Smart cities are rapidly evolving as a result of technological innovation and IoT proliferation into nearly every aspect of urban life. Ten years ago the number of sensors embedded throughout a city was small, and now they are almost everywhere—from roadways to thermostats, and even trash bins. The ability to connect, communicate and remotely manage different devices gave rise to a new trend—fog computing.
The term “fog computing” was created by Cisco and refers to as an extension of cloud computing to the edge of the network. Rather than sending all collected data to the cloud, fog computing implies that data will be processed locally on edge devices, that’s why it is also sometimes called “edge computing”.
It’s wrong to consider that fog computing is intended to replace cloud computing, since these two notions are complementary. Fog computing might be considered as a variety of hybrid cloud computing, since they both provide storage, applications, and data to end users. The key difference is that in fog computing data is processed and analyzed in fog nodes—decentralized devices such as industrial controllers, switches, routers, and video surveillance cameras. They are structurally located between the cloud and the data source, so to speak closer to the ground—hence “fog” is used in the name.
The global fog market has the potential to reach $18 billion worldwide by 2022. Recent studies show that organizations have 40% higher budget for fog computing this year than in 2017. The trend reflects that fog computing will continue to grow in usage and importance, proliferating in such spheres as energy and utilities, healthcare, transportation, manufacturing, and smart cities.
One of the challenges that smart cities face is a wealth of information generated, captured, and analyzed every day. As the number of IoT devices is increasing, the amount of collected data becomes overwhelming and requires heavy-duty computing resources to process it. It is also time-consuming and expensive to transfer the large volume of data between the cloud and data sources. Fog computing makes it possible to reduce the amount of data that needs to be sent to the cloud for processing, thus improving efficiency.
There are huge benefits of fog computing for smart cities:
The key goal of fog computing is to make big data smaller. It is estimated that the volume of data captured by connected devices will exceed 1.6 zettabytes by 2020. Fog computing is capable of reducing the amount of data by applying intelligent sensing and filtering, which allow transmitting only useful information based on knowledge available locally at a given fog device. Data analytics allows minimizing data volumes even more by throwing away raw data.
Fog nodes are able to process data onboard without sending it to remote cloud servers and delivering the results back. This allows to save data travel time considerably and to receive responses in real time. Immediate data processing might be critical for smart city systems, especially when actions need to be taken quickly: for example, turning on traffic lights to green when emergency vehicles with lights and sirens are approaching might save lives.
Transmitting and processing data requires massive bandwidth, which is limited in case of cloud computing. However, this is not an issue with fog computing since all data is distributed between local devices and not sent wirelessly, which allows for a significant decrease in network bandwidth consumption.
Data security is another big reason smart cities are turning to fog computing. It keeps sensitive and confidential data away from vulnerable pubic networks, thus preventing cybercriminals from gaining access to it. With fog computing, malware and infected files can be found at an early stage at the device level before they infect the whole network.
Smart cities create a good environment for the implementation of fog computing, since thousands or millions of things across a connected city are generating heterogeneous data on everything from road traffic, public safety implementations, waste management, and air quality. With the aid of fog computing, the data can be quickly processed and analyzed in order to run systems more effectively.
Smart cities use different types of sensors to monitor and regulate road traffic. Sensors embedded into smart traffic lights detect passing-by pedestrians, cyclists, and drivers; measure speed and the distance between them; analyze the traffic data as they collect it; and take real-time decisions to alter the lights or reroute part of the traffic if it is needed. This contributes to the improved flow of vehicles and fewer road accidents and casualties. The collected data might be later sent to the cloud for long-term analysis.
Fog computing combined with smart traffic lights has already proved to be very effective in dealing with congestions. As an example, a community in Bellevue, Washington installed intelligent traffic lights that respond to traffic conditions in real time: the green lights stay on longer during peak traffic periods. City officials estimated that this led to 36% decreased travel times on the city’s main road and saved from $9 to $12 million to drivers annually.
Palo Alto, California is another city that launched a smart traffic signal project enabling the integration of traffic lights with connected vehicles. The project started in 2016 and is expected to improve traffic signal timing considerably.
Waste management often represents a challenge to smart cities, since this process requires a lot of time, money, and resources. Garbage collectors clean up containers on certain days of a month depending on the schedule, without taking into account how much waste is in it. Collecting waste from a nearly empty container is not efficient as it leads to unnecessary fuel consumption, whereas overfilled bins make the streets look dirty.
The application of smart sensors and fog computing will allow real-time monitoring of the garbage level and provide a way for more efficient waste management. Sensors installed on garbage bins might identify that the fill level is almost reached and alert garbage collectors about this. The fill-level data might be sent to cloud for long-range analysis in order to optimize garbage truck routes and schedule.
By implementing fog computing, relevant environmental parameters and city’s natural resources can be monitored and analyzed. For example, the smart water system is expected to analyze water quality and detect any deviation from normalcy, such as high nitrate or iron levels. Furthermore, it enables water leakage detection and immediate informing of maintenance teams about the need to plug those leaks.
Greenhouse gases control is another area where connected technology and fog computing might be applied to improve environmental sustainability. The analysis of actionable data enables a city government to see the overall picture of greenhouse gases emissions and take reduction measures in time. Based on monitoring results, they might send the citizens reminders of the need to use less heat or hot water in order to help reduce greenhouse gas emissions.
Video surveillance systems equipped with smart sensors contribute a lot to security and safety on city streets. These systems are generating a great deal of information that needs be analyzed in real time in order to ensure effective public place monitoring. Traditional cloud-based models are not suitable for these purposes due to the massive amount of data, associated latency challenges, network availability, and the enormous expenses required to continually stream the data to the cloud and back. This is where fog computing comes in.
With fog computing, data collected from video surveillance cameras is stored and processed on fog nodes close to the edge. Low latency allows for effective surveillance and violence detection in public places, such as a busy airport or a city mall. Once an incident happens, security services will get an alert allowing then to act quickly or track an escaped criminal.
Fog computing has great potential to become the next big trend in smart cities due to its possibility to swiftly and securely process small volumes of data. It helps collect data on city activities from traffic to utilities, ensuring everything is running efficiently and bringing sustainability to urban life. However, the fog-only approach is not able to deal with huge amount of data, so the cloud will continue to have a pertinent role in the IoT ecosystem.