Exploring smartness in public sector innovation - creating smart public services with the Internet of Things

ABSTRACT

The use of the term “smartness” in the context of public service delivery indicates an ambition of the public sector to become more agile and resilient through the adoption of emerging technologies. The Internet of Things (IoT) is an emerging technology that will be key to the realisation of smart public services. The research presented in this paper explored the role of IoT in public sector innovation through a qualitative study of how IoT technology can be leveraged to create and deliver smart winter road maintenance services. We use an existing smartness framework – based on the dimensions of efficiency, effectiveness, transparency and collaboration – to examine the consequences of introducing IoT-based innovation to road maintenance services. The findings suggest that IoT enables public sector innovation and that smartness is created through the combination of technology, people and organisations. The realisation of smartness in public sector innovation requires sufficient management capabilities and robust technology strategies, along with a willingness to explore and adopt new work practices rather than simply implement emerging technologies.

SPECIAL ISSUE EDITORS:

• Raquel Benbunan-Fich
• Kevin C. Desouza
• Kim Normann Andersen

KEYWORDS:

• Public sector innovation
• Internet of Things
• smart city
• smart public services

1. Introduction

Digitalisation initiatives within the public sector have long been described through concepts such as e-government, e-services, e-democracy, computerisation of government and informatisation (see for example, Meijer & Bolívar, 2016; Nam & Pardo, 2011; Van de Donk & Snellen, 1998; Yildiz, 2007), where “e-” denotes the electronic or digital component of the effort. However, the term “smart” has recently come to dominate the discourse on public sector digital innovation, appearing, for example, in concepts such as smart city (Meijer & Bolívar, 2016), smart government (J. R. Gil-Garcia et al., 2014), and smart governance (Scholl & AlAwadhi, 2016). This concept of smartness reflects ambitions that the public sector will become more agile and resilient through the adoption of emerging technologies (J. R. Gil-Garcia et al., 2014) and relays positive assumptions of, for example, interconnectedness, efficiency, sustainability, effectiveness, transparency and collaboration (Nam & Pardo, 2014). These so-called smartness dimensions describe the goals, as well as potential outcomes, of digital public sector innovation and reflect a forward-looking and optimistic view of how information technology (IT) can shape the public sector. Thus, it comes as no surprise that the concept of smartness, along with the technologies and collaboration that provide a platform for public sector innovation, have received so much enthusiasm in the past few years (Nam & Pardo, 2011).

The Internet of Things (IoT) is an increasingly important technological paradigm that is considered to be a key enabler of public sector smartness. Data from sensor-based systems can be integrated and synthesised with public services (Kankanhalli et al., 2019; Saarikko et al., 2020), ultimately culminating in the provision of services that are more responsive and accessible to citizens. The public sector can integrate IoT into a vast array of services, such as traffic solutions, effective waste management, optimised administration, and efficient healthcare (Chatfield & Reddick, 2019; Velsberg, 2018). Indeed, recent research has predicted that IoT technology, if extended to the public sector, will affect almost every public service and significantly improve processes and human efficiency (Tomar & Kaur, 2019).

While earlier research has explored the business potential of IoT (Wielki, 2018) and provided both technical solutions and visions for how IoT could be implemented in a smart city context (Zanella et al., 2014), only a few empirical studies have focused on the role of IoT in public sector service innovation. It is important to state that the use of the term smartness in relation to the public sector may create certain assumptions that could downplay the issues and challenges linked with such technological initiatives (Begg, 2002). This suggests that a more nuanced view of the smartness concept may be needed to reliably assess the feasibility of any presented solutions. We were primarily interested in better understanding how the application of IoT can contribute to public sector innovation, which lead to following research question: How does the application of IoT within public sector service innovation contribute to public sector smartness? Furthermore, we attempted to describe how different dimensions of smartness can be used to assess innovation process outcomes. Hence, this paper is relevant to discussions of smartness in public sector innovation as it explores how IoT solutions can be implemented in the public sector during the transition to smart services.

In order to address our research question, we conducted a qualitative study, collecting data from 30 Estonian municipalities that had all invested in, and used, IoT technologies to improve winter road maintenance services. Countries that are exposed to heavy snowfall have a large risk of public road infrastructure disruptions during winter months and many initiatives, from using road weather stations to vehicle fitted pavement temperature sensors, have therefore been put in place to improve winter road maintenance (see Hinkka et al., 2016). By applying a framework for smartness (Nam & Pardo, 2014) that includes aspects of efficiency, effectiveness, transparency and collaboration, we were able to explore how the digital innovation process affected both internal and external aspects of various municipal organisations. This in turn, allowed us to reflect more deeply upon the different dimensions of smartness and their relevance for public sector innovation.

The paper is structured as follows. The next section will chronicle previous research related to IoT technology and IT-enabled public sector innovation, as well as introduce a smartness framework. Then, Section 3 will describe the research design implemented in the present study. The results will be presented in Section 4, while the paper concludes with a discussion of our findings, implications for the public sector, and suggestions for future research.

2. Related research

2.1. IT-enabled public sector innovation

Innovation can be defined as the generation and adoption of new ideas or behaviours that relate to a product, service or new technology (Denford et al., 2017). On a general level, innovation is often described as the creation and adoption of something new, including products, services or processes (Tidd & Bessant, 2013). However, a closer look at innovation reveals micro- and macro-processes through which agents, organisations, institutions and the economy are transformed by the effects of a novel idea (Dodgson & Gann, 2010). While innovation is often viewed as a source of competitive advantage in the market, public sector innovation is different as it typically involves political decision-making. As such, a decision to introduce innovation in the public sector trickles down to various government departments and local government organisations for development and implementation, often causing added workload to an already stressed organisation that predominantly operates to avoid disaster rather than create value (Potts & Kastelle, 2010). Leaders in the public sector have come to a consensus that due to the large number of stakeholders, complex problems in the public sector cannot be addressed with money or standard solutions. Consequently, innovation is seen as a way to improve public sector services (Krishnamurthy & Desouza, 2014).

Even though the public and private sectors may follow widely different innovation trajectories (i.e. initiation, value creation and evaluation), it should not be forgotten that the public sector represents a large arena for innovation, and that IT solutions have been critical to the success of numerous public sector innovations (Dawson et al., 2016). Public actors implement new IT solutions in a bid to perform tasks more efficiently, which is a central tenet in the strategies of smart governments (Kliksberg, 2000; Mellouli et al., 2014), smart governance (Gil-Garcia & Aldama-Nalda, 2013; Scholl & AlAwadhi, 2016) and smart cities (Anthopoulos & Reddick, 2016; Nam & Pardo, 2014). Smart government/governance is generally described as the next step in government transformation, while the smart city concept encompasses innovative approaches for the provision of public services (Anthopoulos & Reddick, 2016).

Smartness is a multi-dimensional concept that has been linked to a plethora of terms, including efficiency, effectiveness, resiliency, citizen-centricity, openness, equality, sustainability and technological savviness (J. R. Gil-Garcia et al., 2014). In this paper, we apply a smartness framework that focuses on the dimensions of efficiency, effectiveness, transparency and collaboration to explore smartness in relation to IoT and capture both explicit and hidden assumptions about the role of this technology in value creation. These four dimensions have previously been included in assessments of smart city initiatives and their effects on management and service delivery (Nam & Pardo, 2014). In addition to trying to pinpoint which smartness dimensions contribute most to smart city success, researchers have also studied the context in which innovation occurs and how contextual conditions influence public sector smartness (Brewer & Selden, 2000; Kim, 2005; Meijer et al., 2016). For example, while smart city initiatives may have a transformational effect on organisational culture by enhancing data-driven transparency and accountability, the opposite is also true; a strong organisational culture may have a direct impact on effective public sector performance (Brewer & Selden, 2000; Nam & Pardo, 2014). Other contextual conditions, such as performance management measures, actor roles and relations, political systems, and stakeholder boundaries may vary greatly between organisations and the societies in which they function. It is therefore important to take into account both internal and external perspectives when assessing smart city initiatives and public sector innovation.

A smart city can be thought of as an entity that utilises networked infrastructure to improve economic and political efficiency, which, in turn, enables social, cultural, and urban development (Hollands, 2008). Another definition is a city that uses interactive, Internet-based applications to provide citizens and businesses with innovative services (Kuk & Janssen, 2011). The application of IoT to create smart city services is an example of IT-enabled public sector innovation that affects each branch of city government, e.g. administration, transport and healthcare (Velsberg, 2018).

2.2. Role & value of the Internet of Things

The IoT is widely regarded as the most disruptive phase of the Internet revolution (Atzori et al., 2017; Sundmaeker et al., 2010), and refers to physical objects that have network connectivity, processing capabilities and sensors. These characteristics allow the devices to record, process and communicate data (Lee & Lee, 2014). IoT may use wired and wireless communication technologies, allowing devices to communicate with other local devices, and global Internet and mobile network infrastructures to communicate with remote devices (Gluhak et al., 2011). The ability to transmit contextual data over space and time can improve process efficiency and human productivity (Balakrishna, 2012), generate new digital platforms and services (Saarikko et al., 2019), and strengthen the delivery of public services and support citizen participation (Allwinkle & Cruickshank, 2011).

IoT is a promising technology for smart public services (c.f. Atzori et al., 2010; Jin et al., 2014) as it enables the transparent and seamless incorporation of different systems and provides unrivalled access to data for the development of smart city services (Zanella et al., 2014). Equipping a city with IoT devices allows the local government to monitor almost every aspect of a city, take immediate action when problems occur, and provide residents with relevant information and services (Shin, 2009), all of which translate to increased efficiency, transparency, convenience, sustainability and effectiveness (Aldama-Nalda et al., 2012). However, the public sector can only benefit from IoT systems if it integrates and synthesises the IoT data with public services (Tomar & Kaur, 2019). Examples of how the public sector can leverage IoT to enhance public services include smart and connected bus stops, real-time traffic information, and efficient resource use through ventilation systems, autonomously operating lights, and thermostats that control the indoor climate of public buildings based on actual usage.

While these examples illustrate the potential benefits of IoT for local governments, its deployment must be carefully planned to anticipate and avoid potential difficulties. While some IoT devices require little attention, others need frequent updates, generate enormous amounts of data, as well as demand seamless mobility, real-time connections, high availability, and/or advanced security (Zanella et al., 2014; Zouganeli & Svinnset, 2009). In addition, the enormous amount of context-aware data produced through the use of IoT solutions raises concerns about data ownership, privacy and security; these aspects must be taken seriously so that neither citizens nor institutions are exposed to potential harm or wrongdoing (Weinberg et al., 2015).

2.3. An IoT smartness framework

The smartness framework by Nam and Pardo (2011, 2014)) offers both a managerial and service delivery point of view of four smartness dimensions (efficiency, effectiveness, transparency, and collaboration) that can be used to assess smart government initiatives. The management view of a public organisation represents an internal perspective as it focuses on activities and processes within and across governmental agencies. On the other hand, the service delivery perspective, which describes how a public organisation interacts with citizens, private companies and other relevant non-governmental actors, represents an external perspective as it considers interactions that occur across public/private borders.

The four distinct smartness dimensions – efficiency, effectiveness, transparency and collaboration – have all been thoroughly researched in public sector theory and in relation to both smart cities and smart government. Earlier research has, for example, found that improved performance, expressed as increased efficiency and effectiveness, is the most frequently cited motivation for public sector innovation (Cordella & Bonina, 2012; De Vries et al., 2016) and that the adoption of new and emerging technologies is often seen as key to developing public sector smartness (J. R. Gil-Garcia et al., 2016). Furthermore, other researchers have stressed how incorporating IT solutions into government processes may improve access to information and facilitate information sharing, thereby increasing transparency (Bertot et al., 2010; Cucciniello et al., 2017) and fostering collaboration both within governmental agencies and between governmental and non-governmental actors (Ansell & Gash, 2008; Bakıcı et al., 2013).

The smartness framework applied in the present study is illustrated in Figure 1, and indicates that all four smartness dimensions (efficiency, effectiveness, transparency and collaboration) can be viewed from both internal and external perspectives. Consequently, the smartness of an initiative is a combination of how it performs on each distinct smartness dimension and the extent to which it fulfils organisational goals.

Figure 1. The smartness framework (from Nam & Pardo, 2011, 2014).

Various aspects of the smartness framework have been applied in other studies (J. R. Gil-Garcia et al., 2016; Viale Pereira et al., 2017). In the presented research, we chose to adopt the framework as a whole since it provides comprehensive metrics for exploring not only the different smartness dimensions, but also the internal and external perspectives on public sector innovation. As we focused specifically on IoT, we adapted the framework to show how the application of IoT within public sector services contributes to public sector smartness, and how these initiatives perform according to each dimension of smartness. We will now explain the different smartness dimensions in more detail and describe how they are applied in the context of IoT-enabled public sector innovation. The resulting IoT smartness framework can be found in Table 1, which also includes several references that strengthened our understanding of the main concepts.

Table 1. An IoT smartness framework (adapted from Nam and Pardo (2011, 2014))).

Dimensions	Management (internal) perspective	Service delivery (external) perspective	References
Efficiency	How IoT may reduce intra-organisational workloads, minimise waste, time and effort in activities and processes, and maximise intended managerial objectives.	How IoT may improve the production and delivery of public services in terms of minimising expenses, time, waste and effort, and maximising outputs	Cordella and Bonina (2012) De Vries et al. (2016) J. R. Gil-Garcia et al. (2016) Nam and Pardo (2011, 2014) Ostroff and Schmitt (1993) Washburn et al. (2010)
Effectiveness	How IoT may improve the quality of the management of internal processes, for example, by increasing knowledge	How IoT may contribute to improved service delivery to external stakeholders, such as citizens and private companies	Brewer and Selden (2000) Nam and Pardo (2011, 2014) Osborne et al. (2014) Ostroff and Schmitt (1993)
Transparency	How IoT may improve information integration across government offices	How IoT may provide access to information and decision making, in the process of service delivery	Bertot et al. (2010) Coglianese (2009) Cucciniello et al. (2017) Nam and Pardo (2011, 2014) Meijer (2009) McIvor et al. (2002) Porumbescu (2015)
Collaboration	How IoT may foster internal collaboration processes of two or more people working together across and inside the departments of the public organisation	How IoT may foster the process of governmental and non-governmental parties working together	Ansell and Gash (2008) Bakıcı et al. (2013) Bătăgan (2011) McDermott (2010) Nam and Pardo (2011, 2014)

Efficiency and effectiveness are often mentioned together as measures of organisational performance (Ostroff & Schmitt, 1993), and are also traditionally considered the major benefits of public sector innovation (De Vries et al., 2016). Washburn et al. (2010), for example, state that efficient operations management and effective service delivery are key components of smart government initiatives and should be assessed from both internal and external perspectives. Efficiency is defined as the ratio of desired outputs to inputs (Nam & Pardo, 2014; Ostroff & Schmitt, 1993); hence, this term describes the relationship between an investment (time, money, effort) and the outcome (results or returns on investment). Efficiency is generally improved by maximising output while reducing or minimising input, but in the public sector, efficiency also denotes the ability to deliver quality public services that cater to the needs of citizens under improved internal managerial ability (Nam & Pardo, 2014). Public sector investments in IT often target efficiency, more specifically, reductions in transactional costs, to save both time and money (Cordella & Bonina, 2012). In this paper, we focus on how IoT contributes to internal efficiency based on reductions in intra-organisational workloads, waste, as well as the time and effort spent on activities. Furthermore, we consider how IoT solutions can reduce the time that managers need to achieve their objectives. Likewise, we examine how IoT may contribute to external efficiency by exploring whether IoT improves the delivery of public services in terms of maximised output and minimised expenses, time, and waste.

Effectiveness is defined as the capability to produce a desired output through the sound management of internal resources, both personnel and money (Ostroff & Schmitt, 1993). Effectiveness has been shown to be closely tied to organisational factors: “Effective public sector performance is more often driven by strong organisational cultures, good management practices, and effective communication networks than it is by rules and regulations or procedures and pay scales” (Brewer & Selden, 2000, p. 688). Knowledge is seen as a key resource in creating public sector effectiveness, and can be supported by information systems that generate, share and disseminate information across internal and external stakeholders (Bekkers et al., 2011; Osborne et al., 2014). In this paper, we explore how IoT initiatives impact both internal and external effectiveness. The assessments of internal effectiveness mainly measure the management of internal processes, for example, how IoT solutions increase knowledge, while evaluations of external effectiveness reflect how the use of IoT improves service delivery to external stakeholders, such as citizens and private companies.

Transparency is a well-established construct in public sector theory that has been researched for decades to understand its role in improving the quality of government (Piotrowski, 2008; Roberts, 2006). Researchers have, for example, shown that Internet technologies can improve transparency, interactions and responsiveness to citizen needs within the public sector (McIvor et al., 2002). Transparency research predominantly presents two types of definitions for transparency: one that focuses on information availability (breadth of access) and another that concerns the flow of information (who has access). Taken together, an overarching definition is that transparency explains how much information is accessible and who has access (Cucciniello et al., 2017). Transparency requires the availability and flow of timely, comprehensive, relevant, high-quality and reliable information concerning government activities (De Ferranti et al., 2009; Meijer, 2009). Contemporary government transparency research considers external stakeholders (citizens) to be the primary audience of available government information (Porumbescu, 2015). Earlier studies have found transparency to facilitate participation, improve financial management and reduce corruption. However, transparency has also been shown to be ineffective at engendering trust in the government (Cucciniello et al., 2017). As transparency has become a popular topic in the context of government development, any initiatives for the public sector should consider the current level of transparency in various governmental processes so that no unrealistic promises of openness are made to the general public. In this paper, we focus on how IoT impacts internal transparency by studying the sharing of information across government offices (internal transparency) and how IoT can provide better access to information and decision-making in the process of service delivery (external transparency).

Collaboration in the public sector describes how multiple stakeholders are engaged in collective public decision-making processes (McDermott, 2010). Ansell and Gash (2008) conducted a literature review of collaborative governance to identify which variables are critical to such collective processes. Trust building, commitment and shared understanding were found to be critical enablers of collaboration. A public organisation is typically the initiator of collaborative governance, but it is important to state that successful collaboration requires participation by external stakeholders. The framework presented by Nam and Pardo (2011, 2014)) considers collaboration from two perspectives. The internal perspective focuses on how smart government improves collaboration between agencies/departments within the government, while the external perspective is concerned with how effectively the government interacts with non-governmental parties, e.g. communities, citizens, private companies and other public organisations, along with the cooperation between two or more of these entities (Bătăgan, 2011; J. R. Gil-Garcia et al., 2007). In this paper we focus on how IoT may foster internal collaboration processes of two or more people working together across and inside public organisation departments, as well as external collaboration, that is, the process of governmental and non-governmental parties working together.

Table 1. An IoT smartness framework (adapted from Nam and Pardo (2011, 2014)))

3. Research design

We conducted a qualitative field study of the use of IoT in winter road maintenance services among 30 municipalities in Estonia to explore how IoT may affect innovation in public sector services. A qualitative research approach was chosen as it allows an in-depth understanding of the study subject (Cavaye, 1996; Yin, 2003), which – in our case – was the application of IoT within public sector service innovation and how IoT initiatives affect the functioning of public organisations.

3.1. Data collection

Data were collected from 30 municipalities that had applied IoT in winter road maintenance services. The main techniques used to collect data were semi-structured interviews, document reviews and observations of work practices and IoT systems. Various data sources enabled us to perform comparisons and triangulate the data (Miles et al., 2014). All of the data were collected by the first author between March and December 2017, while the second and third authors assisted with the analysis.

At the time of the study, Estonia was in the process of reforming the structure of its municipalities, with the end-goal to reduce the number of municipalities from 213 to 79. The restructuring was to be implemented in several cycles; this study was conducted when the first cycle had been completed and the number of municipalities had been reduced to 202. We sent an email to the mayor/head/vice-mayor of each municipality asking whether the municipality would be interested in participating in our study. We received replies from 81 municipalities, 30 of which had applied IoT for winter road maintenance services. All of these 30 municipalities were selected for this study. We contacted these 30 municipalities and asked to interview people who were responsible for winter road maintenance and who were aware of how IoT was used for service delivery. Four of the municipalities had two people who satisfied these criteria, and the final sample included 34 respondents representing 30 municipalities.

In addition, for all participating municipalities, we contacted private companies contracted to deliver winter road maintenance. Some of the municipalities in our study hired multiple service contractors and some contractors provided services for multiple municipalities. In total, we contacted 34 service contractors and received replies from twelve of them. From those, five maintenance managers from five private service contractors agreed to participate and were interviewed to gather insight on the winter road maintenance, to better understand the external perspective, and to put the gathered information into context. The interviewees from these companies oversaw day-to-day aspects of winter road maintenance from the service contractor’s point of view. One contractor provided maintenance services for three municipalities, another provided services for two municipalities, while the other three each provided services for a single municipality. All of the respondents are summarised in Table 2. The municipality ID corresponds with the respondent ID, so the number 15, for example, refers to both municipality 15 and the respondent from that municipality. In the cases where there are two respondents from the same municipality, they are labelled with the municipality ID, then a separate respondent ID (for example, 1.1 and 1.2 denoting the two respondents from municipality 1). The interviewed service contractors are labelled SC1, SC2, etcetera.

Table 2. A summary of interview respondents.

Municipality & respondent ID	Respondent title	Respondent age	Area by km2	Density (people per km^2)	Service contractor ID
1.1	Head of public administrative unit	45	859,0	15,4
1.2	Head of economics department	43	859,0	15,4
2	Head of economics department	54	480,0	23,0
3	Environmental and municipal advisor	30	533,0	11,5
4	Head of economics department	45	5,99	676,1	SC1
5	Municipal vice-mayor	29	1008,0	10,8
6.1	Head of economics department	62	890,0	11,1	SC2
6.2	Head of public administrative unit	29	890,0	11,1
7	Head of economics department	63	881,0	8,5
8	Municipal vice-mayor	51	1449,0	4,9
9.1	Municipal vice-mayor	55	389,0	14,6
9.2	Head of public administrative unit	47	389,0	14,6
10	Head of economics department	55	494,0	9,8
11	Head of economics department	64	126,0	12,1
12	Municipal vice-mayor	37	706,0	20,0	SC3
13	Municipal vice-mayor	31	599,0	14,6	SC4
14	Municipal vice-mayor	56	1164,0	6,5	SC4
15	Municipal mayor	51	158,1	9,6	SC3
16	Municipal vice-mayor	31	9,85	308,0
17	Municipal mayor	47	10,12	798,2
18	Environmental and municipal advisor	44	177,0	21,8
19	Municipal vice-mayor	53	10,64	1448,6	SC3
20	Municipal mayor	39	58,37	62,1
21	Municipal mayor	44	123,32	8,1
22	Municipal mayor	38	933,0	5,8
23.1	Municipal mayor	52	103,4	23,1
23.2	Head of public administrative unit	30	103,4	23,1
24	Head of economics department	34	398,0	4,7
25	Municipal mayor	46	750,0	21,3
26	Municipal vice-mayor	50	269,81	4,5
27	Municipal vice-mayor	57	300,0	5,5	SC5
28	Municipal vice-mayor	56	952,0	11,3
29	Head of road maintenance	41	3,83	689,5
30	Municipal vice-mayor	58	628,0	34,1

Table 2. A summary of interview respondents

Each interview lasted between 20 and 120 minutes, with an average of 42 minutes. All of the interviews were audio-recorded and later transcribed. The interviews were carried out in a semi-structured manner, which allowed the interviewer to address predetermined topics while allowing the interviewees to discuss additional topics. The pre-defined questions were designed to cover the dimensions defined in the smartness framework and issues such as the role of IoT in effective road maintenance, the collaboration between different actors, the importance of data and information, and the organisation of winter road maintenance. The interview transcript was to a large extent the same for both public officials and private contractors, differing only by the perspective from which the questions were asked and answered. For example, when interviewing municipalities, we asked about trust: “Has there ever been a conflict of trust between you (the municipality) and the service contractors?” When interviewing service contractors, the same question was asked, but from the opposing perspective: “Has there ever been a conflict of trust between you (the service contractor) and the municipality?” Table 3 presents an excerpt of the questions and specifies, where applicable, which dimension(s) from the IoT smartness framework was/were covered by each question.

Table 3. Overview of interview questions.

Interview question examples:	Associated dimension(s) of the smartness framework
What is the most important aspect of winter road maintenance?
How do you decide which roads to maintain during the winter?	Efficiency; Effectiveness
How do you categorise the roads?	Efficiency; Effectiveness
How do you decide in which order the roads should be maintained?	Efficiency; Transparency
What are the biggest challenges regarding road maintenance?
Has there ever been a conflict of trust between you (the municipality) and the service contractors?	Transparency; Collaboration
Why did the municipality decide to use the tracking devices?
How do you verify the integrity of the collected data?	Transparency
Who does the municipality share the collected data with?	Transparency; Collaboration

Table 3. Overview of interview questions

In addition to the interviews, we also collected supporting documentation in the form of 35 procurement documents, 77 public documents, and 47 technical documents. The procurement documents and public documents (such as government reports, municipal information on winter road maintenance, and public statements) provided insight into winter road maintenance and were used to understand the context in which the services were provided; the technical documents provided information on the IoT system. After the interviews, we asked the respondents to demonstrate the IoT system, with the first author taking detailed notes about how the system functions. In 19 municipalities, the respondent and the first author inspected the road conditions by car in order to provide the author with an empirical understanding of the context.

All data collected during the study were gathered in a digital archive containing the interview transcripts, field notes and documents.

3.2. Data analysis

The data analysis employed a qualitative research process (Eisenhardt, 1989; Miles et al., 2014) in which we applied a theoretical framework to the collected data (Walsham, 2006). The analysis was carried out in three different phases, with ATLAS.ti used for the second and third steps of the data analysis.

In the first step, the IoT smartness framework (see Table 1) was used to generate an initial set of eight theory-driven codes. The coding process began with a meeting in which all of the authors participated to agree upon the interpretation of the codes and how they should be applied to the material. During this meeting, the authors first independently applied the codes to an excerpt of the data material. Upon completion, the coded segments were compared and discussed. Through this process, the researchers came to an agreement about how the codes should be interpreted, which resulted in larger overlap scores for each round of coding. After three rounds of coding, overlap stood at 92.6% while Cohen’s kappa was 0.852, which was deemed a satisfactory result.

In the second step, the first author independently coded all of the data from the digital archive using the eight codes identified from the IoT smartness framework. The insights and examples gained from the joint coding guided his work. Table 4 presents coding examples for each of the eight codes.

Table 4. Coding examples.

Code name	Code example (specific code name)
	Internal perspective	External perspective
Efficiency	Public official 15: “It is no longer necessary to drive behind them and check, whether they are working or not.” (Internal efficiency)	Public official 1.1: “The service has actually improved. The drivers maintain problematic roads much faster than before. “ (External efficiency)
Effectiveness	Public official 5: “Of course I do not see whether the snowplough is lifted from the ground or not, however, I can see the speed of the vehicle.” (Internal effectiveness)	Public official 22: “To avoid all of the arguments between the citizens and service contractors, we actually made the data available online.” (External effectiveness)
Transparency	Public official 13: “Every public official learns quickly that it is crucial to defend why they took the decision.” (Internal transparency)	Public official 4: “Many people look up the vehicles online and complain about the route, driving speed and so on.” (External transparency)
Collaboration	Public official 7: “I talk to my colleagues from population registry, regarding where new people have moved in. It helps to manage the maintenance.” (Internal collaboration)	Public official 2: “When we start working with new service contractors, I drive through all roads with them and they save everything to their system – where things are and what needs to be done.” (External collaboration)

Table 4. Coding examples

The eight codes were recorded a total of 615 times, with 381 and 234 of the codes covering the external and internal dimensions, respectively. In Section 4, Table 5 specifies the number of coded segments that represented each of the eight codes.

Table 5. Summary of innovation outcomes.

	Summary of innovation outcomes
	Internal perspective	External perspective	Total
Efficiency	113 coded segments Examples: Reduced time to manage maintenance Better managerial overview of the equipment, contractors and road status Easier to resolve disputes, e.g. through vehicle analysis Streamlined governmental processes for information sharing	91 coded segments Examples: Decreased costs for maintenance services Improved communication with and up-to-date information to citizens and contractors Improved logistics Need-based snow ploughing	204 coded segments
Effectiveness	65 coded segments Examples: Targeted investments in infrastructure Data analytics to follow up routing Allowed to introduce cycle-based pricing models, which emphasised quality over price	108 coded segments Examples: Possibility to identify problematic service contractors through access to data Re-routing of vehicles based on real-time information Increased road maintenance quality, e.g. through real time coordination and control	173 coded segments
Transparency	27 coded segments Examples: Making the decision making process more visible through the use of vehicle reports Increased understanding of road condition	85 coded segments Examples: Increased citizen engagement Increased control of drivers Transparency of the service delivery process Tacit knowledge was documented and made accessible	112 coded segments
Collaboration	29 coded segments Examples: The IoT system served as mediator for new internal collaborations within the municipality	97 coded segments Examples: Change in route planning based on citizen demands New work practice involving the public official in road maintenance work IoT system enabled enhanced knowledge transfer between old and new drivers	126 coded segments
Total	234 coded segments	381 coded segments

The third step entailed an overall analysis of the appearance of the eight codes and their distribution across the municipalities. This analysis helped us better understand the smartness framework and how the different dimensions of smartness varied across municipalities.

4. Results

In this section, we present our results regarding the role of IoT in public sector service innovation in the context of Estonian winter road maintenance.

4.1. The case: smart winter road maintenance in Estonia

The research was carried out in Estonia, which is located in northern Europe. Due to Estonia’s location on the coast of the Baltic Sea, the country has four seasons of nearly equal length and an annual average temperature of around 5 °C. Although snow may appear from October to May, the period between December and February is usually characterised by the harshest climate, resulting in blizzards, black-ice, and snowstorms. Therefore, winter road maintenance is crucial to enabling road transport during winter months in Estonia.

During the winter, each municipality in Estonia is obliged to keep their roads clear of snow and ice. Estonian roads are categorised by importance so that there is a predetermined order in which roads must be cleared of snow and ice. These categories also include time limits for road clearance. Winter road maintenance is financed by the government, although allocated funds are scarce.

To maintain the roads, municipalities hire service contractors for each district through a procurement process. These service contractors have a maintenance manager who oversees drivers and manages their work, while municipalities appoint a public official who oversees all of the districts, assures the quality of the work performed by the contractors and coordinates work between service contractors and their drivers.

Before the use of IoT, the public official would coordinate work and distribute information among different stakeholders mainly by telephone. In order to determine whether the roads had been properly maintained or to resolve disputes, they would have to drive to different locations, call the maintenance manager and drivers, or rely on citizens or co-workers to provide information. Furthermore, these public officials often had other functions besides road maintenance; thus, many of them felt overwhelmed with their workload. To ease the pressure on public officials, as well as in response to citizen involvement, many municipalities have made IoT-equipped service contractor vehicles mandatory. The first municipality made this decision in 2012/2013 and others soon followed.

Typically, the service contractor rents the IoT solution for the duration of the winter road maintenance season (from November to the end of March). The IoT solution consists of a tracking device and various sensors that are hardwired to the vehicle. The sensors can identify the vehicle driver (using a Radio-frequency identification (RFID) card, for example) and which road-clearing accessory (such as snowplough blade or salt spreader) is in use, and measure speed, location, road salt dosage and spread, fuel consumption and level, and other vehicle-related information. Figure 2 gives an overview of data collected by the sensors.

Figure 2. Overview of data collected by sensors.

The IoT device collects and sorts data before sending it to the service provider’s server for further analysis. The collected data enable real-time and retrospective tracking of vehicles and can be used to display which accessory is in use, generate vehicle reports, manage task allocation, and visualise the movement of the vehicles for citizens. The generated data are typically stored by the technology provider for six months. The IoT devices use Global Positioning System (GPS) to identify the vehicle’s position, while the data are transmitted via Global System for Mobile Communication (GSM). Once a device has been installed on a vehicle, it will operate autonomously as long as it is turned on. Tasks can be allocated to drivers through an on-board task management system, which optimises driving directions, updates tasks, and allows the drivers to take pictures of road conditions. The municipality and service contractor have access to the detailed data (seen in Figure 3), while a more general view can be provided to the public.

Figure 3. System display for maintenance manager.

4.2. Efficiency – from internal and external perspectives

The IoT devices provide vehicle data at any given time-point, ensuring that the public official always has a complete overview of road status and how the service contractors are performing. This improves efficiency within the organisation, as the public official enjoys increased productivity, better communication and a clear, real-time overview of the work they are in charge of. Public official 15 commented:

It is no longer necessary to drive behind the service contractor and check whether they are working or not. You can easily see from the system which roads they maintained and which roads they did not.

From the perspective of external efficiency, the IoT solution enabled the public officials to swiftly provide reliable information to citizens whenever they inquired or had comments about the road maintenance (e.g. when would their road be cleared). Public official 17 commented:

When citizens previously called and inquired about maintenance, we first had to call the service contractor, who called the maintenance manager, who then called the driver […] and the information moved back to us in the same way.

Accordingly, the IoT system became essential to acquiring real-time information, enabling the resolution of disputes and responding to citizens in a timely manner, for example, in cases when citizens were worried about roads not being maintained. Municipalities that had previously been overwhelmed by incoming emails and calls from citizens were able to make relevant information about road maintenance publicly available through their website. This significantly reduced the workload of public officials and, as such, they could now concentrate on critical and strategic aspects of road maintenance – which improved service delivery and efficiency – rather than responding to individual citizens. Public official 30 commented:

We decided to make all the information available for the public because the citizens were calling us all the time. We have made it possible for people to see where the vehicles are, who is maintaining the road, which machine (is being used), which roads have been maintained, and we told everyone “if you want real-time information, you can go and see for yourselves”.

Municipalities further benefited from vehicle analysis, as it provided easy verification of whether the service contractors had fulfilled their contractual obligations to the municipality. This was relevant, as there had been instances in which service contractors skipped small roads or detoured to high priority roads for the purpose of financial gain. The automatic analysis of vehicle movements now allows public officials to identify contractual violations, e.g. a service contractor clearing private roads while they are contracted for winter road maintenance or not maintaining roads that required maintenance. Public official 26 commented:

There have been problems with new drivers. Everything is digital, so I’m not sure whether they don’t know how to read the screen or how to use the system, but they are missing roads. When I look at the system and see some roads not maintained, I tell them to return and maintain them.

Additionally, public officials benefited from the automatically generated vehicle reports, which meant that the drivers and maintenance manager no longer had to manually track vehicle-related information – such as working hours and routes taken – to generate invoices.

From an external service delivery perspective, the IoT solution improved efficiency in a number of ways. The use of IoT devices increased the hourly cost of snow- ploughing, but because number of hours spent on maintenance significantly decreased, the total cost of winter maintenance actually decreased. These gains were mainly explained by the ability of the maintenance manager to re-route vehicles and manage their work regardless of time and place, which created a new form of data-driven planning. The maintenance manager (SC1) for municipality 4 commented:

It is easy to organise the maintenance; I can see exactly how vehicles are moving in real-time. When there is a problem with a road, I can instantly notify the closest driver to maintain that specific road.

In addition, the drivers no longer took unsolicited breaks while working, did not maintain roads that were not under a municipality’s jurisdiction, used optimal routes, efficiently identified places that needed more maintenance, and maintained roads according to requirements. The number of neglected roads dropped significantly following the implementation of the IoT system, as the solution includes a control mechanism that enables all parties to track which roads need to be maintained. Finally, when government officials or citizens required some specific road maintenance service – for example, in the case of an emergency call or when a social worker had to visit a patient – it was now possible to re-route the optimum vehicle to do the job. Public official 29 commented:

When I’m told that an ambulance is coming or someone needs to go to the hospital and asks whether it would be possible to maintain the road, I say “Of course, everything is possible”. The driver will be notified that he needs to maintain that road by 9 AM. […] All drivers have a tablet in the car, where they will see which roads they need to maintain and in which order.

4.3. Effectiveness – from internal and external perspectives

From the internal management perspective, equipping road maintenance vehicles with IoT devices provided public officials with in-depth information that allowed them to identify problematic service contractors and drivers (such as a driver sitting in the vehicle with the engine off but logging working hours). However, even though the IoT technology enhanced public officials’ abilities to monitor winter road management, some municipalities still struggled to achieve the required quality of maintenance as service contractors invented ways to circumvent the reporting facilities of the IoT system. Although public officials can monitor vehicle-related information, they were unable to see exactly how contractors worked. For example, some of the vehicles were driven with the snowplough higher from the ground than required, which reduced fuel consumption but resulted in poorly maintained roads. To overcome these issues, the public officials set up occasional speed alerts to track maintenance quality through vehicle speed. Public official 5 commented:

As you know, fuel consumption for an hour of driving can differ a lot when you raise the snowplough from the ground, instead of keeping it close to the ground. […] Of course I do not see whether the snowplough is lifted from the ground or not. However, I can see the speed of the vehicle. […] From the speed I will know whether he was just driving or actually maintaining the road.

The information and experience obtained from using the IoT system was later used to guide investments. For example, roads that were observed to require frequent maintenance were prioritised when scheduling repairs. These targeted investments and the overall use of IoT was shown to cut maintenance costs, increase accountability and improve road quality during the winter. Disputes with citizens and service contractors were solved more easily than before due to increased control and real-time coordination. The gathered data made it impossible for service contractors to say that they had maintained a road when they had not. Furthermore, the IoT system enabled municipalities to set up area notifications that would send alerts when a vehicle left the authorised maintenance area. Hence, drivers could not maintain private roads while contracted to maintain public roads. Flaws in the vehicle reports were found more often than before, both by manual analysis of the device data and by comparing the performed maintenance against the standard cycle through computer analysis. This helped municipalities avoid paying excess fees for maintenance and allowed easier verification of vehicle reports. The real-time monitoring of vehicles and weather conditions significantly increased the public officials’ control over the monitoring of road maintenance. Public official 12 commented:

Sometimes I need to call them and ask why the men are not maintaining the roads when it snows. […] Our municipality is, in a sense, peculiar – we have different weather conditions across the municipality. Here it snows, in another part it might not snow. But I always call them and point out that they should be working.

From the external service delivery perspective, the IoT system significantly improved effectiveness. Municipalities require certain roads to be maintained before others, e.g. those leading to government buildings or serving as main roads for the public. Before the application of an IoT solution, the public officials had difficulties monitoring whether roads had been maintained in the required order and within the required time. Following the introduction of IoT technology, roads were increasingly maintained according to the requirements. It was also found that the service contractors increasingly maintained only the roads they were contracted to clean instead of cleaning roads they were not responsible for. Moreover, the IoT system enabled the service contractors to effectively re-route vehicles to the areas in which they were required the most. Overall, the introduction of IoT led to cost reductions, as well as increased the quality, effectiveness and responsiveness of road maintenance. Public official 10 commented:

Some drivers used to be extremely slow. I found them standing in one place for a long time, sometimes talking to a neighbour, but it is crucial to swiftly maintain roads after a blizzard, so we told them to speed up. […] Another thing is private roads; we do not maintain them, but we saw that contractors were maintaining those roads with our money. We had a talk with the service contractor on that topic. […] It has helped us to avoid such occurrences.

Furthermore, the IoT system enabled the introduction of new business models, as data collected using the sensors were used to change the existing payment system. Municipalities switched from time-based or flat-fee pricing models to cycle-based pricing models. The cycle-based pricing model combined three components: the maximum time it took to clean a district, vehicle characteristics, and an hourly rate for the vehicle. Larger vehicles have a high hourly rate, but are also able to work faster, which makes them cheaper for the municipalities than small vehicles. The IoT system added a layer of control in that service contractors could not intentionally use smaller and less powerful vehicles instead of the ones that had been specified during the procurement process. The new business model reduced the cost of winter road maintenance by a factor of three and improved the quality of road maintenance by a factor of two.

4.4. Transparency – from internal and external perspectives

The implementation of an IoT system affected internal transparency across the studied municipalities in numerous ways. As the public sector has limited resources for winter road maintenance, any expenditures must usually be defended. The information provided by the IoT system enabled public officials to describe and justify their spending decisions, as well as counter false accusations. Officials had, for example, been accused of prioritising roads that lead to government officials’ homes; however, the data gathered from IoT devices allowed them to show that this accusation was not true. Public official 24 commented:

All vehicle reports are sent to the accountant. It prevents anyone from accusing me of doing something wrong later. Everyone in the public sector is browsing someone else’s pockets and you need to protect yourself from that.

Service delivery transparency also improved following the implementation of IoT. A few municipalities made winter road maintenance information publicly available on their websites. This allowed citizens to see when roads were scheduled to be maintained and where the vehicles were moving. However, making this type of information publicly available generated unforeseen problems for the service contractors. Public officials started receiving reports from drivers describing how they were being ambushed by citizens demanding that their roads should be maintained before others. The situation was further complicated by the fact that service contractors and their drivers took direct orders from the municipality, they had no authority to take on additional jobs or to deviate from their assigned routes. Public official 4 commented:

We no longer make the information public. It is mainly burdensome for the driver, but also for me as a manager. Many people look up the vehicles online and complain about the route, driving speed and so on. […] You cannot operate like that, you need to consider what roads you need to clean first and what roads are secondary.

The introduction of the IoT system meant public officials could now monitor vehicle locations and driver behaviour. For example, when a vehicle was driving towards a certain direction, the public official could conclude that the vehicle must have run out of road salt. Furthermore, when the map showed that all of the vehicles were assigned to high priority roads, the official knew why local roads were not being maintained and could relay this information to the public. Additionally, the IoT technology allowed public officials to notify private contractors of unorthodox stops or routes which they suspected to have a corrupt purpose, such as stealing fuel or using the vehicles for personal matters. For instance, a vehicle was observed making an unusual stop for 20 minutes on a remote forest road, which alarmed the public official who notified the service contractor. The service contractor found that the driver had indeed stolen fuel.

When citizens contacted the municipality to get information on maintenance, the public official could rapidly provide information regarding the driver’s work based on real-time vehicle movement and speed data. When problems occurred, such as a citizen being splashed with mud, vehicle data allowed the problem to be solved in a fair, empirical way. Municipalities could further use the vehicle reports to ensure that the service contractors were working as had been agreed. Public official 19 commented:

This winter (2016/2017), a contractor was supposed to be there but for some reason the vehicle was driving in the opposite direction. We called and asked what was happening and the driver said that his window was hit by a rock. This is understandable. These things [IoT-solutions] provide operational information when something, somewhere, is off.

However, public officials from the municipalities also reported several problems. The most common issue was a constant need to modify and develop the system, which can be expected whenever a new technological solution is implemented. This need was driven by a lack of trust in the service contractors. The trust issue was brought into focus by a lack of quality, more specifically, poorly maintained roads, route deviations, and inefficient work. Even though service contractors used the tracking devices, municipalities still had to physically control the vehicle logs and service contractors’ work. For example, some drivers “forgot” to turn on the IoT device in order to avoid being monitored and bypass the system.

4.5. Collaboration – from internal and external perspectives

Information regarding winter road maintenance was available to everyone within the municipal operating unit, which laid the groundwork for internal collaboration. When a problem occurred and a municipal employee was alerted, the employee had the opportunity to forward the location, picture and problem description to the system or directly notify the public official. Some municipalities also had district officials who forwarded road information they had collected from citizens and by physically inspecting the roads. However, the increased availability of information within the municipal organisation could cause tensions between government employees and road management personnel. Because everyone understood that road maintenance resources are scarce, the municipality was exposed to pressure at the times when the devices were being paid for even when there was no snow and the devices were not being used. Public official 27 commented:

One year we did not have any snow in November, and I was called to explain my decision to have the devices on trucks kept on stand-by. […] It is hard to explain that to them in a way that they can understand.

As all of the vehicle information was available, public officials and drivers started to collaborate in new ways. When a public official found or was notified of a road that had not been maintained or needed further maintenance, they directly notified the closest driver. This changed the traditional work practices, as this new form of collaboration cut out the service contractor’s maintenance manager and created a situation in which the public official became more active in the maintenance. For instance, if it started to snow and drivers were not maintaining the roads, public officials started to give out tasks. Furthermore, public officials started to identify drivers who maintained roads carelessly or poorly and notified the service contractors.

Municipalities that published winter road maintenance information on their websites could inform and engage with the public. Hence, citizens were able to access real-time information so that they could assess how, where and when the vehicles were driving. Whenever a problem occurred, the citizens knew they could contact the public official. Moreover, municipalities started to plan maintenance routes depending on citizen requests, whereby roads prioritised by citizens were maintained first.

Furthermore, the available information helped municipalities and service contractors learn about the districts from one another, e.g. which roads were most affected by ground blizzards or which areas were the most problematic. This information was added to the system so that the drivers could see the information on their dashboard. This allowed new drivers to quickly learn the specific characteristics of each district. Public official 2 commented:

When we start working with new service contractors, I drive through all the roads with them and they save everything to their system – where things are and what needs to be done.

4.6. Innovation outcomes of IoT-based winter road maintenance services

Our results suggest that the implementation of the IoT system influenced both internal and external aspects of public organisations. Some of the innovation outcomes, such as increased efficiency in road maintenance and increased control of equipment and work processes, were expected, while others, such as the double-edged sword of citizen involvement, were more of a surprise. For example, some municipalities encouraged and welcomed engaged citizens, while others found the increased transparency they provided the public to be more of a burden.

Table 5 summarises the main innovation outcomes of implementing an IoT system in winter road maintenance. In the table, the innovation outcomes are structured according to the four dimensions of the smartness framework applied in this research. Overall, efficiency was the most commonly coded segment, with a total of 204 coded segments (113 coded segments for internal efficiency and 91 coded segments for external efficiency), followed by effectiveness (173 coded segments, 65 concerning internal effectiveness and 108 concerning external effectiveness), collaboration (126 coded segments, 29 concerning internal collaboration and 97 concerning external collaboration) and transparency (112 coded segments, 27 concerning internal transparency and 85 concerning external transparency). This result demonstrates that applying IoT to public sector service innovation can increase efficiency, effectiveness, transparency and collaboration.

Efficiency was the only dimension in which the internal perspective had more coded segments than the external perspective. Overall, a total of 381 coded segments reflected the external perspective, compared to 234 coded segments for the internal perspective. This shows that IoT implementation in smart public services has the potential to significantly affect the organisation’s relationships with citizens, private companies and non-governmental actors. Thus, the application of IoT can lead to a win-win situation in which the public and private spheres work together to develop effective public services.

Table 5. Summary of innovation outcomes

5. Discussion

The research presented in this paper aimed to contribute to the discussion of public sector innovation by exploring the adoption of IoT solutions and the subsequent transition to smart public services. Based on the research question: “How does the application of IoT within public sector service innovation contribute to public sector smartness?” we adapted a previously presented smartness framework (Nam & Pardo, 2014) into an IoT smartness framework and analysed the use of IoT-enabled winter road maintenance services across various Estonian municipalities. In this section, we extend the discussion of smartness and examine how IoT solutions can influence efficiency, effectiveness, transparency and collaboration within public organisations (internal perspective) and across the government – non-government boundary (external perspective).

5.1. IoT and public sector smartness

Previous research indicates that public organisations most commonly adopt innovative IT-based services to improve efficiency (Cordella & Bonina, 2012; J. R. Gil-Garcia et al., 2016). This was also confirmed in the present study, as can be seen in Table 5. Efficiency is important to the procurement of public services, as public organisations always strive to provide measures of efficiency to demonstrate the responsible use of public funds. In this study, examples of the improved efficiency generated by IoT solutions were linked to the real-time visualisation of processes and included increased productivity through a decrease in the time required to complete tasks, improved communication, a better overview of processes and decreased maintenance costs. While process visibility has been shown to affect effort as well as satisfaction in the service industry (Liu et al., 2015), this study shows that – in the context of public service innovation – process visibility is also strongly connected to efficiency. In other words, a clear benefit of IoT-enabled public services was the increase in process visibility, which was a result of reliable real-time data and led to improved efficiency. However, we also saw that the narrow focus on efficiency meant that most municipalities concentrated on improving existing tasks in order to save time and money. This prevented the organisations from leveraging the underlying potential of the technology to make disruptive or radical changes and create new services or processes. Thus, one could argue that although efficiency is widely regarded as a desirable outcome of digital public sector innovation (Cordella & Bonina, 2012), a solitary focus on enhancing this smartness dimension may end up fostering only incremental innovation rather than encouraging more creative and ground-breaking developments.

Effectiveness was the second most common dimension identified from the study material, and these innovation outcomes were mainly expressed through improvements in the quality of public services. While several benefits of the IoT system, such as access to better information and timelier road maintenance, were noticed right after implementation, other benefits, such as tracking snowplough height from the road to monitor work quality or geographic information system based reporting applications for citizens, emerged later in the service innovation process. Earlier studies have identified knowledge as critical to the successful implementation of IT solutions with the overarching aim of improving public sector effectiveness (Osborne et al., 2014). Our study contributes to this by showing that IoT systems provide an arena for knowledge co-creation between the public actor and its external service partners as work practices became visible and accessible to both parties. IoT can thus support the development of public sector effectiveness through the creation of new business models and processes. However, although the majority of the studied municipalities expressed intentions to use IoT to make winter road maintenance processes more efficient and effective, only a few managed to fully utilise the potential of IoT to digitally transform their processes. The public organisations which, for example, used the available data to share knowledge with diverse stakeholders and create new business models, achieved outcomes that far surpassed the simple automation of existing processes, which was the approach at other municipalities. Effectiveness has been closely linked to organisational culture (Brewer & Selden, 2000). Hence, this study shows that public organisations need to both understand the transformative potential of IoT technology and proactively apply it in a way that changes existing processes if they want to significantly improve effectiveness. This requires clear management strategies that incorporate technological and process knowledge.

Transparency was the third most common dimension identified from the study material, while collaboration was the least often discussed smartness dimension. In recent years, transparency has become an objective for local governments as it is regarded as essential for promoting accountability, building trust, improving access to public services, increasing citizen participation, informing decision-making, and providing information to citizens and other stakeholders (Bertot et al., 2010; McDermott, 2010; Tolbert & Mossberger, 2006). Earlier studies have shown how IT, more specifically, Internet technologies, have allowed the public sector to be more responsive to the needs of citizens, which translates to improved transparency (McIvor et al., 2002). An unbiased data source is the cornerstone of transparent decision-making – whether it is the rearrangement of payment systems and business models or increased citizen engagement. However, this study also revealed that IoT-enabled transparency created unforeseen problems among the service contractors as citizens, who were initially given complete access to drivers’ routes, started demanding customised services based on their individual needs. Our research thus contributes important insights for public sector transparency and IoT implementation, as we show that the issue of transparency requires the balancing of benefits and risks. Bernstein (2017) posits that the benefits of transparency, e.g. learning and control, may become risks when the perspective is changed from the observer to the observed. In the context of emerging technology adoption by the public sector, the possibility to make processes more transparent may create unforeseen risks, such as unsatisfied customers once the public is given access to winter maintenance operational information. Although increased transparency may lead to such risks, it is important to note that keeping data private would hinder benefits related to efficiency, effectiveness and collaboration. Thus, balancing the benefits and risks associated with transparency requires awareness about relevant perspectives and how the benefits outweigh the risks.

Public sector collaboration is regarded as an important part of engaging citizens and other relevant stakeholders (McDermott, 2010). IT solutions contribute information-sharing capabilities that are crucial to public sector collaboration (J. R. Gil-Garcia et al., 2007). Our analysis found that IoT could facilitate both internal and external collaboration following the development of new work practices, i.e. knowledge sharing and activities which joined the government and non-government boundaries. However, even though IoT can help public organisations connect with stakeholders, its application does not automatically lead to innovative collaboration outcomes. For example, only five of the studied municipalities were using the system to improve collaboration with service providers, e.g. teach new vehicle contractors and their drivers about the districts. IoT-generated data can improve the relationship(s) between internal and external actors by building trust and developing common work practices, but this is only possible if the actors consider their previous prejudices and understand how the system can be used to create shared, coherent processes that are aligned with the overall objective. Accordingly, our results demonstrate how diverse stakeholders need to engage in mutually beneficial activities if they hope to experience innovative collaboration outcomes.

5.2. Managing IT-enabled public sector service innovation

In the ongoing discussion of public sector innovation, most experts agree that the adoption of emerging technologies will drive the transition towards smartness by making public organisations more agile and resilient (J. R. Gil-Garcia et al., 2014). However, investing into new technology will not automatically result in innovative services and, for this reason, researchers have called for a more nuanced view of the smartness concept (Allwinkle & Cruickshank, 2011; Begg, 2002; Hollands, 2008). This study contributed to this field by focusing on a case in which public organisations implemented an IoT solution, and then describing how this decision influenced four distinct components of smartness. Changes in organisational functioning (internal perspective) and the ability to engage relevant stakeholders as well as the general public (external perspective) were taken into account in assessments of each smartness component. Hence, we identified the innovation outcomes connected to each dimension, as well as captured the contextual conditions surrounding different stakeholders (public officials, service contractors, citizens) that were connected to the innovation process. Accordingly, the IoT smartness framework applied in this study can be used to explore which parts of a public organisation emerging technologies will most affect, as well as determine how new technology investments will influence internal and external functioning. A such, it provides an overarching perspective to understand where smartness in the public context appears and how it can be understood and researched.

Our research is valuable to the field of public sector digital innovation, as it provides empirical results of how public organisations can implement innovative technologies. More specifically, we observed how smartness is achieved when stakeholder goals align or complement each other, technology can be used to both enhance and transform organisational practices and processes, and the organisational culture can further support change. This study extends our understanding of the smartness concept by highlighting the importance of context, the role of technology and how they are entangled. Previous research has indicated that digital technology must transform organisational processes in order to generate value in an organisational setting (Brynjolfsson & Hitt, 2000). We found that innovation is tightly coupled to the context in which it occurs, as we found that the application of the same technology can yield different outcomes in different contexts. This confirms that contextual factors such as knowledge, routines and people are important in the creation of innovation (Brewer & Selden, 2000; Kim, 2005; Meijer et al., 2016). We also witnessed how the application of IoT, i.e. sensor-based systems that provide access to abundant real-time data, created new innovation outcomes within the public organisations as well as the external service environment. The public organisations that successfully used IoT to develop smart public services proactively considered various stakeholder perspectives to ensure mutually beneficial collaboration. The ability to constantly shift perspectives, from the internal to the external, and back again, stood out as a factor that distinguished the more innovative municipalities from the less innovative ones. Thus, an important insight was that the realisation of smartness in public sector innovation requires management capabilities and strategies as well as a willingness to explore and adopt new work practices rather than the simple implementation of an emerging technology.

From a management perspective, the implementation of IoT enables public organisations to increase efficiency, effectiveness, transparency and collaboration. However, the data that this technology generates, and the insights these data provide, may also open the door for extended (and sometimes unwarranted) governmental control, privacy infringements, and shifting power dynamics. In this study, we saw that although IoT was introduced to generate certain outcomes (improvements in efficiency, effectiveness, and transparency, among others), the system and services were constantly evolving as new applications for the collected data were discovered and processes that were previously largely invisible or had suddenly become antiquated were identified. This suggests that the flexibility required to truly benefit from IoT-enabled services may strain the status quo of how public organisations function. Even though the targeted service and context are known when IoT is installed, both can be expected to evolve as the collected data reveal new application areas for the technology. Hence, emergent technologies can be expected to result in emergent contexts and emergent use, which will require the public sector to stand with one foot in the present and one in the future. This will allow public organisations to solve today’s problems and prepare for tomorrow’s solutions.

5.3. Implications for research and practice

Previous research has only provided vague accounts of the outcomes of IoT adoption within the public sector. Our empirical study covered internal and external perspectives of how IoT implementation affects the functioning of a public organisation. We were able to observe how the implementation of IoT resulted not only in new ways of performing old tasks, but also led to the discovery of new opportunities, processes, and services. The presented results demonstrate that the smartness framework is not only valuable for assessing smart government initiatives, but also relevant for identifying the potential innovation outcomes of emerging technologies. This study thus contributes a more nuanced view on smartness within public sector innovation by considering both contextual conditions, expressed as internal and external perspectives, and the role of technology. While we believe that the dimensions of efficiency, effectiveness, transparency and collaboration will be valuable in forthcoming studies of smart public services, we strongly encourage future research to be specific on the role of technology. Moreover, although positive innovation outcomes were seen across all four dimensions, our results indicate that IoT investments should extensively assess the transparency dimension to minimise the risk of unexpected behaviour and outcomes.

For practitioners, this study shows that emerging technologies such as IoT may foster innovation in the public sector when combined with simultaneous organisational development. We also show that IoT enables new forms of external collaboration and citizen engagement, which can generate both positive and negative outcomes, as the collected data provide unprecedented access to operational information.

Our results show that IoT solutions can help public organisations develop their smartness in numerous ways, and significantly influence both internal functioning and service delivery. As the IoT and other emerging technologies continue to provide new technical possibilities we need to move beyond the characterisations and frameworks from the era of e-government as the core of digital innovation in the public sector. The digital transformation of the public sector will require innovations not only at the core of the organisation, but also at the edge, together with citizens, external companies and other organisations.

6. Conclusions

This study has two main contributions for the field of public sector digital innovation. First, the study shows how the application of IoT in the public sector provides both internal and external benefits and affects several dimensions of the smartness framework. This can be extended to demonstrate that – in the public sector – smartness is created through the combination of technology, people and organisations. Second, we show how the dimensions of efficiency, effectiveness, collaboration and transparency are valuable to identifying the specific outcomes of applying emerging technologies to public services. Furthermore, by taking into account contextual conditions and being specific about the role of technology we were able to provide deep insight into public sector innovation processes. Efficiency is a critical factor when a public organisation is considering investing in IoT, and this study demonstrated how an IoT system can help organisations improve efficiency by streamlining processes. Effectiveness, on the other hand, is difficult to achieve if an organisation preserves the same work practices after IoT implementation rather than using the gathered data to develop new and innovative public services. The increase in transparency provided by IoT enabled transparent decision-making, but also created unforeseen problems when processes and actors became more visible to citizens. IoT also facilitates collaboration, as different actors can access the same data and engage in collaborative work. However, this requires a willingness to be open, as well as active commitment from both internal and external actors.

Finally, this study was subject to several limitations that must be acknowledged. Our focus on public sector organisations and service providers excluded the perspective of citizens. We identified how some of the IoT-related externalities affected citizens, but further studies could explicitly include citizens to provide a broader understanding of innovation in smart public services. Furthermore, this study concentrated on how IoT affects public sector service innovation, and future research is needed to reflect the effects on private sector service innovation under similar circumstances. In addition, this study has focused on one specific type of IoT solution in one particular country, and more research on the role, outcome and effects of other emerging technologies in public sector innovation from diverse contexts are needed to validate whether the presented results are technology- and/or country-specific. Furthermore, emerging technologies can influence how work processes and organisational structures are shaped in the public sector. Research on IoT may address certain critical design features of other emerging technologies, and thus, help researchers understand how and why emerging technologies work. Hence, future research should work to extend the lessons learnt from the adoption of IoT to other emerging technologies such as augmented intelligence and robotic process automation. This will benefit the transition from traditional services to smart, IT-enabled services. Furthermore, researchers should investigate how different emerging technologies are integrated into work processes, how they affect the design of work and, consequently, contribute to public sector smartness. This would be an important topic to study as IoT does not inherently have deterministic qualities that lead to certain outcomes. We would also recommend that researchers investigate how the implementation of emerging technologies leads to potential value conflicts and changing power dynamics between private and public organisations in order to understand how IoT affects not only the different actors involved, but also their relationship to each other. Finally, the smartness framework provided comprehensive metrics for exploring the effects of IoT in public service innovation. Additional studies in smart city contexts could further validate the benefits of using this framework.

Disclosure statement

No potential conflict of interest was reported by the authors.