Smart cities make life better
In the summer of 2010, the 41st World Expo with the theme of “Better City, Better Life” was held in Shanghai, becoming a world-wide event for cultural exchanges, social development, and especially urban development issues. According to the statistics of the United Nations and the World Bank, the urban population accounted for 51% of the world’s total population in 2010 , and the proportion of the urban population just exceeded the key point of the rural population. Therefore, making life better through the development of cities has become a worldwide trend.
Now, ten years later, the proportion of the world’s urban population has risen from 51% to 55% . According to UN forecasts, by 2050, this proportion will further increase to 68% . It can be said that urban life is gradually becoming the norm of human civilization. With the development of cities, the concentration of urban population has brought more convenient life, but at the same time it has also brought about various new problems such as housing difficulties, travel congestion, environmental damage, and waste of resources. People hope to use emerging technologies to solve these new problems in cities, so the concept of “smart city” came into being. In the smart city concept, new technologies such as the Internet, modern industry, and artificial intelligence can be used to integrate urban systems and services, improve the efficiency of resource utilization, and optimize urban management and services, so as to solve urban problems and improve the quality of life of citizens.
The concept of smart city has been put forward by IBM in 2008. After more than ten years of development, some preliminary “smart applications” have slowly entered the lives of urban residents. The map software represented by Google Map combines geographic data with real city scenes, and uses algorithms to understand the city and plan routes without leaving home. On this basis, Uber in the United States and Didi in China integrate car data and user data, and use recommendation algorithms to help users get on taxis faster. In the security field, China built the surveillance camera system “Skynet” in 2017, and the number of cameras put into use in 2019 has exceeded 200 million. The same surveillance network is also being rapidly deployed in other parts of the world. For example, the Domain Awareness System (Domain Awareness System) jointly built by the New York City Police Department and Microsoft includes a large number of cameras and sensors, combined with the background data processing system, which can be used for surveillance. And quickly crack down on illegal crimes.
In the above smart city case, some artificial intelligence algorithms (recommendation algorithms, recognition algorithms, prediction algorithms, etc.) have been used, but most other applications are also focused on data collection, networking, and sharing (such as e-government platforms, remote electrical appliances, control, sensor array, etc.). With the development of artificial intelligence technology, these data can be better used to complete reasoning, prediction and decision-making, and further promote the development of smart cities.
Application scenarios of smart cities
With the rise of artificial intelligence technology in 2012, many new technologies based on deep learning can assist urban residents in their daily needs such as housing travel, help achieve sustainable environmental resources, and help urban managers to grasp information and communicate with residents more quickly. As a result, his life will be more convenient and faster.
The biggest helper of the smart transportation system comes from the autonomous driving system. When self-driving cars become the main force of urban traffic, not only can traffic safety be guaranteed, but also combined with big data and route planning algorithms, self-driving cars can automatically avoid congested areas and choose the best route.
Based on this vision, the research and industrialization of autonomous vehicles are in full swing. In October 2020, Waymo, a subsidiary of Alphabet, released a report on the safety performance of autonomous vehicles . The report pointed out that Waymo self-driving cars have completed a total of 15 billion miles of simulated driving and 20 million miles of unmanned highway driving. Among them, in the past two years of 65,000 miles of road field tests, only 18 real crashes and accidents have occurred. 29 simulated accidents, most of which were caused by the other party’s failure to comply with traffic rules. On the one hand, it shows that autonomous driving technology has matured and can deal with relatively simple situations, but on the other hand, it also shows that the vision of highly intelligent autonomous vehicles has not yet been realized.
Before self-driving cars can completely replace human drivers, some auxiliary driving technologies and road control technologies have entered daily life, such as automatic reversing and warehousing based on sensors, cameras and control technologies, pedestrian collision warning, front and rear collision warning, Lane change warning and so on. Through the comprehensive analysis of vehicle speed, vehicle distance, and images, the computer can intervene in the driving of the car a few seconds in advance, which can improve traffic safety. On the road, artificial intelligence algorithms can already be used to control traffic lights. In 2016, Hangzhou “Urban Data Brain” was tested on some road sections in Xiaoshan District. It used artificial intelligence algorithms to analyze vehicle data and road surveillance cameras to intelligently adjust traffic lights, increasing vehicle traffic speed by an average of 3% to 5%, and some road sections increased 11%.
Another important role of smart cities is to protect the urban environment and optimize the allocation of urban resources. This function can also be assisted by artificial intelligence.
The urban electricity system will benefit greatly. Urban power grids have different levels of power load in different seasons, time periods, weather, and regions. These data and knowledge in the power field can be analyzed through artificial intelligence algorithms to get the working mode of the power grid, and realize based on equipment status, network topology and real-time Grid health assessment based on operating data, monitoring and timely detection of abnormal power consumption. Equipment in the power grid, such as transmission grids and transformers, can also be monitored at a higher frequency. The field robot collects images of related equipment and analyzes the classification algorithm and integrated algorithm, and can timely discover the defects of the equipment in the power grid (such as the fall of shock-proof hammer, missing insulators, etc.) and risks (such as construction, trees, fireworks, etc.).
The urban environment can also be monitored by sensor networks. Take Barcelona, Spain as an example. The government has installed more than 20,000 wireless sensors throughout the city to collect data in various fields such as temperature, humidity, pollution, noise, and traffic flow. In the future, these data can be combined with artificial intelligence classification and regression analysis to predict pollution, weather, and traffic conditions, and also help city managers to take appropriate actions as soon as possible.
Waste sorting can also be supervised by artificial intelligence. It is estimated that by 2050, the amount of waste generated by urban residents worldwide will increase from the current average of 2 billion tons per year to an average of 3.4 billion tons per year . If it is simply landfilled, billions of square meters of land will be occupied every year, which will have a huge impact on the world’s environment. The intelligent waste classification system can replace manual labor and exert its advantages. In the Bin-e smart trash bin in Finland, the camera first collects garbage pictures, and then uses trained image recognition and object detection algorithms to analyze the images. Finally, the garbage is classified and compressed through a mechanical system. The internal sensors are also Users and garbage cleaning companies can be notified to deal with garbage in a timely manner.
In a smart city, the information interaction between city managers and citizens will be more efficient and transparent. The realization of this advantage requires not only the establishment of a data platform but also the participation of information technology represented by blockchain technology.
The characteristics of blockchain technology include distributed storage, multi-party maintenance, and non-tamperability. While satisfying the validity and authenticity of information, it also improves the efficiency of point-to-point transmission of information. Blockchain technology can help artificial intelligence algorithms to be applied, such as product traceability in smart logistics, smart security systems and public security systems form a private chain and transfer data, smart homes communicate securely through the Internet of Things, etc., are inseparable from the zone. Blockchain guarantees efficient data transfer.
Blockchain can also be used for data protection and sharing. In the city’s e-government system, city residents can see the release or change of policies in time, make timely feedback, and see the feedback of others. This will greatly facilitate the communication between city managers and residents.
Medical data contains all aspects of the patient’s privacy, and medical record information is generally only kept by the hospital, which is not convenient for direct use. With the help of blockchain technology, patients can establish a confidential electronic health record (Electronic Health Record), which can be completely and safely transmitted between the patient and the hospital. In addition, blockchain can also help the government and the public to respond in a timely manner when public health incidents occur. In the COVID-19 epidemic, the Chinese government has introduced a health code so that everyone can show their health status and understand the risks of nearby epidemics. Behind this, blockchain technology ensures data security and authenticity, and artificial intelligence algorithms are used to analyze risk levels. China’s effective control of the epidemic and timely response from the government have contributed to the health code, an information service system.
Smart medical care has always been a hot direction for the development of artificial intelligence. Microsoft has launched Healthcare Bot, a chat robot based on natural language processing and speech recognition technology. Patients can diagnose and triage simple diseases through dialogue with the chat robot online. In the imaging field, Chinese companies such as Yitu Technology and Shenrui Technology have developed intelligent diagnosis systems based on image classification and segmentation to help doctors quickly find lung nodules in tomographic (CT) and magnetic resonance (MRI) images. , Locate the location of cerebral hemorrhage, improve the diagnosis efficiency.
In the family, smart homes will gradually replace traditional furniture. With the popularity of the Internet of Things, whether it is traditional home appliances or curtains, doors and windows that originally need to be manually controlled, they can be connected to the “family data brain” for analysis and control, and smart voice assistants such as Alexa and Siri can recognize language commands, and Communicate to the corresponding household facilities. Artificial intelligence algorithms can also analyze various patterns of daily life, so as to directly control home appliances automatically. Now, smart devices have begun to enter the daily lives of ordinary people. Taking smart cameras as an example, smart camera products such as Nanit and Cubo AI in the United States integrate scene segmentation, behavior recognition, and face recognition algorithms, which can help parents monitor every move from babies to toddlers, analyze babies’ sleep conditions, and respond to possible occurrences such as young children climbing high, etc. or dangerous conditions that have occurred (such as covering the nose and mouth of infants, etc.) for early warning.
In the community, residents will enjoy the convenience brought by smart logistics and unmanned supermarkets. Amazon’s warehouse is considered to be the most efficient warehouse in the world, with more than 15,000 robots working in three-dimensional warehouses and distribution and sorting centers to quickly complete the handling and sorting of goods. In the field of unmanned supermarkets, after successfully operating the Amazon Go unmanned supermarket for two years, Amazon opened a larger unmanned supermarket Amazon Go Grocery in 2020, which not only increased the area of the supermarket, but also increased the variety and quantity of goods. This famous unmanned supermarket integrates computer vision, sensor technology and deep learning algorithms, which can simultaneously monitor the movement and interaction of multiple objects, so as to record each person’s shopping images and data in detail. Customers don’t need to scan codes and pay bills anymore, just simply put things in their bags, and they will receive accurate bills after they walk out of the supermarket.
Prospects and challenges of smart cities
It can be seen from the application scenarios of these smart cities that artificial intelligence technology has changed the relationship between people and information. City-related information data has become the training material of artificial intelligence technology, and the capabilities of artificial intelligence’s prediction, decision-making, judgment, and fitting can be widely used in smart cities to better serve people’s lives.
The changes brought about by the application of artificial intelligence in smart cities do not stop there, and the functional areas of the city may also change. Artificial intelligence, autonomous driving, and the Internet of Things technology have changed the way things are connected, and things and people are connected, so that the distribution of resources in and between cities no longer depends on labor, thereby reducing the transportation cost of goods to each community. Coupled with the rise of 5G technology and shared office spaces, people will increasingly move around their homes. The city naturally develops towards multiple centers, and each center becomes a comprehensive community, not just a residential area or a commercial area. This reduces overall travel expenses and naturally reduces carbon emissions.
At the same time, the occupational structure of the urban population may also change. As artificial intelligence can complete functions such as garbage sorting, traffic scheduling, car driving, and unmanned supermarkets, a large amount of simple and repetitive labor can be replaced, saving a lot of human resources. At the same time, large-scale data collection and continuous training of models are required behind these artificial intelligences, and positions such as data engineers, sensor hardware engineers, and artificial intelligence engineers will require more human input. As artificial intelligence participates in various fields such as medical care, education, information management, construction, and real estate, understanding artificial intelligence will be a necessary requirement for many occupations.
Of course, such a perfect smart city cannot be accomplished overnight, and it is even difficult to plan “top-down”. As the development of artificial intelligence technology is cyclical, city managers should formulate short-term development plans and long-term development plans accordingly. In the short term, city managers should support relevant professional enterprises, and use artificial intelligence technology based on deep learning to complete professional applications such as smart transportation, smart medical care, and smart grid, and then jointly form smart infrastructure from the bottom up. . In the long run, artificial intelligence technology may usher in revolutionary technological changes after a period of time, but information and data must be indispensable. Therefore, city managers should digitize city management and digitize city data. With the help of digitization, future cities can also realize “digital twins”, that is, the establishment of data mapping of physical cities, so as to carry out data simulation and prediction of urban planning and urban events. This can not only prepare a data foundation for further artificial intelligence technology, but also a more advanced urban planning and urban construction tool.
In addition to artificial intelligence technology, the construction of smart cities also needs to be combined with the development of other basic technologies. For example, the most far-reaching 5G technology, which can reach nearly 20 times the transmission speed of 4G technology, and can accommodate more communication devices to transmit data at the same time. On the one hand, a large amount of data required by artificial intelligence algorithms can be transmitted to the cloud for processing and the results can be returned instantly, thereby realizing lightweight smart devices that do not contain computing modules. On the other hand, the infrastructure in the city can be connected to the intelligent network as much as possible to truly realize the interconnection of everything. At the same time, the newly installed smart devices can also further promote the digitalization process of the city, enabling the “digitization” and “intelligence” of the city to assist each other.
In addition, smart cities also have some limitations. There may be huge differences in the history, culture, planning, and management modes of different cities, and many experiences are difficult to copy directly. For example, China needs to take into account the extremely high population density and historical relics, while Australia needs to take into account the huge differences between coastal cities and inland cities. Intelligent algorithms are always affected by data, and their work processes and results are more or less inclined to come from data sources. This requires city managers and social workers to supervise algorithms and data collection, and try to ensure fairness to all groups in the entire society. At the same time, urban residents have to surrender their data privacy rights while enjoying the convenience of these algorithms. Therefore, these private data should also be strictly managed. In addition, the cities own environmental constraints determine the upper limit of its development is not unlimited. So, instead of developing and expanding cities, the country should pay more attention to the construction of remote areas and rural areas, so that the convenience of artificial intelligence can be provided in the densely populated areas.
In short, smart cities paint a picture of life that is convenient, fast, intelligent and efficient, and aspirational for urban residents, and these scenes almost all require the participation of artificial intelligence. The construction of a smart city cannot be accomplished overnight. In the process of embedding artificial intelligence technology into smart cities step by step, urban residents will slowly accept the baptism of new ideas and new lifestyles, and human society will then usher in a great change.