Download PDF Spring Bridge on Sustainable Smart Cities March 15, 2023 The world’s cities face increasing threats from natural disasters, aging infrastructure, traffic, and resource constraints. The articles in this issue examine smart infrastructure, sustainability, net zero carbon options, and autonomous driving, among other approaches to smart and sustainable cities. IT for Sustainable Smart Cities: A Framework for Resource Management and a Call for Action Friday, March 17, 2023 Author: Cullen Bash, Ninad Hogade, Dejan Milojicic, and Chandrakant D. Patel Smart cities represent profound and extensive opportunities to achieve sustainability through IT-enabled supply-demand management of resources. Sustainability has become a top priority of industry, government, and academia. Efforts to achieve it encompass a range of actions, from reducing energy consumption to mitigating the effects of negative externalities such as climate change (e.g., Smil 2022). Sustainable smart cities are seen as an important opportunity to drive this transformation at scale. We define smart cities as having automated, to a degree autonomous, infrastructure based on increasing use of data, knowledge discovery, inference, and control (e.g., Founoun and Hayar 2018; Heitlinger et al. 2019; Toh and Milojicic 2021). Smart cities represent a model for sustainability due to their capacity to continuously adjust to changing factors and optimize parameters to varying needs. Integration of information technologies (IT) in city-scale resource management can make cities sustainable through modeling, measuring, and managing to optimally provision resources. Introduction Cities have always been at the forefront of technology, and smart cities are the most recent case in point, with IT as the primary enabler. Cities are centers for goods exchange, typically located near rivers and close to major trade routes for access to critical resources. A city’s size drives the supply and demand of all the goods and activities required to keep the city functional: from food and energy to manufacturing and transportation, among other sectors. Cities also drive advances in the infrastructure around them, by carefully locating agricultural sources for food delivery (e.g., Smil 2019), distributing energy generation for minimizing transfer costs, and building major airports, highways, and ports for goods transfer. Cities are thus part of the much larger ecosystem and value chain (figure 1). IT has become an integral, ubiquitous part of everyday life (e.g., Santana et al. 2017). What electricity meant in the last century, IT means to humanity today. It drives almost every facet of daily life, spanning the sensors at the edge to cloud computing. High-performance computers play a key role in the analysis and control of city-scale physical infrastructure such as power, water, waste management, and transportation. All these technology solutions require standardized interfaces to enable economical management and interoperability. In this paper, we focus on how IT makes smart cities sustainable. We list factors influencing smart cities; define the supply-demand framework; present the smart city lifecycle, high-level design, implementation, and management; and conclude with insights and a call for action. We focus primarily on energy and carbon, but similar arguments can be made for other resources. Factors Many factors impact the evolution of cities. Some prominent ones are briefly presented here and elaborated in the subsequent sections. Economic: The availability of resources, especially for energy, and their transportation and supply chains (e.g., Barcelos et al. 2018) are critical for cities, connecting them to global and regional markets. Today such availability entails colocation with electricity resources as well as internet data centers. We focus on sustainability, but a comparable model can be applied to economics, which is an inseparable part of any sustainable solution. Economic underpinnings apply to the whole of this paper. Ecological: Increasing focus is on the impacts of cities on the Earth, clean energy, water consumption, and global warming. Large data centers such as those required to support smart cities must be evaluated for net zero energy (e.g., Heitlinger et al. 2019). Technological: The evolution of technology, in particular cyberphysical systems, is core to the very nature of smart cities (e.g., Perera et al. 2017; Santana et al. 2017). Technology solutions incorporate IT advances—for example, in computing, communications, data storage, data management, and data analysis—with physical systems. Technology is the instrument that people can use to achieve sustainable smart cities. A holistic supply-demand framework can help city planners prioritize the replacement of aging infrastructure with sustainable alternatives. Sociopolitical: Smart cities are part of their country’s political ecosystem (e.g., Poltie et al. 2020). Therefore, regulators are important in enabling technological changes and energy decisions. Sociopolitical implications motivate some of our “calls for action” in the last section of this paper. In the remaining text, we primarily focus on technological factors, while understanding that they are intrinsically motivated and driven by economic, ecological, and sociopolitical factors. Urban Supply-Demand Framework Especially in light of climate change, environmental sustainability is a key driver behind smart cities, and IT is now a key enabler of sustainability. For example, IT services on the demand side include those from both public and private sectors such as ride-sharing, city information delivered through social media, and health and financial services. We present a holistic supply-demand framework to address sustainability and smart cities. The framework is broad and full implementation may take decades as cities invest on a gradual basis. The call to action is therefore to get started in specific supply-side areas such as power, water, and waste management. The framework starts with supply-side resources, which we break into city-scale verticals—for power, water, waste management, transport, health care, and recreation, among other areas—overlaid with IT elements to enable supply-demand management. An example of integrated supply-demand is the use of IT to shape power usage by scaling power-consuming devices and by allocating workloads, like manufacturing jobs, to specific devices and turning others off to accommodate supply scarcity. Lifecycle Metrics In 2015 the Paris Agreement set the limit on global average temperature rise to 1.5°C from the beginning of the Industrial Revolution, and in 2018 the Intergovernmental Panel on Climate Change set a goal of achieving net zero CO2eq emissions by 2050 (IPCC 2018; United Nations 2015). Cities are major users of energy and, as such, are significant sources of carbon emissions. Integrating IT in city infrastructure can improve energy use and carbon emissions, but this integration will necessarily evolve over time through the retrofit of technology in aging infrastructure or the replacement of infrastructure at the end of its useful life. Decisions about how to prioritize either the retrofit or replacement of infrastructure to improve operational efficiency can be challenging and need to consider embedded resources, like energy and carbon, resulting from the upstream value chain (e.g., manufacturing, transportation) in addition to the use phase of the system. To aid in the decision-making process we propose a metric called net positive impact (NPI): Net positive impact (NPI) = (Value delivered in energy saved)/ (Available energy consumed over lifetime) where the numerator quantifies the energy saved through replacement of the existing system and the denominator is the estimated lifetime energy consumed by the replacement system. Note that the numerator includes only use phase energy consumption since the embedded energy is considered a sunk cost. Leveraging the NPI format, the net positive carbon impact (NPIc) is defined as: Net positive carbon impact (NPIc) = (Value delivered in carbon saved)/ (Carbon emitted over lifetime) Like NPI, the numerator in NPIc quantifies the carbon saved from the replacement of legacy infrastructure while the denominator accounts for the projected lifetime carbon emissions of the replacement system. The amount of time it takes for the operation of the replacement system to achieve an NPIc equal to 1 is termed the carbon payback period (CPP). As an example of how to utilize NPIc, consider the replacement of a legacy carbon-based energy source with a solar photovoltaic (PV) system. Although no carbon is emitted during the use phase of the PV system, there can be significant embedded carbon from the manufacturing process depending on the energy sources. Depending on the size of the PV system and its operational efficiency, the CPP could be greater than 10 years (Bash et al. 2023). Absent other factors, projects with a shorter CPP coupled with high NPIc might be prioritized over those with a higher CPP and lower NPIc. In practice, however, NPIc and CPP should not be used in isolation. Rather they are part of several deterministic factors that include lifecycle stage (i.e., retrofit vs. replacement), economics, overall environmental benefit and impact (i.e., a net reduction in carbon emissions, environmental damage), and technology readiness. A variety of energy sources will be necessary for powering the future energy grid. Figure 2 shows the capacity factors (defined as how often a plant runs at full power) for a variety of energy sources. The factors range from 90 percent for nuclear energy generation to 20 percent for solar, accounting for thermodynamic (in-)efficiencies in the various power generation processes (EIA 2022). These heterogeneous sources present several challenges for their large-scale integration in a city’s electric grid. For example, design should integrate the more stable high capacity factor sources like nuclear or natural gas with lower capacity factor renewable -sources like PV that are characterized by variable diurnal output. Operational challenges include management to shape the demand given the supply. Demand management is enabled by sensing, communications, and policy. This is the context in which we see the role of IT and digital technology in accelerating the integration and adoption of sustainable technology and practices at the city scale—achieving high NPI via the addition of sensing management systems to better control resource use in energy systems. Design and Implementation Design and implementation must be integrated with management (next section) to form a unified system for an energy-sustainable smart city framework (figure 3). Lifecycle and Carbon Metrics Analysis Sustainable cities require a thorough lifecycle analysis to ensure that system design considers the type of energy use (brown vs. green), carbon emissions, and environmental impact. Metrics like NPI and NPIc can be used to help make retrofit and replacement decisions according to the system’s lifecycle stage. As an example, rather than replace an existing system, high NPIc might be achievable through the addition of sensors and analytics that yield insight into the operation of a system. Low-Carbon Decentralized Resource Microgrids Clean or green energy sources such as hydroelectric, PV, solar thermal, biogas, and wind can be an important part of low-carbon energy generation. And zero-energy systems (or buildings), in which the energy both generated and consumed are the same quantity (hence the net consumption can be considered zero) (e.g., Marszal et al. 2011), are made possible through integration with microgrids. Instead of a centralized production model with large distribution and transmission networks, a more distributed model with local resource microgrids should be considered (e.g., Liu et al. 2018), drawing on multiple local sources, such as rooftop PV cells, hydro-electric power plants, and constructed reservoirs. Such a decentralized resource microgrid model is scalable and modular and makes it possible to configure, upgrade, and add more systems as the city grows. Efficient Systems Communication via IT The availability of transmission routes between smart grids, as well as the supply and demand of resources from various geographical regions, can all be determined and managed using an efficient communication infrastructure, which can also provide information such as resource-specific operational characteristics, performance, and sustainability metrics. The information and communication infrastructure, which gathers data on energy consumption and disseminates information about provider rates, underpins the framework for an energy-sustainable smart city. For smart appliances like dishwashers and water heaters, IT can be used to control operations with the right level of energy consumption. It can also be used to buy energy from a variety of sources, including wind turbines, solar farms, and brown energy generation systems. Smart Grids and Smart Metering Smart grids efficiently integrate the actions and behaviors of all connected consumers and generators as well as various energy sources, from fossil fuel–based thermal energy to PV and wind (e.g., Gao et al. 2012). They use IT to assist with demand-response energy management, schedule power generation for various electricity generation plants, and manage dynamic interaction between plug-in electric vehicles and the power grid. Smart energy metering records electricity consumption at predetermined intervals and transmits the data to the utility for billing and monitoring (e.g., Kabalci 2016). This facilitates accurate reading of use without human involvement. In these ways smart grids guarantee reliable, cost-effective, and sustainable energy systems with low levels of loss, and improved safety, security, and fault tolerance. Management Data Analytics, Machine Learning, and Visualization To change the operational state of the systems toward least energy operations, sophisticated data analytics and machine learning techniques can be applied to streaming and archival data to identify patterns and predict trends. Data analytics can be used to build models for fault detection, optimization, and control. Data can be visualized and high-level indicators of the health of each energy system can be monitored. End-of-Life Replacement, Recycling End-of-life replacement decisions include when to replace older systems to improve NPI or NPIc. Recycling policies can be proposed to reduce the carbon footprint of a system or product. Policy-based Control and Operation Policy-based control and operating systems can be devised to promote efficient energy management, achieve sustainability goals (e.g., selection of green energy resources, increased NPIc), and maximize the use of power from renewable resources and thus drive to net zero. As an example, given a sustainability policy, if an onsite PV solar farm at a manufacturing factory generates more energy than required, it can sell the excess energy back to a smart grid (utility company) and help balance the demand and supply of electricity/workload (e.g., Hogade et al. 2018). And the sustainability policy at the same manufacturing plant may allocate certain demand-side fabrication workloads based on local solar power alone. Insights The ability to achieve sustainable smart cities requires configurable design and implementation using digital technologies to integrate supply-demand management that shapes demand commensurate with supply. For example, in IT-enabled digital factories built with 3D printers, workload allocation and execution profiles can be shaped based on supply-side power from the microgrid (Patel 2020). In all cases, solutions are substantially more manageable when applied to specific city-scale verticals (shown in figure 4). Next, we advocate that industrial applications be holistically instrumented using operational technologies (OT) that interface with cyberphysical systems, are managed and secured through supply chain integration, and are protected using IT that interfaces with digital systems. This would guarantee reliable and regulated supplies from trusted and sustainable (ecologically, economically, politically, and socially) sources. The use of loosely integrated OT/IT workflows will be a basis for automation. The complexity of our vision will not be possible to realize with traditional management techniques. AI-based operations will be required to supplement autonomous solutions to enable anomaly detection and a subtle yet vital combination of full autonomy with just enough human engagement to avoid damaging bugs, attacks, and oversights (e.g., Dang et al. 2019; Laplante et al. 2020; Serebryakov et al. 2021). To augment our vision and increase confidence in smart cities, a combination of simulations—-including digital twins (virtual functional representations of physical systems)—will be needed. Connected with the city’s instrumented critical infrastructure and inputs/outputs, such simulations will enable scenario testing, what-if analysis, planning, and prediction for both regular operations and upgrades to the ecosystem. This is critical for catastrophic situations such as pandemics, earthquakes, tsunamis, wars, and other disasters. IT is a key enabler to make smart cities much more resilient under devastating circumstances, when human attention is diverted to survival. In a smart city, cybersecurity will become cyber-physical security. A holistic approach that applies learning from cybersecurity models will prevent bad actors from harming critical supply-side systems such as pumps, motors, turbines, and other equipment. Crippling attacks on power stations could be at least attenuated if not prevented using cyberphysical security. Finally, with significantly more automation and autonomy than today’s cities, smart cities will require human capital with expertise spanning IT, sustainability, and the vertical resource domains shown in figure 5 to operate effectively. Call for Action Many revolutionary technologies will directly influence the evolution of sustainable smart cities (figure 5), in layers of devices, interfaces, networks, software, applications, and markets. Each layer can make a huge difference. But technology alone is not sufficient. For smart cities to be successful, governments, industry, and academia need to take a holistic view that integrates energy, water, climate, carbon emissions, and other critical factors at the city, regional, and global ecosystem levels. To that end, action is needed for the following: Governments: Distributed energy resources require regulations to ensure the safest impacts for climate, carbon emissions, water, energy consumption, and other areas. This needs to be done transparently in attributing and accounting for resources for near- and long-term consumption. Industry: Smart cities require the introduction and adoption of standardized and application programming interfaces to optimize resource management. End-to-end management and regulatory compliance-based rebalancing of resources will require increased automation, analytics, cybersecurity, and the use of AI. Industry: The lifecycle impact of a city’s infrastructure systems needs to be considered when making retrofit or replacement decisions. Metrics like NPI can help guide such decision making as well as effective diversification of systems for risk management. Metrics are also crucial for informing policies for economic, ecological, and social trade-offs and for projections and what-if analysis for long-term resource supply and demand. Academia: Students need to learn about cyberphysical systems and their integration with IT, OT, and AI. While automation and AI techniques will lead to increased autonomous solutions, engineers will need to profoundly understand the underlying technologies and be able to continuously elevate them to the next levels of efficiency and sustainability. Smart cities, through the careful integration and management of technology in the overall ecosystem, will enable sustainability. Technological innovations, together with the items in our Call for Action, will make possible this vision, indeed this inevitable evolution. References Bash C, Hogade N, Milojicic D, Patel C, Rattihalli G. 2023. Sustainability megatrend. IEEE Computer 56(1):125–32. Barcelos M, Barcelos A, Bernardini F, Vaz Silva G. 2018. Analyzing the use of economic and financial indicators in smart cities context. Proceedings, 19th Annual Internatl Conf on Digital Government Research: Governance in the Data Age, May 30–Jun 1, Delft. Dang Y, Lin Q, Huang P. 2019. AIOps: Real-world challenges and research innovations. 41st Internatl Conf on Software Engineering, May 25–31, Montreal. EIA [US Energy Information Administration]. 2022. Electric Power Monthly, Table 6.07.B. Capacity factors for utility scale generators primarily using non-fossil fuels. Washington. Online at https://www.eia.gov/electricity/monthly/epm_table_grapher. php?t=epmt_6_07_b. Founoun A, Hayar A. 2018. Evaluation of the concept of the smart city through local regulation and the importance of local initiative. IEEE Internatl Smart Cities Conf, Sep 16–19, Kansas City. Gao J, Xiao Y, Liu J, Liang W, Chen CP. 2012. A survey of communication/networking in smart grids. Future Generation Computer Systems 28(2):391–404. Heitlinger S, Bryan-Kinns N, Comber R. 2019. The right to the sustainable smart city. Proceedings, ACM CHI Conf on Human Factors in Computing Systems, May 4–9, Glasgow. Hogade N, Pasricha S, Siegel HJ, Maciejewski AA, Oxley MA, Jonardi E. 2018. Minimizing energy costs for geographically distributed heterogeneous data centers. IEEE Transactions on Sustainable Computing 3(4):318–31. IPCC [Intergovernmental Panel on Climate Change]. 2018. Global Warming of 1.5°C: Summary for Policymakers. Geneva. Kabalci Y. 2016. A survey on smart metering and smart grid communication. Renewable & Sustainable Energy Reviews 57:302–18. Laplante P, Milojicic D, Serebryakov S, Bennett D. 2020. Artificial intelligence and critical systems: From hype to reality. Computer 53(11):45–52. Liu Y, Zuo K, Liu XA, Liu J, Kennedy JM. 2018. Dynamic pricing for decentralized energy trading in micro-grids. Applied Energy 228:689–99. Marszal AJ, Heiselberg P, Bourrelle JS, Musall E, Voss K, -Sartori I, Napolitano A. 2011. Zero energy building–A review of definitions and calculation methodologies. Energy & Buildings 43(4):971–79. Patel CD. 2020. Cyberphysical integration. The Bridge 50(4):21–23. Perera C, Qin Y, Estrella JC, Reiff-Marganiec S, Vasilakos AV. 2017. Fog computing for sustainable smart cities: A survey. ACM Computing Surveys 50(3):1–43. Poltie BM, Udoh ES, Luna-Reyes LF. 2020. Understanding smart city projects and priorities across large, medium and small cities in the United States. 21st Annual Internatl Conf on Digital Government Research, Jun 15–19, Seoul. Santana EFZ, Chaves AP, Gerosa MA, Kon F, Milojicic D. 2017. Software platforms for smart cities: Concepts, requirements, challenges, and a unified reference architecture. ACM Computing Surveys 60(6):1–37. Serebryakov S, Janson J, Cader T, Nanjudaiah D, -Subrahmanya J. 2021. Leveraging AI/ML for anomaly detection in system management. Proceedings, CVray User Group, https://cug.org/proceedings/cug2021_proceedings/by_sub_ type.html#pap121. Smil V. 2019. Growth: From Microorganisms to Megacities. MIT Press. Smil V. 2022. How the World Really Works: The Science Behind How We Got Here and Where We’re Going. Penguin Random House. Toh CK, Milojicic D. 2021. Recent advances in smart cities. IEEE Consumer Electronics 10(6):67–68. United Nations. 2015. Paris Agreement: Framework Convention on Climate Change. Bonn. About the Author:Cullen Bash is vice president and lab director, Ninad Hogade is an engineer, and Dejan Milojicic is a distinguished technologist, all at Hewlett Packard Labs. Chandrakant Patel (NAE) is chief engineer and senior fellow, HP Inc.