Click to learn more about author Bernard Brode.
There is a growing consensus that the “smart city” is the only real solution to the challenges faced by urban planners and city governors. With more than half of the world’s population now living in cities, the focus on harnessing the exponentially larger amount of data generated to create infrastructures able to handle the demand becomes more critical.
Different cities have responded to this need in different ways. Some have attempted to develop a centralized model, drawing on the kind of big data analysis that is quickly becoming the standard in private industry. Such models, however, have raised significant concerns — not just over the transparency of data collection in large cities, but also from a technical perspective. In reality, the sheer number of applications that smart city tech can be applied to means that any centralized system is quickly going to become overwhelmed.
This is why many cities are now turning to an edge computing model. In this article, we’ll look at how edge computing is being used, the potential applications of it within smart cities, and some of the challenges that it faces.
Edge Computing in the Smart City
In many ways, edge computing is a “natural” development of the kind of Internet of Things (IoT) networks that are an increasingly common feature of modern cities. At first glance, building a smart city system might appear to involve significant investment and other resources on behalf of city governments, but, in reality, many of the components of our cities are already collecting huge amounts of data on themselves and also processing it themselves.
The most prominent example of this is self-driving cars and other IoT devices, which have been built with such sufficient processing power that they can perform significant levels of analysis themselves, without sending it to a centralized cloud. Driverless car platforms like those developed by EdgeConneX, for instance, makes use of this kind of edge data center. In addition, the rise of smart homes now means that there is already data available on the energy use, pollution levels, and noise levels affecting private homes, and this data is commonly already processed using edge data centers.
Seen in this way, the challenge for building smart cities lies not so much in connecting the city, because the average city is already “connected,” at least through consumer networks. Instead, the challenge for smart city engineers is to access this data and harness the power of existing edge computing systems.
The Applications of Edge Computing
In some cities, these challenges are already being overcome, and an impressive number of applications for edge computing within the smart city are already being realized.
In NYC, the project ListeningNYC aims to produce a crowd-sourced map of noise across New York to inform the city’s noise policies, and this capability relies on data already being processed in consumer-facing cloud infrastructures. In other cities, municipal governments are collecting edge-processed data from driverless cars in order to shape traffic policy.
In many cases, the approach taken by city governments has been relatively hands-off. Several cities have delegated responsibility for developing edge-enabled smart city systems to private firms but are also providing significant levels of seed funding to catalyze this process. The Copenhagen Solutions Lab, for instance, serves as the governing body for all smart city projects in the Danish capital, and Singapore’s Smart Nation initiative aims to engage citizens, industries, research institutions, and the government to “harness ICT, networks, and data to support better living, create more opportunities, and to support stronger communities.”
The Challenges
All this said, there remain major challenges for the application of edge computing within smart cities. These challenges can be broken into two major components: reliability and security.
As Eggplant CEO Dr. John Bates has discussed in SmartCitiesWorld, one of the current challenges of deploying edge computing to manage cities’ systems is that, in some cases, these systems need to be 100 percent reliable if they are not going to cause fatalities. If a traffic light system relies on edge computing, for instance, and the edge server fails, drivers could die. For that reason, writes Bates, “Smart infrastructure needs to be ‘always up,’” and this requires building significant levels of redundancy (at significant extra cost) into edge processing systems for smart cities.
Secondly, there is the issue of security. We’ve previously reported on the concerns raised by smart street lights recording data on citizens, and if such systems are to become more common, city governments will need to take an in-depth look at how they share data — not just when they share data with outside firms but also to develop new standards for email encryption within their networks so that the data they hold is safe from hackers, surveillance, and ultimately fines levied by data watchdog regulations like GDPR. At the moment, and as Kaspersky recently reported, edge computing systems lag far behind centralized cloud processing when it comes to cybersecurity.
The Bottom Line
The key to overcoming these challenges will be a mutual and thorough process of education. At the moment, city planners and tech developers find it difficult to communicate their needs and concerns, especially when it comes to emerging technology like edge computing. Moving toward a world in which the average city planner has some level of knowledge about the way cloud infrastructures work, and the average developer is conscious of privacy concerns, could eventually allow edge computing, and the many applications of this within the smart city, to become mainstream.