Press & NewsSmart Cities and Artificial Intelligence | OMS IoT Systems

July 30, 2019

Artificial intelligence (AI) has not been a sci-fi term for some time now. Smart algorithms that basically make up AI now help make various industries more efficient; the public sector is one of these. Many metropolises around the world are now investing in becoming smart cities using AI for the public good.

A typical example is city traffic management. For drivers in a hurry, there is little more frustrating than the eternal wait for the light to turn green. For this, let us even forget the fact that ordinary networks of traffic lights can make an already bad road capacity in a city worse or directly cause it. The US Department of Transportation is well aware of this and is introducing computer-managed adaptive signal control. Relevant sources say this can save up to 10 – 50 % of drivers’ time. The AI used by the adaptive signal network can analyse in-formation about the current state of traffic throughout a city. The network can then dis-tribute green lights in a way that maximises road capacity.

The Solution to Complex Infrastructure Problems

The above example highlights one of the biggest advantages of AI – it can solve very complex tasks. For a human, it is almost unimaginable to optimise such a city network, as opposed
to modern computers with the capacity for it. Alibaba, the Chinese tech giant, has recently released a second version of their smart city solution – Alibaba Cloud City Brain 2.0. Firstly, it deals with the road capacity in the city, but it can also report traffic offences or, when an emergency arises, it can ensure a clear passageway for emergency vehicles such as fire en-gines. Here AI does not only make the quality of residents’ lives better, it helps save their lives.

Automation Saves Effort, Time and Prevents the Failure of the Human Factor

The next benefit of AI-powered solutions is the fact that ma-chines, as opposed to humans, don’t get tired when perform-ing routine and repetitive tasks, such as passenger ID controls at airports. In a single shift, a customs worker sees hundreds of people when checking their credentials, undoubtedly leading to a fall in their attention span and a rise in being prone to making a mistake in the administrative process. But why should these ID controls be left to humans? A system dealing with this was introduced back in 2014 by the Immigration Bureau of Japan at Tokyo’s Haneda Airport. Automatic gates verify the identity of passengers using fingerprint scanners. Besides making the passport control quicker, the system al-lows employees to focus on other, more important tasks.

Automatic Identification of Persons Results in a Safer City

One could understandably say that reducing the error rates in administrative processes isn’t such a major argument for using AI. But when we consider connecting AI to algorithms able to identify a person by their facial features (facial recognition), we enter a state of play where AI significantly contributes to safety in areas other than air transport. This is the promise of Face-First, an American company providing persons’ identifi-cation by facial recognition. Introducing the system in airport terminals allows for real-time control of passengers against databases of wanted persons. The significance of such a tech-nology in the fight against terrorism is self-explanatory. It is not only terminals, though. Thanks to a very common usage of camera systems, facial recognition can be deployed in the streets, which leads to unprecedented opportunities in Pre-dictive policing. The police can use AI to monitor the presence of suspects and use their behaviour patterns to predict the risk of committing crimes. Such a method has an exceptional potential in the fight against the violence of armed groups,
a problem present in many cities around the world.

How Not to Fear Big Brother

AI in smart cities incurs some risks. First and foremost, solutions using camera recordings and other data recorded in public, such as Predictive policing, carry an association with the term Big Brother from George Orwell’s novel ‘1984’. Many peo-ple are uncomfortable with the idea of being under such sur-veillance, while the objectivity of the solution gets questioned as well. An action resulting from a predictive algorithm can be perceived as beyond what the police are permitted to do. How-ever, my experience is that a deployment of AI is always about the cost-benefit ratio, and this applies to smart cities too. It is important to consider the possible benefits of AI and deploy it only in areas, where these benefits significantly outweigh possible risks. Full-scale monitoring of persons therefore isn’t desirable in the case of Predictive policing. A much more efficient approach to this is deployment in pre-chosen areas, where the risk of a crime happening is large enough to justify the encroachment on residents’ privacy.

 

Artificial Intelligence Only Where it is Needed

The first step in the effort to use AI for the public good should be an analysis of issues and priorities in city development; however, that’s usually not the case, especially in smaller cities. When these decide to invest in a smart city initiative, they usually start wasting funds on many random projects. For example, who benefits from the electric traffic counters on a cycle route recently installed by the municipality of Prague in several locations? The one and a half million CZK (Czech crowns) that one such counter costs could have been invested much more efficiently. Similar unreasonable solutions unfortu-nately influence the people’s image of what a smart city really is. A smart city should provide answers to residents’ most troubling questions. Those often concern transport or a lack of parking spaces. Having a simple algorithm find the data that can be used to considerably relieve the parking system of the city is no longer a problem. Opportunities brought to a smart city by the use of AI are often misunderstood. Smart solutions do not have to be all-embracing, perfect and extremely expen-sive. If the issue they have been deployed to solve is pressing enough, any small improvement using a solution that’s not completely perfect still brings tremendous added value.

Predictive Maintenance Can Save Lives

Tremendous added value also exists in predictive mainte-nance. Had Genoa only invested in predictive maintenance of its infrastructure, it could very well have avoided the disas-trous collapse of a motorway bridge. And we don’t have to go far. Just a year before that a walkway over Vltava collapsed in Troja, a part of Prague. It was only after this event – that miraculously occurred without any loss of life – that old bridge inspections began taking place in Prague. However, shouldn’t municipalities be acting proactively, especially when it comes to safety of the public? Having simple vibration sensors in-stalled on the bridge would allow AI to use predictive models and warn officials that the construction is undergoing signif-icant structural changes. Predictive maintenance also has its value in gas pipes. These may be technically imperfect and cause fatal results. Other mains, such as water, can also have their maintenance costs lowered with its help. Generally it is cheaper to avoid disasters than to deal with their aftermaths.

Data Technology Development Enables the Faster Rise of AI

Technologies generating data are becoming increasingly availa-ble to the public, which also proves that AI can find its applica-tion in many more areas than has been the case to date. These can be regular sensors even in places where a layman wouldn’t expect them. Artificial intelligence in connection with the Internet of Things (IoT) can nowadays optimise trivial tasks such as waste disposal. Why should waste collectors come every week, when instead they can come only when you need them and when it makes the most sense from a logistics standpoint? You, as individuals, may not find any use in this, but the introduction of such a system for the entire city brings substantial savings in transport – not only financial, but also environmental ones. The data does not have to come from conventional sensors. A city does not have to construct an expensive network of sensors to be able to analyse operation-al data. You only need access to data from mobile applications such as Waze, which works on the basis of crowdsourcing. Budgets required for projects that may actually have a positive effect on the quality of life in the city decrease considerably.

Artificial Intelligence for All Sizes of Cities

AI allows a truly wide range of opportunities in the context of smart cities. The recommended best practice is to carefully consider where the value of smart data solutions would be the greatest, and thus where they should be applied. This is the only way cities can fulfil the AI potential to significantly improve the quality of their residents’ lives, regardless of their size.

Author: Adam Votava, aLook Analytics 

http://iotsys.io/wp-content/uploads/2018/05/LOGOIOT.png

OMS, a.s., Dojč 419, 906 02 Dojč, Slovakia, Phone: +421 34 694 0811, Fax: +421 34 694 0888.

OMS, a.s., is registered in Companies register of the District Court Trnava, Section: s.r.o., Insert Nr 1694/T, Company ID No: 341 32 333, VAT No: SK 2020377711

Copyright by OMS IOT.SYSTEMS. All rights reserved.