Artificial Intelligence in Smart Cities

The ever changing technological environment right from ancient civilizations to the modern era we live in right now has impacted the way we live, work and spend our time. For the first time in our history, 50% of the world’s human population now live in cities and this is expected to move to 70% by 2050 [1]. Currently, cities are facing challenges in dealing with carbon emissions, energy consumption, traffic and dealing with an aging infrastructure [2]. Cities are now starting to integrate technology in their core functions, including municipal operations, safety, infrastructure and transport. This has driven the application of IoT devices for a more connected and an informed city. However, it means that a large number of data is continuously gathered from several sources and to meet this growth of data sizes, we need a system that can efficiently monitor and analyze this data to add value to city councils.

Managing Big Data & Smooth Operations

Algorithms used in AI are improving everyday and when systems are created to learn from past experiences, the applications are endless. AI and smart cities are a perfect fit as smart cities gather a significant number of data from its sensors and citizens while AI is only as good as the amount of data it gets. However, care needs to be taken when building the AI system so that it sits well with regulations, needs of authorities and citizens [3]. The data that comes in from different applications can be converted into actionable information to improve the city’s operational efficiency and cut down costs in the process. For instance, energy use data can be utilized to manage energy production and real time traffic data can help in visualization and decongesting prime areas in the city. This also sheds light on the population’s behavior which can be predicted well after several iterations. The city can then preempt its population’s needs and serve them better.

Putting Citizens First

Smart cities are developed around its citizens’ needs. It allows the personalization of experiences by giving meaningful insights, reliably and accurately. For example, a visit to the shopping mall or walking around streets can be personalized through holographic projectors or even by sending relevant information to smartphones. Public safety will be improved as AI systems can predict and take action during emergencies. For instance, once a traffic accident is detected the automated response management system takes over and communicates to the relevant authorities to take further action.  Furthermore, AI systems that are built taking into consideration citizens’ perspectives and participation can not only help in making a more cohesive environment but also assist citizens in making smarter choices [4].

Pattern Recognition

Perhaps, the most important aspect of AI in smart cities is its ability to learn about citizen-city interaction. It uncovers information that was unknown before and empowers decision makers to make the right choice when dealing with significant changes in their operations. The complex nature of cities can be challenging to narrow down the application of a clear model. AI can help in determining the right model for a particular city before complete deployment, thus reducing the financial risk of huge upfront capital costs.

In conclusion, the integration of artificial intelligence in smart cities only seems natural as AI is more effective when it has large datasets to work with it. It can uncover previously unknown information about a city’s population by recognizing patterns in data that was simply too complex to do until now. It brings in many benefits to cities like convenience to citizens, lowering energy costs and reducing environmental concerns.

References  

[1] – http://www.un.org/apps/news/story.asp?NewsID=48240#.WYTHxsKpCfU

[2] – http://www.mdpi.com/1996-1073/10/4/421/pdf

[3] – http://meetingoftheminds.org/smart-cities-artificial-intelligence-balancing-opportunity-risk-21706

[4] – http://smartcitiescouncil.com/article/how-will-ai-transform-cities-3-experts-weigh

1+

10 comments on “Artificial Intelligence in Smart Cities”

  1. Hi Rahul!

    Great article you got here. AI (combined with big data) indeed has the potential to make cities smarter which could manifest in better traffic regulation, energy grid optimization, and even disaster preparedness.

    However, gathering data from huge cities involves a daily observation, investigation, and analysis of the lives of its citizens. For example, let’s focus on the topic of energy. When studying the energy demand patterns of families, you would be gathering data on their electricity usage all day and for more effective analysis, even classify the specific appliances and devices that generated the different levels of electricity demand.

    This of course allows me to bring the topic of conversation into privacy. How would a family feel with the idea that their local utility knows when their lights are on/off in their masters bedroom, or when the water heater is on in their shower? Exiting the topic of energy, how about the idea the street cameras will allow city councils to know where your location is throughout the entire day?

    I do hope that cities can get smarter because it will help tackle global problems like climate change, overpopulation/congestion, waste management, etc. I am on that side with you. However, we have a lot of human ethical issues to climb over before we are able to fully utilize the capabilities of AI and provide it with the big data that it needs to get smarter together with our cities.

    1+
  2. Hi, Rahul! This is a nice article talking about how AI could make the city ‘smarter.’!
    I do agree with your point: In these days, the systems that could add together to form a smart city seem to be separated: Google maps to help us get traffic info and find a better route, store apps that could direct us to the store to what we need. And if they share the data and work together, the city will be smarter. Especially, I think that applications of AI in cities could bring many conveniences like more intelligent transportation system, a smart shop as you mentioned. I believe that this process not only requires some smart devices, but also needs the cooperation of the city government to provide necessary data like traffic condition, the police department to receive real time notification about situations happening in the city, from many other smart types of furniture at home, and from many existing apps. I believe that there could be one platform that could cooperate data from all these, but I also think that it might be a problem to make everyone and every store to perticipate because there will be people worrying about privacy. The idea will only be valid if all the departments in one city work together and come up with a privacy policy.
    (Su17-MS&E-238A-01)

    1+

    Users who have LIKED this comment:

    • avatar
  3. Thanks for the insightful post, Rahul! I agree with many of the points that you bring up. One area that cities must remain weary of as they continue to innovate and use technology in their day-to-day operation is the protection of privacy. I was recently shopping for ink cartridges on Amazon and a few hours later, while browsing Facebook, I saw an ad for the very same ink cartridges (which, needless to say, was a bit creepy). With cities using this technology, they have access to many more personal details about each citizen and thus, great care must be taken to protect each individual’s privacy. Strong cybersecurity systems are also a must to ensure sensitive data isn’t leaked.

    0
  4. Hi Rahul! Thank you for this very interesting post, you write very well. I think that one of the main technologies to come in the near future that will allow the evolution to a smart city as you describe it is 5G. The distribution of 5G will really constitute a disruption in the way that we will be connected : it will for an ubiquitous wireless network that will marry data collection and computation with billions of devices. I cannot wait to see what 5G will enable at any level in the future, especially for the development of smart cities!

    1+
  5. Thanks for sharing your opinions on this interesting topic. I like the vision you have for future technologies in city building. And you also pointed out the AI would be valuable only when it can get enough data to learn patterns. However, as we know, accessibility of data and privacy of it are two counter sides. I hope laws can catch the fast speed of technologies developed as soon as possible. So that we can have more people live in smart cities.

    0
  6. Great post! Do you see an opportunity for the integration of smart grids for electricity and potential reuse of waste heat/energy as part of the smart city strategy?

    0
  7. Class 238A

    Hey Rahul! very well written and you have managed to get relevant points to the table, but three problems to highlight on this topic:

    Sensors:
    Building smart cities need deployment of sensors without which collecting valid data is kind of impossible. Sensors will measure temperature, traffic patterns, foot traffic, air quality and infrastructure integrity, among many other things. Lux Research, an innovation research and advisory firm, has a report that suggests the world will deploy 1 trillion sensors by 2020. now this means that you would need 1 trillion batteries to power these sensors.

    Power:
    Batteries are not only difficult to manufacture but also will bring along other challenges of pollution and energy consumption (even if we have a direct power source connected on wires). this means we need to first develop an mechanism to transmit power over wireless channels to overcome the above challenge.

    Political Will:
    We assume that by tapping into the collective intelligence of both devices and people we can create better living environments. I believe we can. But data is not a magical cure to all our woes.
    Smart cities will need leaders who have the courage to defend data, say what it means and establish it as a truth upon which cities make decisions. If officials don’t stand behind their data, neither will the public.

    I trust with powerful data and political will, enabled by wireless power, we’ll come up with some great answers.

    0
  8. Hi Rahul!

    Thanks for you post.

    I’d like to add to the discussion some perspectives on the challenges that smart cities are facing.

    Smart cities are, for instance, inhibited by hurdles with budget constraints and limited resources and expertise being the most frequently cited inhibitors (Black & Veath, 2017). Related to open data is the issue of whether whether data should be made available free of charge or not and managing implications of open access to data if data is in fact to become a new “natural” resource. Another data related issue is deciding on the appropriate data velocity, a challenge which is particularly evident for municipalities where lack of expertise and resources is one of the most widely recognized hurdles. When making a city smarter, it is also of paramount importance to acknowledge that one system (e.g. the building system or the transportation system) only represent one system in the system of systems that makes up a city. Acknowledging that systems are interrelated is important because applying a reductionist view on cities leads to a siloed view which can inhibit positive gains. This is a widely accepted notion, for instance within IT architecture research (e.g. Weill & Ross, 2010).

    0
  9. Many cities are trying to embrace the ‘smart cities’ approach to running an efficient city. In my past article, I gave a few notable examples (http://mse238blog.stanford.edu/2017/07/popanga/city-2-0-building-smarter-cities/). The main areas where smart cities can be applied are efficient use of water and energy, seamless transportation, agriculture- smart irrigation, food supply, implementing and monitoring security, disaster preparedness, space usage – parking, highways and sewage management. IoT devices can be implanted on almost anything; traffic lights, buildings, roads, bridges and alleys, underground pipes, dams, and even on plants and animals. For some applications, the data collected must be analysed in real time by AI especially in urgent situations such as crime detection, accidents and other emergencies. Some applications could need less periodical updates such as energy consumption systems or irrigation systems. So much work is needed to make these systems efficient and profitable. Network security is the major challenge in such systems as they are susceptible to hacking.

    0
  10. Thank you Rahul for sharing your thoughts! 🙂

    I also think that AI and Smart Cities is the next logical step that has to be focused on.
    Smart cities build the intersection of the Internet of Things, analytics and AI. With public infrastructure, public services and city life, AI could be useful in all these sectors that define a city and its lifestyle. Furthermore cities are the fastest growing area of the IOT, with 3.3 billion devices due to be connected by 2018 and data becoming a new natural ressource, it is clear that this data needs to be analysed. AI will be that something, enabling multiple systems to be optimized together, detecting emergent patterns, and providing wholly new capabilities in ways that traditional analytics tools cannot.
    http://meetingoftheminds.org/smart-cities-artificial-intelligence-balancing-opportunity-risk-21706

    0

Comments are closed.