PaperlayerView exact revision on arXiv

Source-linked AI summary

Analyzing the Adoption Challenges of the Internet of Things (IoT) and Artificial Intelligence (AI) for Smart Cities in China

Ke Wang, Yafei Zhao, Rajan Kumar Gangadhari, Zhixing Li

arXiv:2205.01067v1cs.CYcs.AI

TL;DR

Urban service pressures accompany the challenge of adopting AI and IoT in smart-city development. The paper reviews adoption barriers and analyzes their interrelationships using expert opinions and DEMATEL. It classifies four barriers as causes, with insufficient funds having the greatest D-R score.

  • Problem

    The study examines what hinders AI and IoT adoption for smart cities facing pressures on urban services.

    Source: §1 · p. 1, §3 · p. 8

  • Method

    A PRISMA literature review identified ten challenges, whose interrelationships were analyzed with DEMATEL using expert opinions.

    Source: §3 · p. 8

  • Results

    Infrastructure (C1), insufficient funds (C2), cybersecurity risks (C3), and lack of trust (C10) were cause-group challenges; C2 had the greatest D-R score, 1.422.

    Source: §5 · p. 22

  • Takeaways & Limitations

    Organizations planning AI and IoT in smart-city development should focus first on infrastructure, funds, cybersecurity, and trust challenges.

    Source: §5 · p. 23

    The study did not use a Grey scale to measure the accuracy of expert opinions.

    Source: §6 · p. 24

Abstract

from arXiv · show

Smart cities play a vital role in the growth of a nation. In recent years, several countries have made huge investments in developing smart cities to offer sustainable living. However, there are some challenges to overcome in smart city development, such as traffic and transportation man-agement, energy and water distribution and management, air quality and waste management monitoring, etc. The capabilities of the Internet of Things (IoT) and artificial intelligence (AI) can help to achieve some goals of smart cities, and there are proven examples from some cities like Singapore, Copenhagen, etc. However, the adoption of AI and the IoT in developing countries has some challenges. The analysis of challenges hindering the adoption of AI and the IoT are very limited. This study aims to fill this research gap by analyzing the causal relationships among the challenges in smart city development, and contains several parts that conclude the previous scholars work, as well as independent research and investigation, such as data collection and analysis based on DEMATEL. In this paper, we have reviewed the literature to extract key chal-lenges for the adoption of AI and the IoT. These helped us to proceed with the investigation and analyze the adoption status. Therefore, using the PRISMA method, 10 challenges were identified from the literature review. Subsequently, determination of the causal inter-relationships among the key challenges based on expert opinions using DEMATEL is performed. This study explored the driving and dependent power of the challenges, and causal relationships between the barriers were established.

1. Introduction

Urban growth and sustainability pressures frame the need for smart-city services, while IoT and AI offer tools for monitoring and managing city systems. The study examines adoption challenges through a literature review and expert-based DEMATEL analysis.

1. Introduction

  • The paper describes urban migration as burdening big-city services and identifies sustainability and environmental impacts as motivations for smart-city development.Cities are framed as needing to meet public needs while limiting environmental harm.Source: §1 · pp. 1–2
  • Smart cities aim to improve resource use, city management and services, and citizens’ quality of life, with AI and IoT applied across sectors.The paper notes applications in transportation, security, energy, buildings, education, and health.Source: §1 · p. 2
  • China’s smart-city pilot projects have increased in recent years, according to a synthesis of government documents and public reports.The sources include China MOHURD, China NDRC, and Forward the Economist.Source: §1 · p. 3
  • China’s smart-city market spending was projected to reach USD 25.9 billion in 2020, up 12.7% year-on-year and second only to the United States.The cited spending guide is IDC’s Worldwide Smart City Spending Guide, released in July 2020.Source: §1 · p. 4
  • The study uses PRISMA literature review and DEMATEL analysis of expert opinions to examine causal relationships among barriers to AI and IoT adoption.The authors state that the findings are intended to help policymakers, practitioners, and researchers understand the effects of these challenges in China.Source: §1 · p. 4

2. Development of AI and IoT for Smart Cities

The section presents smart-city development examples and explains how IoT and AI contribute to connected urban systems, including building-energy management. It describes China’s smart-city industry and reviews IoT applications and AI-based building-performance methods.

2.1. The Case for Smart City Development

  • Singapore examples include IoT-enabled fleet, air-quality, energy, and agriculture services, while smart transport and e-government are among its most developed services.The section also describes intelligent lighting, water-meter monitoring, and smart waste collection as applications in other cities.Source: §2.1 · pp. 4–5
  • China had more than 700 smart-city pilots by 2019, and its market reached RMB 10.5 trillion, with Hangzhou and Shenzhen illustrating city-scale data integration and AI use.Hangzhou’s City Brain served 11 systems and 48 application scenarios; Shenzhen’s construction center centralized control of more than 400 sites.Source: §2.1 · pp. 5–6

2.2. Features of the IoT

  • The IoT connects sensor-equipped objects to the Internet for information exchange, identification, location, tracking, monitoring, and management.Examples discussed include city-scale testbeds in Antwerp and Santander.Source: §2.2 · p. 6
  • Building-energy studies combine IoT and data-mining methods to support building operations, analyze energy-use patterns, and manage consumption.Methods reviewed include association mining, regression, classification, clustering, and neural networks.Source: §2.2 · p. 6

2.3. Implementation of AI in Smart Cities

  • Building-performance research uses data-driven or physical models to predict energy use; ANNs, SVMs, and decision trees account for 76% of reviewed algorithms.Nearly half of the studies reviewed focused on predicting overall building energy consumption.Source: §2.3 · pp. 6–7

3. Literature Review and Challenges to IoT and AI Adoption in Smart Cities

The review applies PRISMA to identify adoption challenges, then examines infrastructure as a foundational requirement for smart-city systems.

3. Literature Review and Challenges to IoT and AI Adoption in Smart Cities

  • A PRISMA review screened 439 initial papers, removed 128 duplicates and 87 papers that did not meet criteria, and included 171 articles to identify ten challenges.The review drew on English- and Chinese-language academic search engines and used expert opinions with DEMATEL to establish challenge interrelationships.Source: §3 · p. 8

3.1. Lack of Infrastructure

  • Smart cities require advanced, Internet-connected infrastructure, and connected IoT devices collect physical-world data to support decisions and urban services.The reviewed literature also identifies adapting existing infrastructure for multiple purposes as a need.Source: §3.1 · p. 9

3.2. Insufficient Funds or Capital

Insufficient funding can delay smart-city projects and raise their costs, making access to capital a central implementation concern.

3.2. Insufficient Funds or Capital

  • Insufficient funds can delay smart-city implementation and increase project costs, while governments may seek private-sector funding for these projects.The cited literature describes investment and capital funding as a major threat to smart-city strategies.Source: §3.2 · pp. 9–10

3.3. Cybersecurity and Data Risks

Smart cities must manage data security, privacy, and cyber risks as their networks connect governments, industries, and citizens. AI and IoT also offer approaches to waste collection and processing.

3.3. Cybersecurity and Data Risks

  • Smart-city networks generate actionable data but raise privacy and security concerns, while the paper calls for shared responsibility and effective attack detection.China's personal data privacy law requires separate consent for sensitive information; the passage also describes a cyberattack that disrupted a US pipeline company.Source: §3.3 · p. 10
  • Robust connectivity is required for smart-city success, and the paper reviews attack-detection approaches alongside their stated limitations.The reviewed work includes centralized and distributed monitoring, DoS detection methods, and insider-attack surveys.Source: §3.3 · p. 10

3.4. Smart Waste and Hygiene Management

  • AI-enabled waste segregation and ant colony optimization can support waste sorting and disposal, while centralized monitoring can improve service quality.The passage frames AI and IoT as relevant across collection, transportation, treatment, recycling, and disposal.Source: §3.4 · p. 11

3.5. Lack of Professionals

Smart-city adoption depends on skilled professionals, while connected devices and changing consumption increase energy needs. The section discusses technology and renewable-energy approaches to managing those demands.

3.5. Lack of Professionals

  • A shortage of skilled professionals can lag AI and IoT adoption and make intelligent urban management, operational efficiency, and quality assurance harder to achieve.The passage highlights the need for cross-field computer-technology knowledge and notes the added importance of attracting professionals after the pandemic.Source: §3.5 · p. 11

3.6. Managing Energy Demands

  • Connected smart devices require substantial energy, while rising demand and costs make energy provision and renewable-source migration challenging.The paper describes AI and IoT approaches including automated streetlights, peak-time price increases, and equipment upgrades.Source: §3.6 · p. 11
  • Regional grids powered by solar, wind, or hydro are proposed as a way to meet growing energy demand locally rather than relying on central supply.The passage presents this as Northfield's argument.Source: §3.6 · p. 12

3.7. Managing Transportation

The paper presents efficient transportation and environmental-risk response as important smart-city challenges. It describes AI and IoT applications in transport analysis, pollution monitoring, and risk detection.

3.7. Managing Transportation

  • Intelligent transportation systems are treated as essential to smart cities, with AI analyzing transport data for congestion, accident causes, and possible preventive measures.China's shared-bicycle systems use real-time road conditions to propose manual scheduling plans, addressing the last-mile problem.Source: §3.7 · p. 12

3.8. Environmental Risks

  • AI and IoT can support environmental-risk detection and mitigation through design, automated drones, and air-quality sensors on public online platforms.The passage connects urban population growth with environmental pressures including pollution from transport and waste.Source: §3.8 · p. 12

3.9. Managing Public Health and Education

The section describes AI and IoT applications in public health and education, alongside concerns about protecting personal health data. It also outlines ways these technologies could support medical services and learning.

3.9. Managing Public Health and Education

  • AI could assess patients using medical histories and current conditions, while digitized records and IoT-enabled equipment could support healthcare access and management.The passage also notes blockchain as an alternative for storing medical records and cites an AI-based online triage model for smart-city hospitals.Source: §3.9 · p. 13
  • The use of personal health data can risk breaches or misuse that compromise patient or public rights.Examples include data about personal journeys during COVID-19 and visits for difficult medical conditions.Source: §3.9 · p. 13
  • AI could tailor educational programs to industry and academic needs, while IoT can facilitate student interaction with peers and teachers.The passage also describes a low-cost, flexible educational platform using Raspberry Pi development boards.Source: §3.9 · p. 13

3.10. Lack of Trust in AI and IoT

Trust is presented as a barrier that may slow AI and IoT implementation in smart cities. The broader review identifies 10 challenges and calls for more attention to efficient, secure algorithms.

3.10. Lack of Trust in AI and IoT

  • A lack of trust in AI and IoT might slow their implementation; awareness, transparency, and case studies are proposed as ways to improve trust.The passage frames trust as systemic in modern life and sustainable smart-city development.Source: §3.10 · p. 13
  • Data- and citizen-centric governance depends on trust, and participatory AI is argued to increase community trust.Smart-city data systems may generate revenue while needing to maintain people's trust.Source: §3.10 · p. 14
  • Many Koreans reportedly lost trust in smart-city development through u-City experiences, though new technologies may change the situation to some extent.Source: §3.10 · p. 14
  • The review identified 10 challenges and flags efficient, secure algorithms as an area requiring more attention.Algorithms and smart devices are described as relevant across factors including infrastructure, water management, and health education.Source: §3.10 · p. 14

4. Research Methodology

The study uses DEMATEL to examine interrelationships among challenges to AI and IoT adoption in sustainable smart-city development. Its three-stage process moves from literature review to expert input and DEMATEL analysis.

4. Research Methodology

  • DEMATEL is used to identify interrelationships among challenges hindering AI and IoT adoption in sustainable smart-city development.The method is described as a decision-making technique that considers expert attitudes and has been used to analyze barriers and challenges across disciplines.Source: §4 · p. 14
  • The methodology first identifies challenges through a PRISMA-based literature review, then collects industry-expert information, and finally uses DEMATEL to analyze and rank relationships.The study consulted 10 experts from industry, government organizations, and universities.Source: §4 · p. 14
  • Experts were selected by convenience sampling, with relevant degrees or experience and preference for research interests in AI, smart cities, or IoT.Data were collected from July to August 2021.Source: §4 · p. 15

Various Steps in DEMATEL Methodology (Stage 3)

The study turns expert pairwise ratings into thresholded DEMATEL relationships and separates challenges into cause and effect groups. Infrastructure, capital, cybersecurity, and trust are identified as the main challenges affecting implementation.

Challenge identification

Step 1: Formation of a direct relationship matrix (DRM)

Step 2: Formation of a normalized direct relation matrix (NRM)

Step 3: Calculation of the total relation matrix (TRM)

Directed relationships

Results

Directed relationships

  • The relationship graph distinguishes strong, moderate, and weak links with solid red, dashed red, and dotted black lines, respectively.The authors recommend prioritizing strong relationships when considering how to address adoption issues.Source: Various Steps in DEMATEL Methodology… · p. 21

5. Results and Discussion

DEMATEL classified four challenges as causes affecting six others; insufficient funds had the greatest D-R score. The analysis also identified specific interactions and effect-group priorities among the challenges.

5. Results and Discussion

  • Lack of infrastructure (C1), insufficient funds or capital (C2), cybersecurity risks (C3), and lack of trust (C10) formed the cause group, affecting the other six challenges.The effect group comprised smart waste and hygiene management (C4), lack of professionals (C5), managing energy demands (C6), managing transportation (C7), environmental risks (C8), and managing public health and education (C9).Source: §5 · p. 22
  • 1.422 was the greatest D-R score, for insufficient funds or capital (C2); infrastructure (C1) had the second-highest effect degree, 10.462.The authors identify funding as the most important challenge and infrastructure as requiring further consideration during implementation.Source: §5 · p. 22
  • Lack of trust in IoT and AI (C10) had the second-greatest D-R value, while cybersecurity risks (C3) had a D-R value of 0.3352.The paper notes that connected systems and extensive public data make smart cities vulnerable to cyberattacks.Source: §5 · p. 22
  • Environmental risks (C8) had the highest R value, 10.641, and a D-R value of −1.2071; managing transportation (C7) had the procedure's second-largest D+R.Although C8 was in the effect group, the authors describe its influence on other elements as substantial.Source: §5 · p. 22
  • Managing energy demand (C6) and public health and education (C9) had R values of 10.359 and 10.292, and D-R values of −1.122412 and −0.8052, respectively.The authors consider both challenges priority candidates; lack of professionals (C5) had a D-R value of −0.0624.Source: §5 · p. 22
  • The causal-interrelationship graph showed strong interaction between lack of professionals (C5) and managing energy demands (C6); transportation (C7), environmental risks (C8), and public health and education (C9) were also highly related to other challenges.The graph was based on M4 matrix results, with arrow-headed lines indicating causal interactions.Source: §5 · p. 23

6. Conclusion and Future Discussion

The study used a systematic literature review and expert-based DEMATEL analysis to examine barriers to AI and IoT adoption in sustainable smart cities. It identifies four dominant challenges and describes limitations in the expert-opinion analysis.

6. Conclusion and Future Discussion

  • Infrastructure, insufficient funds, cybersecurity risks, and lack of trust were the most prominent barriers affecting AI and IoT adoption in sustainable smart city development.The study places waste management, professional capacity, energy demand, transportation, environmental risks, and public health and education among the challenges highly affected by other factors.Source: §6 · p. 24
  • The study used PRISMA to identify ten challenges from 171 articles and assessed them with DEMATEL using opinions from ten smart-city experts.The authors say the experts had experience in smart city development and came from diverse industries and organizations.Source: §6 · p. 24
  • The authors present the findings as a basis for understanding barrier interrelationships, with potential usefulness for smart-city projects in developing countries such as India, Brazil, and Malaysia.The study itself is based on data from China.Source: §6 · p. 24
  • The study did not use a Grey scale to measure the accuracy of expert opinions; the authors suggest Grey-DEMATEL with more challenges and other methods for future work.They also suggest consulting more experienced experts to help identify possible solutions.Source: §6 · p. 24
Loading 2205.01067v1…