Artificial Intelligence: A smarter way to build smart cities
Source – financialexpress.com
Artificial intelligence, or AI, a term that a few decades ago circled around in the realms of sci-fi and fantasy, is a reality today. The most evolved AI of our fantasy tales is often sentient machines scheming to take over the world. We are not there yet, and perhaps reality is far more benign than our imagination.
AI, as we know today, has the potential to provide solutions to many of our real-life problems. In India, which is the world’s fastest growing major economy and has the second largest population in the world, AI can be transformational.
With the publication of a discussion paper on AI by the NITI Aayog recently, the government has not just signalled its priority to this disruptive technology, but also thrown it open to the public for ideas and suggestions.
One of the problems requiring urgent attention, where AI can demonstrably make an impact, is urban infrastructure, and here we have taken the first key steps with Smart Cities Mission. Our scorching pace of urbanisation means India is no longer a nation of villages. Every minute an estimated 30 people are moving out from their village, lock, stock and barrel, to become city dwellers. According to some studies, around 40% of the Indian population will be living in cities by 2030.
The question is: Do we even have the cities to accommodate this gigantic population? Experts say we will need at least 500 new cities to house our urban citizens or risk turning the existing ones into unmanageable slums.
Already, our metropolises that were planned at a time when resources were abundant and urban living was a luxury not necessity, are bursting at the seams with their hopelessly inadequate infrastructure and lack of planning. And while we have been extremely reluctant urbanisers, the late start has given us some significant advantages, one of which is the use of technology to leapfrog stages of development.
Across the world, cities are using technologies to become smarter, in the sense they are using it to manage key functions horizontally—city services, transport, communications, water, smart grids, public safety, education and health—all through a digitally-managed central command room. The government has embarked on an ambitious plan to turn 99 of our urban clusters into smart cities with an expected investment of `2.04 lakh crore. The strategic components of these smart cities include city improvement, renewal and extension, in addition to a pan-city initiative in which smart solutions are applied covering large parts of the city. Many have already begun implementing projects involving smart command and control centre, smart roads, solar rooftops, intelligent transport systems, smart parks, etc.
While these ‘brownfield’ cities are being retrofitted, we have ‘greenfield’ cities being built up from scratch under the Industrial Corridor programme. Creating a brand new smart city requires a comprehensive, long-range plan to direct the future growth and development of the entire area. For these greenfield cities, some of which are bigger in size than the country of Singapore, we don’t just have physical master plan, but, for the first time, it is dovetailed with a digital master plan to manage, operate and control all citywide utilities through one command and control centre.
The challenges of developing brownfield cities are different from greenfield cities, but common to both is the understanding that technology, especially emerging ones like AI and Internet of Things (IoT), will be the cornerstone of realising India’s aspirations of building the ‘intelligent’ cities of tomorrow. The basic premise of AI is the development of intelligent machines that are capable of high-level cognitive processes like thinking, perceiving, learning, problem-solving and decision-making. You provide the system with lots of data and get actionable intelligence in return; for example, in the way Netflix or YouTube gives you suggestions based on your browsing patterns.
Cities of today are sitting on goldmine of useful data, churning petabytes of it daily through video cameras, sensors, traffic management systems, smart meters, vehicles and mobile phones. AI has the potential to make sense of the humongous data and use the intelligence to increase performance of cities, optimise operational costs and resources, and enable sound citizen engagement.
For a smart city operation, the basic ICT functions include collection of data, say using sensors, CCTVs, smart energy meters and even social media engines, for real-time human activity. This would be relayed through communication systems like fibre optics, 3G/LTE, internet, Bluetooth. The data would then need to be analysed using AI and other tools, and the intelligence generated used to direct decisions and actions.
How can AI be potentially used in smart cities? The answer is, in countless ways. For instance, for management of traffic that has become the bane of many cities today—AI can understand real-time traffic, calculate permutations in a flash, process data at local level of traffic junctions, and even turn traffic lights off and on intelligently.
Public safety and security is enhanced by AI through sophisticated surveillance technologies, accident pattern monitoring, linking crime databases, combating gang violence and so on. AI can help with crowd management, estimation of size, predicting behaviour, tracking objects and enabling rapid response to incidents. It can be invaluable for managing utilities and optimal use of resources such as distributed energy and water. AI can lead to smarter homes with resources-saving applications and easing domestic workloads. It can greatly ease citizen services delivery, processing of files and applications, through chatbots for responding to enquiries with smart conversations. These bots can free up operational staff at help centres to address more complicated and time-sensitive queries.
An inevitable outcome of digitisation is cyber-attack and cyber-crime, targeting sensitive and personal data, which again AI can help manage to some extent by detecting vulnerabilities and taking remedial measures automatically. AI can similarly assist in numerous other city functions to improve public accessibility, including maintenance of outdoor spaces, lighting, parking management, learning, education, skill development and so on.
Technology can be a great enabler in achieving our primary goals of a smart city—protecting citizens and planning for the future; building a sustainable, resilient infrastructure that optimises resource use and ensures essential needs of all are met; and enhancing the quality of life of citizens, enabling individuals to improve their health and productivity.
But while we do have the advantages of access to a rapidly evolving technology ecosystem, there are other challenges we need to overcome. One of these is coordination between stakeholders—residents, administrators and operators. With AI, data sharing and a well-defined activity chain will be essential for efficiency. Large migrant populations in cities may skew data, so we will need to adjust a number of variables. For example, apart from actual ingress of people, increasing business activities, infrastructure expansion easing connectivity, community exchange must also be analysed.
A more technical issue with data-reliant decision-making is the accuracy of the data itself. We need to bring in a robust standardisation system for accurate data inputs. For example, in detecting wearers of safety helmets among bike riders without proper standards for helmets, AI will count as positive even those that are not approved by law. Also, ICT infrastructure requires uninterrupted network, power and regular maintenance. Difficulties remain with relation to costs, getting people with domain expertise and end-to-end solutions for this level of work.
The Smart City model has never been implemented on this scale in any part of the world. Many systems have been adopted and implemented in silos, but all the functions of the city have never been integrated right from scratch. It is an exciting new challenge for India; we could be creating new models for future cities and setting benchmarks for urban infrastructure.
The government has steered this mission in the right direction; the budget allocation for Digital India—the government’s umbrella initiative to promote AI, machine learning, 3D printing and other technologies—was almost doubled to `3,073 crore ($477 million) in 2018. Roping in the private sector, especially Indian companies that have a solid technological expertise, to shape the overall AI strategy would be a good move.
A standardised ICT infrastructure for seamless integration—state and countrywide—is essential. Start-up companies may be given access to data so they can create business models around use cases, helping both government and start-ups. What is important this time is we make a start somewhere. In many ways, public engagement is also critical for this mission since our talent, skills, creativity and ideas are largely invested with our people.