Written September 2020, revised 2026.
Written about smart-city programs, though the argument is really about critical infrastructure anywhere: efficiency and resilience are different design goals, and optimising for the first does not deliver the second.
From the beginning of human civilization, cities have endeavored to change environmental, financial, and social aspects of urban life. Smart cities encompass the transformative spirit of this evolution: from a number of small, discrete initiatives that used technology to improve residents’ lives, smart cities have evolved to a sizable market opportunity that encompasses investments in government transparency, large-scale IoT integration projects, and public safety. With the notion that common frameworks may increase efficiency, promote inter-agency data sharing, and reduce IT spend for common projects tied to smart missions, smart city components have become an essential part of urban critical infrastructure. Smart initiatives notwithstanding, urban areas with high population concentrations remain particularly vulnerable to predictable disasters, including diseases and epidemics, along with earthquakes, tsunamis, volcano eruptions, sea-level rise and flooding, food shortages, wildfires, extreme heat, hurricanes, tropical storms and typhoons, terrorist attacks, civil unrest, cyberattacks, war, nuclear or chemical contamination, or extreme air pollution. Additionally, as the COVID-19 crisis has shown, cities are not only vulnerable to loss of life, but also to a loss of economic output and value. To combat these unknowns, cities have focused on emergency management and redundancy. But with integrated technology and AI-driven analytics, a smart city framework should evolve to include discussion on urban resilience.
Smart City Framework
On a global scale, smart cities are the most densely populated areas in the world. With some major city locations being home to more than 10 million people, research shows that they take up more than 60 percent of the world’s energy consumption. In fact, according to the United Nations Department of Economic and Social Affairs, 68 percent of the world’s population will inhabit cities by 2050. Initiatives to make city programs “smarter” are on the rise, and some cities have already begun the intricate process to develop and deploy a framework that supports smart initiatives and help them align with urban strategic objectives. Strategic urban objectives in themselves are often focused on the intractable challenges urban development initiatives commonly face, and smart projects find themselves in unique positions to solve. For instance:
These issues, and others, can be mitigated through the adoption of smart, scalable solutions (S3) that take advantage of smart technologies. At its core, a smart city framework is a simple decision methodology that enables both the public and private sectors to plan and implement smart urban initiatives to respond to the above and provide an increase of efficiencies for future projects, reduction of costs for ongoing initiatives, and to full-fill the primary promise of smart cities: to enhance a citizen’s quality of life.
Resilience Framework
In a framework analysis by Dr. Roland Kupers and Song Hsi Ching “A Resilience Framework for Smart Cities,” Kupers and Ching state “The stresses a city faces can be different in nature: they may be chronic or acute, exogenous or endogenous, natural or social, knowable or unknowable. The system can be a bungee cord, someone’s body, a company or a city, etc.”. Smart city initiatives are designed to help monitor and manage the potential for these disasters using innovative technology and the Internet of Things (IoT). They give government and city agencies the ability to better respond to emergencies and plan for long-term growth and sustainability.
Many cities have become acutely aware of the potential for predictable and unpredictable disasters and have taken the steps toward adopting smart city initiatives. Over the last decade, various countries have begun the creation of smart city projects. However, as the IoT and its technology continue to expand, cities and government agencies worldwide are finding a need to create a more resilient smart city infrastructure, and they’re updating many of the technologies currently in place. In fact, some have already appointed chief resilience officers responsible for cross-developmental risk assessment, organization of readiness and preparedness, and putting in place technology-enabled manual and automated response management.
The technology behind smart cities and IoT is critical for any preparation against the unexpected, so updates and improvements are necessary whenever possible. Major areas like Kansas City, Missouri, and Las Vegas, Nevada, are implementing multiple-phase updates to improve their more dated technologies. Kansas City, for example, launched its smart cities plan back in 2016. The plan consisted of various phases that would update the city’s Wi-Fi access points and over 600 traffic sensors over a nine-mile long span. In Las Vegas, smart city improvements focus on public safety and transportation, utilizing high-definition video cameras, sound and motion sensors, and other IoT devices.
These updates and improvements are essential for sustainable smart city infrastructure and aren’t limited to only public safety and transportation. Processes to leverage robotics for the delivery of goods and transportation, liberating additional capacity for housing, mobility, and freight during emergencies, building automated and flexible production lines and supply chains, and massively deploying online and remote capabilities for education, healthcare, business, and entertainment are also other ways to improve resilience in smart cities, and will go a long way in being better prepared for disasters.
A Layered Response to Resilience
Most cities establish resilience criteria in an intuitive way rather than as a clearly structured component in an established framework. A structured method to include resilience allows for overlapping efficiencies in smart infrastructure, but also provides transparencies into the “why’s” and “how” smart cities use citizen data. As city leaders define actions or initiatives by their impact on stated city objectives, it may be helpful to look at adding resiliency in four layers congruent to the intractable challenges we mentioned previously. Additionally, a four-layer framework provides a logical flow that enables stakeholders to “test” initiatives for proof of value while engaging in a feedback loop that enables stakeholders to deploy best practices, thus aligning urban objectives with potential smart projects.
A Resilient Global Supply Chain
Many smart city initiatives are reliant on global supply chain, IoT suppliers, and IT contractors. As emergencies start to affect urban development, some of the hardest-hit areas are manufacturing and global supply chain. Smart city frameworks must be able to account for the slowdown or potential shut-down of global trade activity for IoT hardware, or technological components that make up the backbone of smart technologies.
Additionally, reliance on foreign markets such as China may see some of the greatest impact due to reduced manufacturing of IoT devices in case of a global emergency (such as we say with the global pandemic), or security issues or strict advisories provided by the Federal government. Near term, China will see some of the greatest impact due to reduced manufacturing of IoT devices because China has the largest IoT device shipments and connections today. Additionally, the supply of products in markets such as China may be even more reduced as these markets may be “…the recipients of a “one-two” punch led by the trade war for Chinese IoT suppliers as companies reassess their supply chains and dependence on Chinese products” (ABI Research). For many cities, supply chain assessments have already been in motion due to the trade war. But now is a good time to formalize those assessments and establish alternatives in various IoT markets, either for the potential supply slowdowns or shutdowns due to reduced shipping volumes or trade wars or Federal policies associated with technology supplied by foreign vendors.
Update — September 2026
This essay was written before the models that would change the question. In 2020 a smart city was a set of instrumented systems, each with its own dashboard and its own operator. What has since become plausible is a coordinating layer — a superintendent intelligence sitting above traffic, utilities, emergency response, permitting and public safety, reasoning across all of them rather than reporting each one separately.
That is no longer speculative architecture. Agentic orchestration moved from isolated pilots into production at large regulated organisations during 2026, and the design patterns now under discussion — task decomposition, shared state, hierarchical coordination — are precisely the shape a city-scale superintendent would take. McKinsey’s April 2026 analysis puts 60–80% of routine infrastructure work within reach of agentic automation.
And the failure mode is the one this essay was already about. Gartner projects that more than 40% of agentic AI projects will be cancelled by 2027 — not for want of capability, but for governance gaps, fragmented agent inventories, unclear auditability and poor integration. That is the efficiency-versus-resilience argument below, arriving on schedule in a new domain. A superintendent AI that cannot show its working, cannot be audited, and has no defined behaviour when a subsystem fails is not resilience. It is a single point of failure with better throughput.
The question a city should ask of such a system is not how much it automates. It is what happens on the day it is wrong, and whether anyone can reconstruct why.
Sources: Agentic orchestration in the enterprise, April 2026 · Multi-agent orchestration patterns, 2026
What we built from this. Critical infrastructure is one of the threat vectors DeepWatch covers, alongside health, thermal, natural disaster, severe weather, terrorism, cyber, transportation and public sentiment — and it is built to keep running at forward and disaster-impaired sites, reconciling when the link returns. Resilience is not a feature added afterwards. Situational Awareness.
An earlier version of this essay was published on Medium in September 2020. This version has been revised.