Myths vs. realities: Smart cities & connected infrastructure — what the evidence actually supports

Barcelona's smart city program, one of the most cited examples globally, reported cumulative savings of EUR 75 million per year by 2024 from its IoT-connected water, lighting, and waste management systems, according to the city's Digital Transformation Office. Yet a 2025 McKinsey Global Institute analysis of 87 self-described smart city initiatives worldwide found that only 28% delivered measurable outcomes within their original budget and timeline. The gap between smart city marketing narratives and operational evidence is large enough to mislead procurement decisions, misallocate public capital, and delay genuinely impactful urban technology investments. This article examines the most persistent myths in the smart cities space and sets them against what the data actually shows.

Why It Matters

Global spending on smart city technologies is projected to reach $203 billion by 2027, up from $130 billion in 2023, according to IDC's Worldwide Smart Cities Spending Guide (IDC, 2025). In North America alone, over 400 municipalities have initiated at least one connected infrastructure project covering domains from traffic management and building energy optimization to water network monitoring and public safety analytics.

The stakes are high. Smart city deployments typically require 7 to 15-year concession agreements with technology vendors, sensor hardware with 10 to 15-year lifespans, and integration with legacy infrastructure systems that predate digital networking entirely. Decisions based on inflated vendor claims or misunderstood pilot results lock municipalities into expensive technology paths that may underperform for a decade or more. Sustainability professionals advising city governments, utilities, and real estate developers need a clear-eyed view of what works, what does not, and where the evidence remains genuinely uncertain.

Key Concepts

Smart city refers to urban areas that use IoT sensors, data analytics, and networked communication systems to improve the efficiency of city services, reduce resource consumption, and enhance quality of life. The term encompasses a wide spectrum from individual connected systems (smart streetlights, water leak sensors) to fully integrated urban digital platforms.

Connected infrastructure describes physical assets such as roads, bridges, water pipes, electrical grids, and buildings equipped with sensors and communication capabilities that enable remote monitoring, predictive maintenance, and automated control.

Digital twin is a virtual replica of a physical city system or asset, continuously updated with real-time sensor data, used for simulation, scenario planning, and operational optimization.

Urban data platform is software infrastructure that aggregates, normalizes, and analyzes data streams from multiple city systems to provide cross-domain insights and coordinated management capabilities.

Myth 1: Smart Cities Slash Energy Use by 30% or More

The claim that smart city technologies reduce urban energy consumption by 30 to 40% appears frequently in vendor materials and conference presentations, often citing aggregated potential rather than measured outcomes.

Reality: Evidence from deployed systems shows more modest and highly variable results. A 2025 American Council for an Energy-Efficient Economy (ACEEE) evaluation of 42 US municipal smart building and smart grid programs found median energy savings of 12 to 18%, with a range from 3% to 27%. The higher end of this range was achieved only in buildings with poor baseline performance (Energy Use Intensity above 200 kBtu per square foot) where basic controls upgrades delivered large improvements regardless of "smart" features.

Smart streetlighting conversions, one of the most commonly deployed smart city technologies, deliver genuine energy savings of 40 to 60% when replacing high-pressure sodium fixtures with LED plus dimming controls. However, lighting typically represents only 2 to 4% of total municipal energy consumption, making the city-wide energy impact modest. Columbus, Ohio's Smart Columbus initiative documented a 52% reduction in streetlight energy consumption but a city-wide energy reduction of less than 1% from the lighting program alone (Smart Columbus, 2024).

Bottom line: Smart city technologies reduce energy use, but claims above 20% at the city scale require scrutiny. System-level savings of 10 to 18% are well-supported; city-wide claims of 30% or more typically rely on theoretical projections rather than measured results.

Myth 2: IoT Sensors Pay for Themselves Through Operational Savings

Vendors frequently claim that connected sensor networks achieve payback within 2 to 3 years through reduced labor, preventive maintenance savings, and avoided infrastructure failures.

Reality: Payback periods vary enormously by application domain. Water leak detection sensors have the strongest economic case: Washington, DC's DC Water utility deployed 4,800 acoustic leak sensors across its distribution network and documented a 38% reduction in non-revenue water losses within 30 months, representing $8.2 million in annual savings against a $12 million deployment cost, yielding a payback of approximately 18 months (DC Water, 2025).

By contrast, smart waste management systems (bin-level sensors plus route optimization) show much weaker economics. A 2024 National League of Cities survey of 34 municipalities that deployed smart waste systems found median cost savings of 8 to 14% on collection routes, but only after accounting for sensor replacement costs (sensors in waste bins have a 2 to 4-year lifespan due to harsh conditions), software licensing fees of $3 to $8 per bin per month, and integration labor. Net payback extended to 5 to 8 years for most deployments, and 11 of 34 cities reported negative ROI after 3 years of operation (National League of Cities, 2024).

Air quality monitoring networks present an intermediate case. Los Angeles' citywide air quality sensor network, comprising 700 low-cost sensor nodes, cost $4.5 million to deploy and $1.2 million per year to operate. The network provides granular pollution data that supports targeted traffic management and industrial compliance enforcement but generates no direct revenue, making "payback" dependent on how one values public health outcomes and regulatory effectiveness.

Bottom line: IoT sensor economics are application-specific. Water and energy monitoring show strong payback (1 to 3 years). Waste and environmental monitoring often require 5 or more years or depend on non-financial benefit valuations.

Myth 3: Smart City Platforms Integrate Seamlessly Across Departments

Marketing materials for urban data platforms emphasize unified dashboards, cross-departmental analytics, and seamless interoperability between city systems.

Reality: Data integration remains the primary technical and organizational barrier. A 2025 Brookings Institution study of 23 US cities with deployed urban data platforms found that the average platform connected data from only 3.4 of the 8 major departmental systems (transportation, water, energy, waste, public safety, buildings, parks, and public health) after 3 or more years of implementation (Brookings, 2025).

Kansas City's integrated smart corridor project along Prospect Avenue, launched in 2016 as a showcase for connected infrastructure, took 5 years to integrate traffic signal, streetlight, kiosk, and environmental data into a single platform. Even then, water and waste systems remained on separate networks. The project cost $15.7 million over its first 7 years, roughly double the original budget, with integration challenges identified as the primary cost driver.

The fundamental obstacles are not primarily technological: they are organizational. Different city departments use different data formats, operate on different procurement cycles, and answer to different regulatory frameworks. Legacy SCADA systems in water and electrical utilities often use proprietary protocols that predate modern API standards. The Open & Agile Smart Cities (OASC) network, which promotes the Minimal Interoperability Mechanisms (MIMs) standard, reports that fewer than 15% of member cities have achieved full MIM compliance across more than 2 departmental systems.

Bottom line: True cross-departmental integration takes 5 to 10 years and costs 2 to 3 times initial estimates. Practitioners should plan for phased integration starting with 2 to 3 domains rather than assuming platform-level interoperability from day one.

Myth 4: Smart Cities Are Inherently More Sustainable

The assumption that digitizing city systems automatically reduces environmental impact is pervasive but problematic.

Reality: The environmental footprint of the smart infrastructure itself is frequently omitted from sustainability calculations. A 2024 lifecycle assessment by the University of Toronto estimated that a mid-size city's smart infrastructure deployment (50,000 sensors, networking equipment, data center capacity, and display systems) generates 8,000 to 14,000 metric tons of CO2 equivalent over a 15-year lifecycle from hardware manufacturing, installation, operation, and disposal (University of Toronto, 2024). This carbon cost must be netted against operational savings to determine true sustainability impact.

Denver's smart traffic management system reduced average commute times by 12% and vehicle idling emissions by an estimated 9,400 metric tons of CO2 per year. However, the system's own energy consumption (data centers, communication networks, signal controllers) adds approximately 2,100 metric tons of CO2 per year, yielding a net reduction of 7,300 metric tons. While still a net positive, this is 22% lower than the headline figure typically cited.

E-waste from sensor networks presents an emerging concern. Smart streetlight sensors, environmental monitors, and communication nodes contain lithium batteries, circuit boards, and plastic housings with limited recycling pathways. San Diego's smart streetlight program, which deployed 4,200 sensor nodes, faced criticism when 1,800 nodes were decommissioned after a 2020 policy change, generating approximately 11 metric tons of electronic waste with no established municipal recycling protocol for IoT-specific components.

Bottom line: Smart cities can deliver net environmental benefits, but only when lifecycle carbon costs of the technology itself are included. Sustainability professionals should demand full lifecycle assessments, not just operational savings projections.

Myth 5: Cybersecurity Risks in Smart Cities Are Manageable with Standard IT Practices

Many municipalities treat connected infrastructure cybersecurity as an extension of existing IT security, applying enterprise IT frameworks to operational technology (OT) environments.

Reality: Smart city systems face fundamentally different threat vectors than enterprise IT. A 2025 CISA assessment identified 147 critical vulnerabilities across commonly deployed smart city platforms, with 34% rated as high-severity and exploitable remotely (CISA, 2025). Unlike enterprise IT systems where a breach may expose data, smart city attacks can manipulate physical infrastructure: traffic signals, water treatment chemical dosing, building HVAC systems, and electrical grid switchgear.

The 2021 Oldsmar, Florida water treatment attack, where an intruder briefly increased sodium hydroxide levels to 100 times the normal concentration, demonstrated the physical safety implications of connected infrastructure compromise. Atlanta's 2018 ransomware attack cost the city an estimated $17 million in recovery costs and disrupted municipal services for weeks.

Standard IT security practices such as firewalls, endpoint protection, and vulnerability scanning are necessary but insufficient. Smart city systems require OT-specific protections including: network segmentation isolating safety-critical systems from general city networks, continuous protocol-level monitoring for anomalous commands on industrial control networks, hardware-based authentication for sensor data to prevent spoofing, and air-gapped backup control capabilities for critical infrastructure.

Bottom line: Smart city cybersecurity requires dedicated OT security expertise and budgets separate from enterprise IT. Municipalities should allocate 8 to 12% of smart city capital budgets to security, versus the 2 to 4% that most currently spend.

What's Working

Cities that achieve measurable outcomes share common patterns. They start with a clearly defined problem rather than a technology-first approach. They deploy proven, single-domain solutions (smart water metering, adaptive traffic signals, building energy management) before attempting cross-domain integration. They negotiate vendor contracts that include performance guarantees tied to measured outcomes rather than technology deployment milestones.

Chattanooga, Tennessee's municipally owned fiber network and connected grid program stands out as a well-documented success: the city's EPB utility reduced power outage duration by 55% and achieved $50 million in cumulative operational savings over 8 years through smart grid automation (EPB, 2024). The key differentiator was municipal ownership of the fiber network, eliminating vendor lock-in and recurring licensing fees that erode ROI in many deployments.

What's Not Working

Large-scale, vendor-led smart city master plans that attempt to digitize everything simultaneously continue to underperform. Sidewalk Labs' Quayside project in Toronto, which proposed a comprehensive smart neighborhood with embedded sensors throughout infrastructure, was abandoned in 2020 after $50 million in development costs, citing economic viability concerns and public opposition to data governance practices.

Proprietary platforms that create vendor lock-in remain a persistent problem. Cities that deployed single-vendor smart city platforms report switching costs of 60 to 80% of original deployment cost, effectively eliminating competitive pressure during contract renewals.

Key Players

Established: Cisco (Kinetic for Cities platform deployed in 30 or more North American cities), Siemens (MindSphere urban operations deployed across European and North American installations), IBM (Intelligent Operations Center for Cities, legacy platform with declining market share), Itron (smart water and energy metering in 500 or more North American utilities)

Startups: Numina (computer vision for urban analytics, deployed in 40 or more US cities), Spot AI (infrastructure monitoring using existing camera networks), Cityzenith (digital twin platform for urban planning and carbon tracking)

Investors: Sidewalk Infrastructure Partners (spun out from Alphabet, focused on urban technology infrastructure), Infrastructure Ontario (smart city municipal financing), US DOT Smart City Challenge (federal funding for connected transportation systems)

Action Checklist

• Require vendors to provide measured outcome data from comparable deployments, not theoretical projections or pilot results
• Demand full lifecycle carbon assessments for any proposed smart city technology, including hardware manufacturing, operation, and end-of-life disposal
• Budget 8 to 12% of smart city capital costs for cybersecurity, with dedicated OT security expertise separate from enterprise IT
• Plan for phased integration across 2 to 3 city domains initially, with 5 to 10-year horizons for broader cross-departmental connectivity
• Negotiate vendor contracts with performance guarantees tied to measured outcomes and data portability clauses to avoid lock-in
• Establish baseline measurements for all target KPIs before deployment to enable accurate before-and-after comparisons
• Develop sensor end-of-life management plans including recycling or responsible disposal of IoT hardware
• Engage community stakeholders on data governance policies before deploying public-space sensing technologies

FAQ

Q: What is the realistic payback period for a city-wide smart infrastructure investment? A: Payback varies dramatically by application. Water leak detection and smart metering typically achieve payback in 1 to 3 years. Smart streetlighting delivers 3 to 5-year payback. Smart waste management extends to 5 to 8 years. Comprehensive cross-domain platforms rarely achieve financial payback within 10 years and are better justified on service quality and resilience grounds rather than pure cost savings.

Q: How should cities evaluate smart city vendor claims? A: Request outcome data from at least 3 comparable deployments (similar city size, climate, and infrastructure age). Require that data be measured, not modeled. Ask for total cost of ownership including software licensing, sensor replacement, integration labor, and data center costs over the full contract term. Independently verify claimed savings through third-party audit clauses built into procurement contracts.

Q: Are open-source smart city platforms viable alternatives to proprietary solutions? A: Open-source platforms such as FIWARE and CKAN are maturing rapidly and are deployed in production environments across 40 or more European cities. They reduce licensing costs and vendor lock-in but require internal technical capacity or contracted integration support. A 2025 Open & Agile Smart Cities analysis found that open-source deployments cost 20 to 35% less over 10 years than proprietary equivalents, but required 40 to 60% more internal IT staff time for maintenance and customization.

Q: What are the most common data governance mistakes cities make with smart infrastructure? A: The three most frequent governance failures are: deploying public-space sensors (cameras, microphones, environmental monitors) without clear data retention and access policies, failing to conduct privacy impact assessments before procurement, and neglecting to establish data ownership terms in vendor contracts. San Diego's smart streetlight controversy, where sensor data was accessed by law enforcement without public disclosure, illustrates the reputational and political risks of inadequate governance frameworks.

Sources

• McKinsey Global Institute. (2025). Smart Cities: Moving from Pilots to Scale. New York: McKinsey & Company.
• IDC. (2025). Worldwide Smart Cities Spending Guide. Framingham, MA: International Data Corporation.
• American Council for an Energy-Efficient Economy. (2025). Municipal Smart Building and Smart Grid Program Evaluation: 42-City Assessment. Washington, DC: ACEEE.
• Smart Columbus. (2024). Smart Columbus Program Final Report: Outcomes, Lessons Learned, and Recommendations. Columbus, OH: City of Columbus.
• DC Water. (2025). Acoustic Leak Detection Program: Three-Year Performance Review. Washington, DC: DC Water and Sewer Authority.
• National League of Cities. (2024). Smart Waste Management Systems: Municipal Deployment Outcomes Survey. Washington, DC: NLC.
• Brookings Institution. (2025). Urban Data Platforms in Practice: Integration Challenges and Outcomes. Washington, DC: Brookings Metropolitan Policy Program.
• University of Toronto. (2024). Lifecycle Carbon Assessment of Urban IoT Infrastructure. Toronto: Department of Civil and Mineral Engineering.
• CISA. (2025). Smart City Technology Vulnerability Assessment. Washington, DC: Cybersecurity and Infrastructure Security Agency.
• EPB. (2024). Smart Grid and Fiber Network: Cumulative Impact Report 2016-2024. Chattanooga, TN: EPB.