Smart cities: The 5-year outlook

I was recently asked by a client where I thought smart cities would be in five years’ time, by the end of 2021. Here is a summary of what I said, drawing on several of my earlier blog posts, although I have not referenced these separately.

I began with Gartner’s observation that smart cities are the fastest-growing area of the Internet of Things (IoT), and that by 2018 there will be 3.3 billion connected devices in cities. I then tried to identify the consequences that I thought would follow from this, both good and bad — things to welcome and things requiring attention.

Accentuate the positive

The explosion in numbers of connected devices will continue to offer opportunities to improve service delivery, reduce environmental impact and reduce costs, much as it has already: examples are legion, and there is no reason to suppose the phenomenon will stop.

I suspect, however, that certain sub-sectors within smart cities will get more attention. For example, as Gartner reflects, smart commercial buildings are a slam-dunk for IoT technologies due to the cost and energy saving potential, and there are of course many, many buildings that remain to be retrofitted.

But beyond that, it is evident that body-worn cameras on police and first responders are an idea whose time has come, to protect them and the public; by 2021 the water industry globally surely will have to have gotten with the program (won’t it….?); it seems more likely than not that autonomous cars will be a frequent reality by then, even if not ubiquitous; and as sea levels rise and climate events become more severe, more cities will be thinking about the use of smart-city technologies for resilience, disaster mitigation and response. 

Another growth area that increasingly will come within the orbit of smarter cities is the use of the IoT and accumulated data to provide both economic and community development platforms.

IoT-derived open data is enabling community engagement around things such as air and water pollution.

The open data movement is one of the drivers of this — that data is creating businesses (Zillow and Helios are examples); and IoT-derived open data is enabling community engagement around things such as air and water pollution. Elsewhere, some cities are starting to look at their cloud computing infrastructures as ways to provide computing services that support start-up businesses, or support NGOs.

This trend, part of a movement that is sometimes labeled Government as a Platform, or GaaP, seems likely to become more established, and given that it overlaps smarter cities where IoT data is involved, it seems likely that by 2021 the two will merge.

U-shaped information flows

By 2021, more cities will have come to appreciate the need for a "U-shaped" set of information flows. This is where the traditional focus on "top-down," government-to-citizen data and functionality is complemented by enablement for "side-to-side" (citizen-to-citizen) interactions and information sharing, as well as "bottom-up" data movement from citizens to government. These three dimensions make up a “U” when drawn on a piece of paper. 

As of today, too much smart-city activity is focused on the top-down only, with the result that the idea is criticized for being technocratic, alienating and undemocratic. That is a shame, as smart cities absolutely do not need to be that way, and the way to avoid it is using IT (frequently, mobile-based "systems of engagement") to enable the other dimensions of the "U."

The good news is that in my anecdotal experience more and more cities seem to be understanding this.

It is easy to predict that smart cities by 2021 will be starting to see the first fruits of how AI can support optimization and automation of many city systems.

One of the truly fascinating things about the IoT is that it has come into existence at the same time that various forms of artificial intelligence are, at last, commercially viable. It is, therefore, easy to predict that smart cities by 2021 will be starting to see the first fruits of how AI can support optimization and automation of many city systems, from autonomous cars to energy and water demand-response. It won’t be plain sailing (see below), but the first benefits will be verifiable.

Finally, on the positive side of the equation, we will start to see the availability and uptake of smarter-city standards, whether governing data and functionality, as in the City Protocol; or specifying performance measures, as in the forthcoming ISO standards 37151, for measuring the effectiveness of community infrastructures, and ISO 37120, for sustainable and resilient development.

Eliminate the negative

One of the issues with smart cities today is that in many cases they are politically “unnatural":  Where some IoT technology integrates hitherto separate systems or data, it requires departments to work together in entirely new ways for the available benefits to be realized.

More devices on the IoT, all other things being equal, means more integration opportunities and more potential benefits — which means in turn that pressures on city governments to break down functional barriers, or at least to find new ways to work across them, will become much harder to resist. Expect to see new forms of organization becoming more common by the end of 2021 to enable governments to leverage the new technologies more effectively.

Privacy will become a huge issue for smart cities by 2021. I predict that utilities, in particular, may feel the brunt of this. This is because smart energy and water meters are, besides being usage monitoring and billing devices, a huge treasure trove of information about appliances running in a household, and therefore whether someone is at home, their lifestyle, whether there are kids in the house (and if so of what gender and age) and so on.

Utilities will not be alone, however. Cities that collect traffic tolling data, for example, also may be in the spotlight. I predict that by 2021, cities and national governments will be implementing "charters" for how smart-city data will be used and not used, in some cases backed by a legal framework. 

Spectacular cascading failures

Because smart cities often include the integration of once-separate systems and data, they enhance the possibilities for more spectacular cascading failures — "failure chains," as I call them — spanning those systems. The 2003 energy system failure in the Northeast United States and Southeast Canada is perhaps the best example to date, where a tree branch hitting a powerline caused the line to trip. This, in turn, propagated across multiple segments of the grid until 60 million people lost power, and systems as disparate as water treatment, phones and Amtrak were impacted.

The key point here is that, apart from a bug in one alarm system that blinded system operators to what was happening for an hour or so, each grid segment was operating as designed: No single authority knew that the set of linkages in question even existed, or that the fault could propagate in the way that it did.

I predict that by 2021, several trends will take hold: further "system of system" failures will have happened. There will be discussion of where the sensible limits of integration are, given the current state of technology. And system design methodologies will be under development to address "system of system" interactions.

Finally, every one of Gartner’s 3.3 billion devices, in the absence of action taken to prevent it, will be a potential cybersecurity weakness. Given the systemic inter-linkages I have just discussed, cities will be only as cyber-secure as their weakest link. In my observation, cities today are not even close to being as aware as they need to be about this.

Sadly, by 2021, I predict that important pieces of infrastructure will have been attacked and broken in more than one city, and that lives may have been lost and economies will have been damaged as a result.

The silver lining in this rather gloomy picture is that awareness of cybersecurity issues will have increased and we may see the emergence of legally mandated standards of prevention and liability, as well as the widespread adoption of cybersecurity insurance.