challenges and barriers to adoption for specific sections of consumers, may only emerge after problematic design patterns within the IoT interventions become entrenched [12]. Therefore, it becomes imperative to scrutinise the fairness of the deliberative processes that lead to the choice of IoT interventions, particularly when assumptions about their universal adoption go unquestioned. This examination should involve a thorough assessment of the decision to replace existing alternatives, considering factors such as the justification for asserting the superiority of IoT interventions over current alternatives and whether the valid merits of existing alternatives are being overlooked [61].

Fairness of outcomes involves avoiding prejudices, discrimination, or bias based on sensitive attributes like race, gender, and so on [55]. IoT interventions can inadvertently perpetuate discriminatory biases based on data collected, necessitating mechanisms for individuals and organisations to identify and challenge bias while ensuring freedom from discrimination [72]. Reviewability of the intervention helps consumers to understand data sources, lineage and flow, demonstrates legally appropriate data processing, and supports them in raising any fairness concerns [59].

In essence, fairness entails impartial service delivery outcomes from smart UIS, accommodating consumers from diverse socio-economic backgrounds, and ensuring unbiased intentions behind any proposed interventions.

Dignity: Dignity, at its core, embodies the inherent worth and respect owed to every individual for their intrinsic value, not just their capabilities. As outlined in the Universal Declaration of Human Rights, recognizing, and upholding the dignity of all individuals is the cornerstone of freedom, justice, and global peace [74]. Nevertheless, smart city IoT interventions have the potential to impact the dignity of work, especially for operations personnel. These interventions may improve convenience for some users, but they can also restrict the autonomy of others by forcing them to conform to centralised instructions, and in some cases, the perspectives of the designers of these interventions [39, 73].

Smart UIS can affect the dignity of operations personnel by assigning them less creative and often fragmented tasks, potentially depriving them of fulfilling work. For instance, data-driven applications on IoT platforms, especially those using artificial intelligence, have the potential to automate manual and routine cognitive tasks typically carried out by operations personnel [7, 29]. Operations personnel may also find themselves engaged in producing the data necessary to fuel these applications. It is crucial to acknowledge that this kind of fragmented work is frequently associated with low wages and involves repetitive tasks performed under precarious labour conditions [79]. Unfortunately, when the adoption of these technology interventions is primarily driven by efficiency and productivity goals, such concerns are frequently overlooked. It is essential to recognize that the dignity of operations personnel hinges on viewing income and employment as rights, not acts of charity [5], and therefore it is necessary to have measures in place to improve their working conditions.

In essence, dignity revolves around offering meaningful roles to operations personnel and mitigating the precarious nature of their working conditions within smart UIS.

3 Methodology

This study is part of a larger project funded by the government of Karnataka to arrive at a high-level framework to govern IoT interventions across different application domains. The project broadly stipulates four ethical dimensions to be explored from a governance point of view. These include Justice, Fairness, Trust, and Dignity. Based on a yearlong study organised between May 2022 till May 2023, we attempt to build such a framework for the application domain of Smart Cities, by analysing an IoT intervention—the Integrated Command and Control Centre (ICCC) through a CAS perspective. The proposed IoT intervention has been commonly envisaged for implementation in each of the seven designated smart cities in the state of Karnataka. These include Bengaluru, Belgavi, Davanagere, Hubli-Dharwad, Mangaluru, Shivamogga, and Tumakuru.

Our data collection and analysis progressed through three main phases leading up to an empirical case of a smart city IoT platform—the ICCC. The three phases in data collection include: (1) literature review, (2) interviews of stakeholders from the government, industry, and civil society, engaging with Smart City projects, and (3) follow-up review of official project documents related to Smart City IoT interventions, follow-up interviews of operations personnel, and field observations. Insights from literature review not only directed our data collection strategy but also helped us conduct analysis of the resulting data to present the case of an ICCC. Below, we describe each of these phases.

3.1 Phase 1: Literature Review

In the initial phase of our research, we conducted a comprehensive review encompassing research articles pertaining to complexity of UIS, smart cities, urban governance and IoT governance with a particular emphasis on ethical considerations related to Justice, Trust, Fairness, and Dignity. Most of these articles have been referenced in previous sections. Additionally, we examined technical standards and policy drafts relevant to data-driven technologies, including prominent ones such as IoT standards published by IEEE [38] and the Industrial IoT consortium [46], policy drafts by a European research cluster on IoT governance [39], and policy drafts from India regarding the data protection bill (2019), data-driven mobility [58], and Smart Cities [52]. In the realm of ethics-based governance, we analysed governance frameworks and reports issued by multinational organisations such as UNESCO [73], the European Commission [23–26], and the World Economic Forum [28], as well as technology and consulting companies, such as McKinsey [49] and KPMG [45].

3.2 Phase 2: Interviews

This literature review helped us conduct semi-structured interviews with stakeholders from the government, industry, and civil society working on smart city projects. A total of 13 respondents were interviewed in this stage. They include: (a) four smart city officials from the Karnataka State—one officer responsible for smart city data management, and three officers responsible for selecting vendors to build/procure and operate smart city IoT solutions. These officials were responsible for the procurement and management of data-driven smart city solutions across the seven designated smart cities in the state of Karnataka; (b) four industry members—working on IoT interventions primarily in transportation and logistics. Two of them are senior technical managers in a multi-national company that leverages its proprietary IoT platform to offer solutions to buyers and service-providers in the space of logistics and mobility. Another is a co-founder in a startup that aims to build IoT- and Blockchain-based solutions, in general, and which is currently also exploring possibilities in the space of supply chain logistics. The fourth industry member is a senior director in one of the technology giants that offer data science solutions to several industries such as healthcare, transportation, and business verticals such as operations, marketing, sales, and so on; (c) five members from the civil society, including four field-based researchers from three different educational institutes and one from a non-governmental organisation, all working on projects related to technology interventions and ethics in smart cities.

Of the 13, 2 interviews were telephonic, and the remaining were in-person. For each respondent, the total interview duration spanned between 2 and 4 h, and in most cases, included multiple interactions with them over the one-year duration of this study. The interviews were largely semi-structured and focused on the following aspects of smart city IoT projects: key technology components; stakeholders involved; and potential ethics-based governance challenges related to justice, fairness, trust, and dignity. To inquire about the ethical concerns, we posed the following broad questions to the respondents as presented in Table 2 below.

Table 2.

Ethical dimension	Broad questions
Justice	−What is the lawful and intended purpose of the intervention? −How are the costs and benefits of this intervention distributed among different stakeholders? −Is the system accessible to citizens irrespective of socio-economic or digital divides?
Fairness	−Is the intervention based on a fair assessment of the existing socio-economic and digital divides? −Can the intervention result in any biased outcomes to citizens along socio-economic or digital divides?
Trust	−What are the potential concerns that the intervention may pose w.r.t. privacy, safety, and security? −In what way does the intervention affect existing channels of grievance redressal or accountability available for consumers?
Dignity	−In what way does the intervention affect the role, and autonomy of operations personnel?

Table 2. Broad Set of Questions for Semi-structured Interviews

In addition to these interviews, one of us also participated in three separate day-long workshops dealing with potential IoT interventions in the transportation and telecommunication sectors. All three of them were organised by a leading higher science and technology education institution of Bangalore. The first workshop was attended by 12 members including senior representatives from large software and networking companies, not for profit organisations, research scholars, and entrepreneurs. The second workshop was attended by a total of six members including senior roles from a multinational engineering and technology company and faculty from the institute. With a limited number of participants, these two workshops allowed us to observe viewpoints from different stakeholders more closely. The third workshop was attended by a large audience of over 50 members including senior representatives from the industry, government, and civil society.

3.3 Phase 3: Archival Study and Field Observations

The semi-structured interviews with respondents from the government and the civil society resulted in discussions primarily around smart city IoT interventions within two UIS—solid waste management and transportation. These interviews led us to simultaneously conduct a detailed review of official project documents to gain deeper insights into their architecture, objectives and stakeholders affected by these interventions. We relied on openly available documentation about these interventions mainly from three designated smart cities—Bengaluru, Mangaluru, and Tumakuru. Smart city proposals of these three designated smart cities helped us identify the current and futuristic objectives of these interventions. Detailed plan proposals, official tenders, and project reports, concerning these interventions helped us to understand the architecture and stakeholders affected by these interventions.

While we noted architectural similarities among the interventions in the three cities whose documentation we reviewed, further discussion with smart city officials unveiled that this resemblance extended to the remaining designated smart cities as well. The officials pointed out that the architectural similarity is necessary to effectively integrate them into an overarching ICCC platform within each city. Nevertheless, they said that when it comes to the nature of smart-city applications, which may get hosted onto the ICCC platform, Bangalore will have an advantage compared to other cities owing to its entrepreneurial environment. In Bangalore, the development of data-driven or AI applications over ICCC can be outsourced to startups or established technology vendors to address city-specific UIS problems. But in the other smart cities, readily available or off-the-shelf solutions may have to be relied on.

Our study of smart city project documents in three designated smart cities also clearly indicated that the IoT interventions for both the UIS, namely, solid waste management and transportation, were proposed to be integrated into the ICCC in every city. ICCC is a standard smart city IoT platform that is prescribed for all designated smart cities across the country. To gain a further understanding of ICCC, we reviewed official documents published by the state and the central governments that laid down the objectives, architectural standards, and maturity assessment frameworks for ICCC platforms. Triangulating insights obtained from the interviews, official documents, and literature review, together helped us gather the necessary data that formed the basis to analyse and build a detailed empirical case of ICCC—an IoT platform for smart cities.

The semi-structured interviews, reflections, and review of project documentation provided a top-down perspective on the two UIS in terms of their primary stakeholders (suppliers, government, and consumers) and interactions, and the potential changes imminent from the proposed IoT interventions.However, they only provided a superficial glimpse of the nature of interactions between actors at the last mile within a given UIS, for instance, between the operations personnel and consumers associated with a particular UIS. To gain an understanding of these interactions, we conducted additional interviews with both operations personnel and consumers at the last mile of service delivery. With the help of two additional research assistants, we conducted unstructured interviews of 12 stakeholders at the last mile of service delivery in solid waste management in one of the cities. These included six frontline workers responsible for gathering household waste, and two field supervisors. To understand the nature of interactions of different consumers with these operations personnel, we also interviewed two residents (one from an apartment complex and another from an independent house), two hotel managers (one from a small hotel and other from a large restaurant).

We documented data gathered from interviews, observations, and reflections in the form of field notes. Most of the data that entered field notes were gathered in the first six months (May 2022–Sept 2022) of the project. The field observations of the last-mile of civic service delivery were conducted and recorded between Feb 2023–May 2023.The field notes were jointly reviewed on a weekly basis by the three of us and reflections about ethical concerns corresponding to justice, fairness, trust, and dignity, were also appended to it. The resulting field notes became a dense narrative of 36 pages, single spaced with 10-point font in Microsoft Word (roughly 26,500 words). It primarily constituted of (a) transcripts of semi-structured interviews with smart city officials, stakeholders at the last mile of civic service delivery, and people from the industry and civil society, (b) field observations from the last mile of civic service delivery, and (c) weekly reflections from the three authors about the ethical concerns around justice, trust, fairness, and dignity that seemed to emerge.

The respondent narratives, observations, and reflections in the field notes were periodically revisited, triangulated, and validated with the information about the interventions presented in the official project documents. The resulting data formed the basis for our analysis, which we briefly present in the following section.

3.4 Data Analysis

In addition to weekly meetings on field notes, the three of us also had regular (monthly or sometimes bi-monthly) and prolonged meetings to analyse the field notes and consolidate emergent inferences till that period. The government of Karnataka, who were funding our project, mandated intermediate progress reports from us. This necessity also prompted us to conduct our data analysis periodically and diligently.

Literature review of articles on CAS and on articles that view urban infrastructure systems as CAS together gave us the necessary framing to analyse and present the case of ICCC from the view of the major human or organizational stakeholders (suppliers, consumers, and the government) and their interactions. It also helped us to pay attention to the complexity features in each of the UIS—for instance, the diversity within each of these stakeholders (e.g., differences in suppliers, different consumer types, etc.) and the diversity and redundancy in the channels of their interaction. Review of literature on the chosen ethical concerns (justice, trust, fairness, and dignity) helped us to define these concerns, identify them from the field notes, and present them in the case. Here again, the framework of CAS helped us delineate the plausible factors leading to different ethical concerns (as we depict in subsequent sections—Figures 2, 3, and 4).

Figure 1 provides a diagrammatic view of the process that led to the presentation of an empirical case of ICCC.

Fig. 1.

According to Eisenhardt [20], some of the key activities to build useful cases for theory building include, among others—laying down clear research objectives, adopting multiple data collection methods to triangulate information that make up the case, periodic review by multiple investigators to see the data through multiple lenses, and most importantly focusing on possible a priori theoretical constructs to build theoretically useful cases. The aforementioned strategy of data collection and analysis that we adopted also largely aligns with some of these activities outlined by Eisenhardt [20]. The overarching aim of our study was to construct an ethics-based governance framework for smart city IoT interventions, and our data collection and case development, as described earlier, were guided by this objective. As described, we also employed various data collection methods to triangulate the information gathered for case development, ensuring its robustness. Subsequently, each of the three researchers independently analysed the case using a CAS framework. We engaged in weekly discussions to delve beyond initial impressions and gain insights from multiple perspectives. CAS was chosen as the most suitable framework for our data analysis, because it provided a structured way to interpret smart UIS in terms of its impact on existing agents and their interactions within these systems. Furthermore, it gave us a structure to bring out potential ethical concerns associated with the IoT intervention and propose relevant questions for governance. In the following section, we present our case analysis, and in the subsequent sections, we borrow and extend the findings from the case to highlight the governance aspects of smart UIS in addressing different ethical concerns.

4 Smart City IoT Intervention: Case of ICCC

India's Smart Cities Mission aims to enhance UIS and improve the quality of life of citizens, who are consumers of the corresponding urban services. These UIS include water supply, electricity, sanitation, transportation, housing, solid waste management, and so on. The Mission's goal is to comprehensively develop these UIS, as cities are expected to house a significant portion of India's population by 2030 [52]. Initially, a set of cities are chosen as smart cities, and funded by the Mission to improve the efficiency and sustainability of their respective UIS. These cities are expected to set high standards for others and use data-driven smart solutions to achieve such goals.

ICCC is a smart city IoT intervention that integrates data generated from different UIS and provides a platform to build various applications. In this case, we focus on two key infrastructure systems within the cities of Karnataka that ICCCs aim to integrate: transportation, and solid waste management.

4.1 Smart Transportation System

The efficiency and sustainability goals set in the context of transportation include increasing the share of public transport in meeting transport demand, regaining road, and parking infrastructure as public good, and reducing transport sector contribution to air pollution. To achieve the first of these objectives, the smart city interventions propose to install vehicle tracking units in public buses to track them in real-time. The data from buses shall help city authorities, the primary supplier of transportation services, to dynamically optimise bus routes and schedules. Regarding information accessibility, these solutions envisage mobile applications that empower commuters (or consumers) to access real-time data concerning bus schedules, recommended routes, estimated travel times, and more.

The smart transportation systems are also expected to leverage data to influence behaviour of commuters and private mobility providers (other suppliers) to regain road space and minimise pollution. To this end telemetry-based tracking is to be made mandatory for all four-wheel private vehicles and intermediate public transport vehicles (such as autos, cabs, etc.) in the city. Measures are proposed to make these stakeholders bear the full cost of externalities—in terms of their contribution to congestion and pollution. Some of these include dynamic congestion fees for usage of personal vehicles during specified peak hours on major arterial roads, and linking vehicle registration with pollution emission. Telemetry-based tracking and automated number plate recognition are expected to also facilitate automated enforcement of some of these pricing policies. Real-time data about parking usage in the city shall also be leveraged to dynamically price the usage of public parking infrastructure.

Another intervention under smart transportation infrastructure is the promotion of digitally connected non-motorised bikes to improve last mile transport and reduce transport sector contribution to air pollution. In this regard, a public bike sharing application is planned and outsourced to private vendors for development. It is envisioned to help commuters locate and use non-motorised bikes for last-mile transport. Commuters will have to access the GPS-enabled bikes through an application or a smart card so that the bike sharing application can identify the commuters. This real-time location data about bikes and commuters helps the application to optimise utilisation of bikes, improve service delivery, and reduce theft and vandalism. A fully digital fare collection is envisaged by relying on online payment gateways like internet banking, credit/debit cards, mobile wallets, or prepaid smart cards.

4.2 Smart Solid Waste Management

A smart solid waste management system is another smart UIS that is proposed for integration into the ICCC. It is an IoT application to monitor and track municipal solid waste (MSW) collection. Some key objectives of this intervention include: increase the household level coverage of MSW collection, improve the efficiency of collecting MSW, and minimise MSW. The solution is expected to augment the decision-making capacity of city authorities, the primary supplier of solid waste management services, by providing real-time insights into the household waste collection process in a city.

The proposed IoT application, outsourced to private technology vendors, provides a specific type of access to the solid waste management system for suppliers (city authorities), consumers, and operations personnel. For authorities, it facilitates a view of MSW collection trends from different localities, real time location of waste pick-up vehicles, and track the attendance and movement of operations personnel. Consumers (households, for example) can raise on-demand waste collection requests through this application and advanced analytics is then used to dynamically allocate the requests to specific pick-up vehicles and operations personnel. These operations personnel are mandated to log information about waste collection from every household, which automatically gets linked with the household's property tax and solid waste cess information. In future, this is expected to help city authorities levy dynamic solid waste cess on households, based on the costs incurred toward collection, transportation, processing, and disposal of waste. Consumers are expected to raise complaints or on-demand waste collection requests through online application, and nearest operations personnel will be centrally assigned to address such grievances.

The ICCC platform hosts applications related to smart transportation, solid waste management and many more. As per the smart city proposals, there are four overarching objectives expected from the platform and its applications. First, the platform is expected to gather data from different UIS and enable data-driven applications to augment decision making capacity of respective city authorities. Second, the platform targets standardisation (through digitalisation) of processes related to service access, service delivery, and grievance redressal across different UIS, to reduce the overall response times. Third, data-driven applications enabled by the platform facilitates city authorities to centrally monitor and control the performance of operations personnel to improve their productivity. Fourth, the platform unlocks a city's data assets for private entrepreneurs in a standard manner, so that the latter can unleash rapid innovation in terms of building novel smart city applications to solve pressing city problems.

In the context of unlocking data assets, the Indian government proposed secure and privacy-preserving data exchange protocols between data producers (like public and private service suppliers, and, service consumers) and data consumers (like startups, researchers), through an open platform standard called India Urban Data Exchange (IUDX).¹ Following IUDX standards, ICCCs in cities are expected to serve as data marketplaces and enable smart city applications to be quickly developed at reduced costs, especially by relying on startups or private entrepreneurs. Researchers can now conduct data-driven experiments in real-world settings through these platforms.

4.3 Potential Impact of ICCC: A CAS Perspective

Below, we highlight the potential impact of ICCC on the following agents—consumers, suppliers (public and private), and operations personnel. In the context of each of the agents, we also use the CAS perspective to articulate the potential ethical concerns expressed by the respondents.

4.3.1 Consumers and Interactions.

ICCC envisions that consumers gain access to urban infrastructure services or to voice their feedback, by default via online channels (via mobile applications). In the context of smart transportation systems this can be seen in the form of online applications for citizens to access dynamically changing information about public transport routes and schedules, for planning their travel. The public bike sharing application envisions a complete reliance on online channels without any human intermediaries for accessing non-motorised bikes. In smart solid waste management, online channels are the default option for consumers to raise requests related to on-demand waste collection or grievance redressal.

As per a government respondent, online channels are considered cost effective compared to other options.

“Earlier the preference used to be to display information about when a particular bus arrives at a bus stop, in the dashboard of the bus stop itself. But this is very costly, because one must procure devices to be installed in each bus stop. Instead of this, building an application for customers to access the optimal route and bus information, is a lot easier and cheaper. This is now preferred in many of the smart cities.”

This predominance of online channels may cause concerns related to justice and fairness by raising entry barriers for consumers along the existing socio-economic and digital divides. For instance, in the context of the proposed dynamic solid waste cess, a civil society respondent believes that “low-income households might expend more time and effort to raise their grievances compared to residential complexes who may have dedicated staff to take care of such matters.” Digital divide can potentially exacerbate when waste collection requests and grievances need to be made only through online channels as opposed to direct interaction with last mile operations personnel. Income and digital divides also manifest when private technology vendors, outsourced with the operations responsibility of smart UIS, control policies, such as pricing, for access to services.

“E-ticketing vendors typically provide their devices for free in exchange for a shared ownership of data about buses, locations, number of commuters travelling, and so on. [Similarly] In the state-owned parking stations the IoT systems to enable smart parking applications (which enable personal vehicle owners to identify available paid parking slots) are outsourced to private vendors who not only share ownership of the data but also use it to control tariffs. Vendors may set tariffs in a manner that may inadvertently exclude some citizens from accessing the said services.” —a government respondent.

The above example also highlights a potential concern related to consumer trust, specifically data privacy and security. This is because vendors, in pursuit of profit-making use-cases tend to outsource consumer data to third-party vendors for further processing and analytics often without seeking their consent. Absence of appropriate legal and technical accountability measures that can distribute responsibilities for any unwarranted user data breaches between these chains of vendors, may exacerbate the problem of consumer trust.

The emphasis on digitally mediated service delivery often sidelines traditional human intermediaries thereby potentially limiting consumer choices, regardless of their capabilities. As a civil society respondent questions, “A lot of city services may be digitally maintained, but not everybody is capable of engaging, nor everybody wants to engage. If such a choice is not considered, then it essentially means to deny service for some. As a city authority, why would you want to do something like that?”

Some consumers may be capable enough to voice their concerns about the security and privacy risks from digitally mediated services and pursue them with the concerned authorities. However, as the respondent points, “For others it's something that they have to comply with, [as] they are not being given a choice… beyond that they don't even ask any questions… [because] asking questions itself is a luxury because… you need time and resources to be able to follow up.” The respondent also believes that limitation of choice constrains consumer agency and therefore it is important to retain access/feedback channels through human intermediaries, even if they seem redundant from a technology perspective.

Figure 2 highlights the potential impact of ICCC on consumers and their interactions within UIS, and how it might lead to ethical concerns of justice, fairness, and trust.

Fig. 2.

4.3.2 Suppliers, Operations Personnel, and Interactions.

ICCC enables suppliers of urban services with centralised data-driven decision making and control. In the smart transportation system, some examples of this include applications for city authorities to impose charges (congestion/pollution) on personal vehicle usage, automated penalty enforcement, algorithmic assignment of non-motorised bikes to requesting users, and so on. Assignment of operations personnel and pick up vehicles for on-demand municipal waste collection requests, and dynamic solid waste cess, are some such examples under smart solid waste management system.

An emphasis on centralised data-driven decision making often motivates suppliers to decide upon the nature of data that gets collected, indicators against which service delivery gets evaluated, and the smart city applications to be prioritised. A civil society respondent believes that suppliers may prefer standardisation of data collection and processing strategies across different UIS in pursuit of greater process automation and reduced costs. Such strategies often treat a particular form of data as conducive and ignore other forms of data that may be more relevant for use-cases related to a given UIS. According to the respondent, the demand for producing such standard usable data can potentially lead operations personnel to perform precarious labour such as manual data collection and labelling tasks.² A potential example of this could be frontline workers logging extensive information about household waste collection in the future in case of smart solid waste management.

ICCC, as per its stated objectives, also enables city authorities to remotely monitor and control the productivity of their operations personnel. Data-driven centralised control over last mile operations personnel is directly evident in some of the proposed applications. Examples include central assignment of pick-up vehicles and frontline workers for on-demand waste-collection requests in smart solid waste UIS, and dynamic adjustment of bus routes, schedules, and assignment of drivers and conductors in smart transportation UIS. Such control, if excessive, may cause dignity concerns for operations personnel who may now see reduced decision-making autonomy in their day-to-day operations.

Similar data-driven centralised control is also exercised over individual consumers while imposing user charges or dynamic cess. A potential outcome of this is the possible neglect by suppliers of the informal social fabric (made up of interactions between consumer, operations personnel, and other informal suppliers³) at the last mile in the context of service delivery. As per the respondents interviewed from the last-mile in case of solid waste management, we infer that the central algorithmic interventions can potentially disrupt existing informal ways by which different households and businesses negotiate with the frontline workers.

“Small hotel owners offer breakfast and lunch to the frontline workers in exchange for waste collection toward the end of breakfast and lunch times each day. Big restaurants not only do this but also directly interact with field supervisors to call for immediate waste pick-up in extreme cases. Individual households pay a mutually negotiated monthly amount to the frontline workers for the collection of waste pick-up at specific times or days as per their permitted schedule. Residential complexes, akin to big restaurants, have informal contracts with both frontline workers and the field supervisors. The reason frontline workers tend to enter these informal contracts is due to the low wages they get and the contractual nature of their job. They also help these workers to slightly deviate from the otherwise specified time schedules for waste-collection to address the contextual requirements of different users.” —Paraphrased from the narratives of last-mile stakeholders in solid waste management.

Disruption to this informal social fabric at the last mile can raise potential concerns related to dignity, trust, justice, and fairness. Issues of dignity arise when operations personnel are forced to give up their current informal service negotiations with consumers and replace them with digitally mediated interactions abiding by data-driven, centralised, and often dynamically changing control instructions. Regarding trust, informal negotiations among stakeholders at the last mile frequently serve to bridge the trust gap for consumers in service delivery. This is exemplified by frontline workers who can accommodate specific consumer requests, as illustrated in the previous narrative.

Unfortunately, such a trust may get disrupted when online channels become default and sideline or bypass the existing interaction channels through human intermediaries. As a faculty respondent puts it, “absence of human intermediaries to support locally, can severely constrain service delivery experience, especially for the digitally disadvantaged.” Moreover, issues related to justice and fairness arise when these interventions fail to acknowledge the existing socio-economic and digital divides, which may otherwise be bridged by the informal social fabric at the last mile. For instance, when data-driven central instructions are enforced on the frontline workers, the different ways by which they fulfil civic service delivery to the diversity of consumers at the last mile becomes difficult or even impossible. This could exacerbate socio-economic or digital divides by being disadvantageous w.r.t. service delivery for some.

Figure 3 highlights the potential impact of ICCC on suppliers, operations personnel, and on the interactions within an informal social fabric (between consumers and operations personnel) at the last mile of service delivery. It also highlights how this impact may lead to ethical concerns of dignity, justice, fairness, and trust.

Fig. 3.

4.3.3 Large Private Technology Vendors.

Cost concerns also motivate city authorities to promote one-size-fits-many solutions through ICCC platform. A greater leeway is given to large private technology vendors who can build such solutions at low cost by leveraging off-the-shelf components related to data collection, processing, and analytics.

“If a vendor is packaging the services for three cities, they can offer the same service at less cost. As a result, there could be a natural preference toward big technology vendors to take up these projects in multiple cities. Furthermore, these vendors will also get access to rich data, which they can leverage to build innovative products to retain and expand their user base while managing costs.” —a civil society respondent.

Data privacy and security concerns naturally accompany these one-size-fits-many solutions as they are often developed and tested in contexts different from the cities in focus.

“IoT-based solutions are…. Readily available boxed solutions developed for other country contexts and provisioned by big technology vendors… [These solutions come with] edge-devices and interlinked cloud service in a single package posing concern related to cyber security… one standard measure adopted is to not use the cloud service offered and use the central data centre (common for all smart cities) as the cloud.” —a government respondent

When the choice of IoT solutions is left to the technology vendors the resulting outcomes may sometimes perpetuate disparities between well-served and under-served localities or sets of population in the city. For instance, in case of public bike sharing, vendors are given freedom to decide profitable locations to install bike stations. One of the respondents from the industry believes that a natural option for them would be to deploy in already well-served areas where there is relatively certain demand. He argues that, “Of course, the vendors will go by demand, and since it is not easy to assess the demand at the lower serviced areas they may often get left out.”

According to this respondent, the vendors usually follow go-to-market strategies that focus on the early adopters, who naturally tend to be more digitally immersed. In response to this, a civil society respondent says that such strategies can potentially exacerbate service divides across consumers leading to concerns related to justice and fairness.

“Since data about the consumers from the less serviced areas may not be used for expansion plans, there is a possibility that the service provision may be biased toward certain sections of consumers, like those belonging to areas with high density of digitally immersed population, those belonging to certain social groups, economic class, or occupational groups (for example, areas dominated by IT sector employees). The intervention may also affect other local alternatives such as metered or shared auto-rickshaws making them more costly for some consumers.”

Concerns related to justice arise because the intervention may offer opportunities for service access or feedback primarily to those who are already capable enough (digitally immersed, for example). As the above narrative indicates such interventions may also ignore the importance of other local alternatives that are more accessible to certain sections of consumers. An example of this is also discussed by a government respondent in the context of enforcing interoperable digital payment standards across public and private mobility modes. He believes that it may sideline local mobility providers who fail to adapt sufficiently to such standards.

Concerns related to fairness arise when the data from underserved localities (which often intersect with other socio-economic categories of population) may become underrepresented in the data that informs the smart city applications. Moreover, as the service delivery processes get standardised over time, such biases may persist longer.

“A dominant tendency among service providers [or vendors] is to deploy IoT applications first, and correct issues after they start to interact with the real-world. A potential danger is that once any such intervention gets deployed the service delivery processes may get standardised and become very difficult to reverse later.” —opinion expressed by a civil society respondent in one of the workshops around IoT platform standards.

Figure 4 highlights the potential role of large private technology vendors that ICCC may create, and the impact it could have on consumers and other local service providers. It also highlights how this impact may lead to ethical concerns of justice, fairness, and trust.

Fig. 4.

5 Ethics-based Governance of Smart UIS: A CAS View

Our findings discussed above show that smart city IoT interventions tend to equip city authorities, who are the primary suppliers of urban infrastructure services, with data-driven decision-making and control. There is also a preference to outsource the operations of smart UIS to large private technology vendors. Given this scenario, city authorities must now take on a more prominent role in formulating effective governance measures to guide smart UIS in a desired direction—a direction that minimises potential ethical concerns.

The case findings highlight that IoT interventions dominate online channels for consumers overlooking their existing interactions with operations personnel and small businesses in accessing urban services. These interventions, often characterised by a one-size-fits-all approach, appear to compel local small businesses providing alternative service delivery options to either conform or face competition from larger private players. The centralised data-driven control facilitated by these interventions can encroach upon the autonomy of operations personnel. Furthermore, the standardisation of data gathering processes can potentially increase labour precarity for operations personnel by requiring extensive data collection to support algorithmic decision-making. In summary, these interventions risk sidelining previously trusted, informally negotiated channels of interaction between consumers, operations personnel, and other stakeholders like small suppliers in the last mile of service delivery. Essentially, this implies a reduction in diversity among agents by neglecting their hitherto roles, and in redundancy, by limiting the ability of these agents to provide alternative channels for consumers toward access/feedback w.r.t. the urban services.

This potential reduction of agents’ diversity and redundancy also foregrounds ethical concerns related to justice, trust, fairness, and dignity. For example, the smart city IoT interventions can raise justice concerns by sidelining the role of human agents who bridge the service-delivery gap for less digitally immersed sections of consumers, which may also intersect with other socio-economic divides. Ignoring their role could imply a fairness concern, or bias against disadvantaged sections of consumers currently supported by these agents. Large private technology vendors laying excessive control over consumer data could also lead to trust issues in the absence of appropriate accountability measures. Consumers might find it challenging to rely on their previous trustworthy choices, like small suppliers who could struggle due to intense competition, or operations personnel who are now obligated to follow data-driven instructions instead of engaging with consumers in informal and potentially trustworthy ways. The data-driven instructions, which impact the autonomy and working conditions of operations personnel and, consequently, their sense of dignity, in a way, could be viewed as contributing to a reduction in diversity within smart UIS. Figure 5 summarises a CAS view of ethics in the context of smart UIS.

Fig. 5.

Ignoring the diversity and redundancy in the existing UIS can potentially limit their self-organisation, and adaptation to changes [56, 57], for example, in the form of IoT interventions. Diversity of agents, for instance, could help agents in UIS to evoke contextual responses to adapt to any changes forced by such interventions. When the intervention disrupts some of the existing interaction channels, agents offering redundancy through alternative channels can help UIS to self-organise and restore despite any central control.

As the above discussion shows, the framework of CAS helped us to foreground ethical concerns in a structured manner as depicted through Figure 5. This framing allows us to pose relevant questions to city authorities in the context of an ethics-based governance of smart city IoT interventions. Following Singh et al. [67], we categorise these questions along legal and technical dimensions. Table 3 summarises the structured set of inquiries to assist city authorities in formulating appropriate governance strategies for smart city IoT interventions.

Table 3.

Recognise diversity and redundancy of agents in UIS: Currently who are the suppliers and associated agents (operations personnel and other providers) participating in the fulfilment of urban services to consumers?
	Legal Dimension	Technical Dimension
Justice	(a) How does the intervention affect the role of suppliers, operations personnel, and other providers? (b) Does the intervention acknowledge their role in bridging service delivery gaps for different sections of consumers?	(a) What kind of technical measures are required in the intervention to facilitate a continued participation of agents, and support their ways in bridging service-delivery gaps?
Fairness	(a) Does the intervention acknowledge the existing role of these agents in supporting disadvantaged sections of consumers? (b) What are the measures required to empower agents within the UIS to raise fairness concerns? (For example, identifying the role of civil society in voicing such concerns on the consumers’ behalf)	(a) What kind of technical measures are required to offer entry points for consumers, and other stakeholders such as the civil society, to raise fairness concerns?
Trust	(a) Given the undue control of technology vendors on consumer data, what accountability measures are needed to mitigate consumer trust? (b) Previously, what role did different agents play in bridging such a data-related trust gap between consumers and the UIS?	(a) What kind of technical measures are required to review the IoT intervention in terms of its potential privacy, safety, and security concerns?
Dignity	(a) Is the intervention favouring a centralised data-driven decision-making and control, if so, are such control instructions affecting the autonomy of operations personnel? (b) Does the intervention rely excessively on manual data collection and labelling tasks, if so, how are the working conditions where such tasks are performed?	(a) What kind of technical measures are required to distribute autonomy among agents, and avoid excessive data collection and labelling tasks?

Table 3. Ethics-based Governance Framework for Smart City IoT interventions

6 Conclusion

Our research delves into the intricacies of complexity, ethics, and governance within smart UIS. We employ an empirical case study focusing on the ICCC, an IoT platform slated for implementation across all seven designated smart cities in Karnataka state. Although the platform aims to integrate IoT interventions from various UIS within a city, our case builds upon two interventions related to solid waste management and transportation.

The complexity of these UIS is illustrated through the lens of its primary stakeholders such as suppliers, consumers, and the government, and the interactions centred around them. The case highlights the diversity within each of these stakeholders, in terms of the capacities of small vs large technology vendors or service-providers in the space of urban service provision, and customers differentiated along socio-economic dimensions and digital divide. It also highlights the redundancy of service provision channels enabled by different agents such as the small suppliers or technology vendors, and operations personnel.

Existing literature underscores the significance of diversity and redundancy within UIS, emphasizing their role in the self-organization and adaptation of urban systems. For instance, Mukhopadhyay [53] highlights the role played by informal and small private enterprises in delivering a significant portion of public or merit goods (like water supply, sanitation, transport, housing, etc.) especially at the last mile of civic service delivery. These small enterprises work in tandem with the operations personnel who facilitate service delivery of similar goods through the traditional government apparatus. Any disruption to the traditional apparatus is locally mitigated by these small enterprises who have established informal interaction channels with the citizens who are the consumers of civic services [61]. In addition to these, actors from civil society, local elected representatives, and many others also play a role in offering redundant channels to voice consumer feedback thereby contributing to the bottom-up processes that help UIS to sustain and grow in changing environments.

Our study indicates that smart city IoT interventions often downplay the significance of existing diversity and redundancy features within UIS, thus overlooking their essential contribution to UIS resilience. This disregard for diversity and redundancy can in-turn lead to ethical concerns such as justice, trust, fairness, and dignity, within the context of smart UIS. Given that the governance of smart UIS involves regulatory actions arising from the interactive relationship among various agents within the UIS, it becomes more meaningful to address even the ethical governance concerns from the perspective of these agents. Our study therefore encourages the exploration of mitigation measures for ethical concerns in smart UIS by examining the current roles of different agents in addressing such concerns within the existing systems and evaluating how their roles may evolve post-intervention.

We find that this CAS-informed view of ethical concerns has allowed us to frame a structured set of inquiries to assist city authorities in formulating appropriate governance strategies for smart city IoT interventions. Not just city authorities, but we posit that this framework can serve as a valuable reference for any stakeholder representing the industry, government, or civil society—the primary pillars of urban governance. It can assist them in articulating their viewpoints and formulating essential governance measures to address ethical issues associated with smart city IoT interventions.

The contributions presented in this study are limited by smart city as an application domain and the context of urban infrastructure systems in India. The proposed framework only covers a set of base questions that need to be operationalized further. This is the task we plan to take up in our further studies. We plan to gather such operational inputs from stakeholders of industry, government, and civil society, and build appropriate assessment procedures for evaluating any smart city IoT intervention along each of the above ethical dimensions. Last, we also hope future studies may extend these contributions by building on or testing this framework in other urban infrastructure and smart city contexts, or even other application domains.

Footnotes

https://iudx.org.in/

To understand the fragmentary nature of these tasks and how it equates to labour exploitation in today's data-driven world, see Reference [79].

A significant proportion of public or merit goods (like water supply, sanitation, transport, housing, and so on) to the last-mile in the Indian cities are provisioned by the informal and small private enterprises [53, 61].

References

[1]

Ali M. Adil and Yekang Ko. 2016. Socio-technical evolution of decentralized energy systems: A critical review and implications for urban planning and policy. Renew. Sustain. Energy Rev. 57, (2016), 1025–1037.

Abstract

1 Introduction

2 Theoretical Framework

2.1 Complex Adaptive Systems

2.1.1 Properties of CAS.

2.2 Urban Infrastructure Systems as CAS

2.2.1 Smart Cities: IoT-based Transition of UIS.

2.3 Ethics-based Governance of Smart UIS

2.3.1 Governance of IoT Interventions.

2.3.2 Ethics Informed Governance Objectives.

3 Methodology

3.1 Phase 1: Literature Review

3.2 Phase 2: Interviews

3.3 Phase 3: Archival Study and Field Observations

3.4 Data Analysis

4 Smart City IoT Intervention: Case of ICCC

4.1 Smart Transportation System

4.2 Smart Solid Waste Management

4.3 Potential Impact of ICCC: A CAS Perspective

4.3.1 Consumers and Interactions.

4.3.2 Suppliers, Operations Personnel, and Interactions.

4.3.3 Large Private Technology Vendors.

5 Ethics-based Governance of Smart UIS: A CAS View

6 Conclusion

Footnotes

References

Cited By

Index Terms

Recommendations

Review paper on technology adoption and sustainability in India towards smart cities

IoT technologiesfor smart cities

Governing Smart Cities

Comments

Information

Published In

Publisher

Publication History

Check for updates

Author Tags

Qualifiers

Funding Sources

Contributors

Other Metrics

Bibliometrics

Article Metrics

Other Metrics

Citations

Cited By

View options

PDF

eReader

Get Access

Login options

Full Access

Figures

Other

Share

Share this Publication link

Share on social media

Affiliations