by Anna Gerber | Updated January 3, 2018 - Published September 20, 2017
This article is part of the IoT 201 learning path, a next-steps developer guide for IoT.
Connected cities emerge when IoT technologies are applied across an entire metropolitan area. When you think of connected cities, you might think of large cities that have more high-profile smart cities initiatives like London, New York, Chicago, Rio de Janeiro, or Amsterdam. However, small towns can also benefit from connecting people, services, and infrastructure. In this article, I’ll explore connected cities and some of the challenges that are involved in developing city-wide IoT solutions.
Many cities and towns around the world are turning to IoT to solve urban problems, such as traffic congestion, and to improve the safety and quality-of-life of their citizens. Smart sensors that are installed throughout the city, in vehicles and buildings, and apps and devices that are used by people who are living or working in the city produce data that is used throughout these connected cities. The IoT data is used to inform decisions on how public spaces are designed, how to make the best use of resources, and how to deliver public services and utilities more efficiently and effectively.
Some of the key issues that are being addressed by applying IoT technologies at a metropolitan scale include:
IoT technologies that are being applied to solve energy management issues within connected cities include smart grid technologies, smart metering technologies, and smart street lighting platforms.
Smart grids (and smart grid technologies) make electricity delivery more efficient by applying predictive analytics to data that is collected from sensors that are installed throughout the grid in order to match capacity with demand. Smart sensors monitoring the grid are typically connected to neighborhood area networks (NANs) or Low-Power Wide-Area Networks (LPWAN) through networking technologies like SigFox, LoRa, NB-IoT, or LTE-M.
These sensors include temperature sensors and phasor measurement units (PMUs), which measure current, voltage, and frequency of the electrical signal. These sensors are used to monitor the efficiency of renewable energy generators that are connected to the grid, such as solar panels or wind turbines, and to identify where to place generators to maximize the energy that is generated. Data from sensors on generators, transmission lines, cables, transformers, and substations is also used by providers to detect faults and to determine when maintenance should be scheduled.
Smart meters (and smart metering technologies) that are installed in homes and smart buildings allow energy usage to be monitored remotely and for supply to be controlled remotely, which leads to cost savings over manual meter reading and switching. Sensor components that are incorporated into smart meter devices include hall sensors, accelerometers, shock sensors, anisotropic magneto resistance (AMR) sensors, and PMUs. These sensors monitor energy usage and efficiency, monitor the health of the smart meter device itself, and also detect tampering of any of the devices. Consumers benefit from real-time energy monitoring when the data that is produced by these sensors is aggregated and presented through in-home display devices, visualization dashboards, and reporting dashboards that are integrated into mobile or web applications. These dashboards and apps allow them to track costs and consumption patterns, enable them to identify activities and appliances that use the most energy, and to modify their behavior in response to these data analyses.
In combination with smart appliances that have built-in actuator components such as relays that act as remote switches, smart meters assist with managing load. For example, energy-hungry devices like pool pumps or HVAC (heating, ventilation, and air conditioning) systems are automatically switched to run at off-peak times, which helps to prevent outages and brownouts and saves money for consumers through off-peak tariffs. Similar programs are being rolled out for other metered utilities like water and gas, too. For example, the city of Barcelona adopted smart water meters. With these smart water meters, the city can apply data mining and analytics along with real-time visualization and reporting tools that use the sensor data that is produced by the smart meter devices to better inform consumers, which has led to more efficient water usage and ultimately to cost savings for citizens.
Read more about the Chicago Smart Lighting project, which implemented smart street lights, on the Chicago Infrastructure Trust site.
In public spaces, energy-efficient, LED-based smart street lights, such as Cisco’s Smart+Connected Lighting, Philips connected-lighting, or Silver Spring’s street lights and sensors, have been tried or installed in hundreds of cities and towns around the world, from Barcelona (Spain) to Adelaide (Australia). Over 300 million street lights are operating around the world. These smart LED street lights results in significant energy savings not only because power draw of LEDs over traditional street lighting is reduced but also because the lights can be centrally controlled and the brightness of the lights can be adjusted based on whether people or traffic are nearby. These adjustments are achieved by analyzing data from proximity and motion detection sensors, such as passive infrared sensors (PIR), ultrasonic sensors, or microwave (Doppler) sensors, or by applying computer vision algorithms to detect vehicle or pedestrian presence by using live video streams from cameras. Citizens can also opt in to provide location data from GPS trackers that are built into their mobile phones or connected cars.
Smart lighting platforms often provide the backbone for connecting other sensors across a connected city, typically implemented as a wireless sensor network (WSN).
Environmental sensors are used to monitor public waterways, parks, and green spaces, and the sensor data can be used to identify spaces that require cleanup or protection. These environmental sensors are also used to track ambient environmental conditions at locations throughout the city, such as temperature, humidity, rainfall, and most notably air quality.
Environmental sensors are often rolled out by adding additional sensor components to extend the capabilities of the smart sensor source nodes within the wireless sensor network (WSN) that is provided by the smart grid or street lighting platform. In a typical WSN, the smart sensor nodes are low-power microcontroller-based devices, which are powered by batteries or solar cells and connected through a mesh network that uses 6LoWPAN plus IEEE 802.15.4 or RF networking standards. Mesh topologies, where the sensor nodes are interconnected, and are all involved in communicating data through the network, allow the range of the network to be extended, while also increasing the reliability and self-healing capability of the network.
In an urban environment, wireless sensor networks are prone to interference that is triggered by weather conditions like rain and fog and from reflective surfaces on buildings and water that cause signal interference as a result of multi-path fading when the signal takes multiple paths. The redundant paths provided by the mesh network topology allow the network to adapt by intelligently routing traffic around these problems. Also, channel hopping techniques can be adopted so that environmental (and other) sensor data can be propagated upstream to cloud services for processing, storage, and analysis.
Air quality sensors help tackle air pollution problems that many cities face, arising from vehicular or industrial emissions. Emissions can be monitored directly through CO2 sensors installed on vehicles. The data that is collected from the air quality sensors that are attached to the wireless sensor network nodes is communicated over the mesh network, and through gateway devices to cloud services that analyze the data. The data is analyzed across batches of data to provide historical reporting and insights, but can also occur in real time by using stream analytics services that are offered by IoT Platforms (you can see a demo of how you can use Apache EdgeNet, IBM Watson IoT Platform, and the IBM Streaming Analytic service to implement streaming analytics in your IoT solution). These services enable air quality incidents to be predicted through real-time analysis of sensor data, which allows early warnings to be issued so that people can avoid the most polluted areas, which helps to improve the health and well-being of the citizens who live or work in the affected areas. Analysis of air quality data coupled with emissions data can also be used to reroute traffic to prevent emissions building up in those areas of the city.
Managing waste is another area where sensor data is used to reduce costs and improve efficiency in a connected city. Sensors can be retro-fitted as part of existing waste disposal processes. For example, connected cities can add cellular-based smart sensors to trash cans so that trucks can be scheduled to collect trash only when they require emptying, and can use on-street sensors or computer vision algorithms over camera feeds to identify areas where litter builds up, where additional trash cans should be installed.
In Chicago, monitoring where garbage was building up and integrating data on weather and the location of empty buildings enabled data analysis that helped to predict where rats were nesting so that authorities could bait the areas in advance. This implementation resulted in a reduced number of rats, and a 20% cost saving over the previous approach of baiting after complaints were lodged.
In greenfield connected city developments, like the South Korean city of Songdo, waste can be processed even more efficiently by eliminating manual collection of garbage altogether. Songdo requires citizens to tag different types of trash with coded RFID smart tags and uses a reader built into the automated pneumatic garbage disposal system so that each type of waste is drawn away without any manual collection or secondary sorting required, to be processed separately and either buried, recycled, or burnt as fuel based on the data encoded in the tag.
Connected cities improve the experience of commuters by analyzing data from road reporting systems including road sensors, roadside video cameras, and variable speed signs. Applying IoT technologies to solve transportation problems involves feeding the data that is gathered from sensors into analytics services to produce actionable insights that are used directly to trigger actuators that are connected to smart devices such as adaptive traffic signals, or applied indirectly, to inform decisions on policy and to streamline processes. In Songdo, this solution involves monitoring geolocation data from GPS trackers and RFID tags on vehicles, analyzing the progress of vehicles to detect incidents or congestion, and then directly adjusting traffic signals in real time to control the traffic flow and reduce delays.
Adaptive traffic signals have been adopted in cities around the world including Sydney, New Jersey, and Toronto. Historical analysis of traffic and road sensor data can also be used to adjust speed limits and tolls, which manipulates traffic flow in the longer term. In addition to being used to route traffic around incidents, sensors also report on the condition of roads and bridges so that maintenance can be scheduled when required.
Road reporting data from sensors and cameras can be used to manage on-street parking. For example, the data can be published through smart parking mobile apps that display available parking spaces, navigate commuters directly to the nearest available parking space, and manage payment of parking fees to make parking as painless as possible.
Read more about smart parking in this IBM blog, ” Parking in the smart city.“
Public transportation can be improved adaptively too, by using usage data from smart ticketing systems and route timings from sensors and GPS trackers on board the vehicles. This IoT solution can provide real-time reporting on service availability and on delays to commuters who are waiting at stations and stops. It can also adjust timetables in the longer term to more accurately reflect the recorded timings and can use analytics to predict demand for different services at different times of the day and adjust timings or introduce additional services to improve efficiency.
Read more in this IBM white paper, IBM Intelligent Operations Center for Emergency Management.
Data from sensor networks provides real-time visibility into what is happening in the city for law enforcement agencies and emergency responders to make better decisions. This situational awareness can be used for day-to-day prediction, for planning and what-if analysis, and, in times of crisis, to assist with rapid response to incidents. For example, road sensors that monitor traffic can be used in ordinary circumstances to route law enforcement vehicles around congestion. Or, in an emergency situation like a flood, the same sensor data indicates which roads have limited or no access (less traffic than usual) and can be used to prioritize which areas should be evacuated and which roads should be cleared and repaired afterwards.
The city-wide sensor network that hooks into the smart lighting or smart grid infrastructure often includes cameras that can be used to monitor availability of parking spots or to detect the presence of people in order to adjust lighting levels. These cameras can also be used by law enforcement agencies for surveillance by applying video search and analysis tools to the raw camera feeds. This data from cameras and sensors, combined with other sources such as content from social networks, can be analyzed by using machine learning and artificial intelligence techniques to predict when crimes might be about to occur.
Read more in this white paper: IBM Smarter Cities Public Safety—Law Enforcement
Many of the benefits of connected cities arise from applying cognitive computing to produce insights from data that is gathered from sensors and other instrumented data. However, cities are citizen-centric, so the data that is captured from sensors must be complemented by input from citizens. Mobile and web apps provide opportunities for citizens to engage with local government, and to communicate requests, provide feedback, or report faults with utilities and infrastructure – a form of crowd-sourcing called crowd sensing. De-identified and non-confidential data, such as crowd-sourced air quality observations over time, can be treated as public assets and published as open data. Adopting standard data formats for both citizen-contributed and sensor-generated data is also important to ensure that the data remains accessible to individuals and businesses to extract value.
Many of the challenges that are involved in developing connected cities are not purely technical challenges. Developing a connected city involves establishing partnerships, developing strategies and business models, and consulting with the community, before any technologies are rolled out. Some of the challenges that have been identified from existing connected cities projects include:
Large-scale IoT solutions need someone in a leadership role to champion the project and facilitate collaboration and communication between the stakeholders.“
Establishing a connected city involves gathering input from many stakeholders and gathering data from disparate sources including private and public sectors. However, this level of co-operation is only possibly by breaking down communication blockers and facilitating data sharing between cross-sector stakeholders. This process also involves developing a governance structure and a city plan so that all of the stakeholders are working together towards the same goals.
One of the lessons learned from connected cities like Amsterdam was the importance of appointing a coordinator — that is, a CTO; large-scale IoT solutions need someone in a leadership role to champion the project and facilitate collaboration and communication between the stakeholders. Also, connected cities initiatives succeed by fostering an inclusive, participatory culture, where all citizens are encouraged to take an active role in the decision-making process.
When teams make decisions and set priorities, they need to decide whether to initially adopt a people-driven or efficiency-driven approach. This process involves balancing requirements for maximizing efficiency and cost savings against the needs of citizens from all demographics. In large cities, the return on investment after teams introduce IoT technologies is likely to be high, due to economies of scale and the savings brought about by improvements in speed and efficiency. However, for smaller cities and towns, the investment in the infrastructure and technology that is required may take many years to pay off, especially as retro-fitting existing cities with smart technologies can be more expensive than developing greenfield projects.
One approach that is often adopted by existing connected cities initiatives has been to start with a pilot that is focused on application areas that provide immediate cost savings, like introducing smart lighting or smart grid technologies. Then, teams can iteratively apply the lessons learned and the savings that are achieved to inform and fund subsequent pilots that address other requirements.
Making sense of the data is key to the success of connected cities. Ensure that your IoT platform supports real-time and historical data analytics.“
I’ve discussed the benefits of adopting an IoT platform and listed general-purpose platforms, including IBM Watson IoT, AWS IoT, and Microsoft Azure IoT, in my IoT platform guide. Smart-city-specific IoT platforms include:
Scalability and resilience in IoT Platforms are important features to consider when teams develop robust IoT solutions at this scale. Making sense of the data is key to the success of connected cities. Ensure that your IoT platform supports real-time and historical data analytics, as many of the adaptive technologies that are adopted within connected cities, such as adaptive traffic signals or smart lighting, require real-time analytics. Also, consider platforms that support analysis of unstructured data sources, including video feeds from cameras, or text from citizen feedback and requests, in combination with structured data from the many types of sensors deployed around the city.
Designing the network architecture, and deciding on standard data formats and networking protocols to adopt, will also have an impact on the effectiveness of the IoT solution.“
Communications infrastructure, which can include cables, cellular towers, small cell network access points, and wireless access points, enables devices including smart sensors and actuators to communicate with gateways and gateways to communicate with cloud apps and services that provide analytics, rules, and storage to process the large volumes of data that is being produced by the sensors. As soon as the sensors are deployed across the city and start to generate data, the communications infrastructure can become a limiting factor, so this infrastructure often needs to be upgraded as part of the transition to a connected city. The load on the network, gateway devices, and services should be monitored so that the infrastructure can be scaled to meet network bandwidth and performance demands as more devices are deployed. Designing the network architecture, and deciding on standard data formats and networking protocols to adopt, will also have an impact on the effectiveness of the IoT solution.
You should adopt devices and IoT platforms that implement security best practices.“
Security challenges for connected cities include maintaining resilience against cyber-attacks including targeted attacks, ensuring compliance with regulatory frameworks, and maintaining the confidentiality and integrity of citizen’s private data.
One of the biggest challenges in maintaining resilience of connected cities systems is in securing the smart devices and sensors themselves. With so many heterogeneous devices, there are many potential points of vulnerability. Devices often include actuators that can trigger real-world behaviors (with potentially life-threatening consequences if things go wrong). Examples include adaptive traffic lights that allow traffic signals to be controlled remotely, which might lead to traffic accidents, HVAC systems that can be turned on or off, and smart electricity meters that can disconnect the supply of electricity to the premises remotely. Hence, connected cities must be designed with security as a priority from the ground up.
You should adopt devices and IoT platforms that implement security best practices: IoT Platforms provide services such as authentication and access control services for devices, users, and services, in addition to encryption to ensure confidentiality and data signing for data integrity. Device management is also an essential security-related IoT platform feature and should include automated, over-the-air update support so that deployed devices can be kept up to date and patched “en masse” should a vulnerability be discovered, and they can decommission devices when they reach end of life. However, you should assume that devices will be compromised and develop strategies and vulnerability management plans to ensure that even if they are compromised then you limit the amount of damage that an individual device might do. These strategies include securing the network architecture and cloud services and designing devices to fail safely.
Lastly, you should test the security of the systems continuously and apply activity logging, anomaly detection, monitoring, and analytics, including video and unstructured content analytics in order to detect physical security incidents. Develop strategies for how vulnerabilities are disclosed and managed, and establish partnerships with security intelligence vendors and law enforcement agencies so that security incidents are dealt with quickly and appropriately.
As connected cities come of age, they must rapidly scale and evolve to meet changing citizen requirements. Ongoing challenges for mature connected cities include integration, future-proofing, and assessing impact.
More than half of the world’s population live in urban areas. As urban areas expand, and more cities embrace IoT, it is likely that connected cities will expand to connected regions and beyond. Connected cities may eventually subsume data and services from the broader IoT scope, for example smart education and healthcare.
The technologies that are used to implement connected cities will need to be upgraded over time, at the very least to keep them up to date with the latest security patches and performance enhancements. Also, as the city grows, connected cities will want to take advantage of any new technologies and business processes that become available, for example, to use aerial drones for retail deliveries.
The way that citizens interact with IoT devices, and their expectations of behavior within the connected city, will inevitably change over time, as our lifestyles and culture evolve. Connected cities implementations will need to change to reflect those new requirements.
Throughout this process, legacy infrastructure and technologies will need to be maintained, modernized, and extended either through refactoring or reimplementation. Some degree of future-proofing will be necessary to make it possible to extend and maintain the solution over time. For example, cities must pay attention to industry best practices, such as adopting open standards and adopting a microservices approach to designing the architecture for the connected city platform, and using loosely coupled services and technology-agnostic, abstract APIs. Then, services can be upgraded independently and incrementally. And, IoT solutions must be designed for flexibility and scalability. However, IoT moves at a very fast pace, so it is best to be driven by citizen requirements, rather than trying to anticipate requirements too far into the future.
Connected cities initiatives must also be prepared to use historical data that has been collected from smart sensors to demonstrate how well they are progressing towards solving the issues that they set out to address through the application of IoT technologies. This involves quantifying cost and time savings over time and describing and assessing KPIs relating to improving sustainability, reducing traffic congestion, improving emergency response times, increasing citizen engagement, and so on. Demonstrating the effectiveness of the solution is usually a precondition of securing ongoing investment and buy-in from the community. IoT Platforms provide analytics services and visualization tools that can assist with this process.
Implementing connected cities is a long-game process. The benefits of connected cities will not likely be immediate and are more likely to be incremental to begin with. However, in the longer term, the efficiencies and cost savings that are achieved through the application of IoT to urban scenarios enable cities to scale their municipal infrastructure and grow sustainably while offering significant economic benefits.
Learn how to use a NodeMCU board, a sensor, and MQTT in Watson IoT Platform to build a practical IoT…
After presenting a broad overview of the concepts and skills that IoT developers need, this learning path guides you through…
To be a part of the IoT technological revolution, you’ll need to explore the architectures, designs, and development best practices…
Back to top