Article |Chen Long

This article is reprinted with permission from the Collective Intelligence Club

Introduction

Wang Peng, a veteran smart city researcher and senior consultant of Huawei's Smart Cities, was invited to speak on the topic of "From City Data to Smart Cities" at the Tencent Research Institute×Collective Intelligence Club's AI&Society Salon. "The author reviewed Wang Peng's views on urban data and smart cities. The author reviewed Wang Peng's discussion on urban data and its application, and combined with the latest research on human digitization by Long Ying's team at Tsinghua University, he proposed a discussion on the virtualization of cities and individuals. The video transcript of the talk is available in the applet and web link at the end of the article.

Digital Twins from Industry 4.0

The concept of the Digital Twin can be traced back to a presentation given by Prof. Michael Grieves to industry at the University of Michigan's PLM Center in 2002 (there is no written evidence of this, but it is still widely recognized). evidence, this is still widely recognized as the earliest source of the digital twin).

In 2014, Michael Grieves elaborated in his whitepaper "Digital Twin: Manufacturing Excellence through Virtual Factory Replication. Michael Grieves elaborated on this in his white paper "Digital Twin: Manufacturing Excellence through Virtual Factory Replication". He argues that it is possible to build a virtual entity and subsystems in virtual (information) space from the data of a physical device that can characterize that physical device, and that this connection is not unidirectional and static, but is linked throughout the product lifecycle.

After this, the concept of digital twins gradually expanded to include analog simulation, virtual assembly, and 3D printing. With the continuous development of the Internet of Things technology, artificial intelligence and virtual reality technology, more industrial products and industrial equipment have intelligent characteristics, and the digital twin is gradually extended to the complete product cycle stage, including manufacturing and services, and constantly enrich the form and concept of the digital twin.

The business community is at the forefront of the digital twin. The digital twin in Industry 4.0 is being given its own understanding by major software vendors, who are integrating it with their own businesses and working to create solutions that merge the real world with the virtual world.

General Electric (GE) and ANSYS with the concept of digital twins, the realization of the fusion of physical machinery and analytical technology to achieve the way, so that each engine, each turbine, each nuclear magnetic **** vibration have a digital "twin", and through the digital model in the virtual environment. The digital model can be used to simulate robot debugging, testing, and optimization of operating conditions, so that the optimal solution can be applied to the machine in the physical world, thus saving a lot of maintenance and debugging costs.

Siemens invokes the concept of the digital twin to describe the data modeling that occurs across all aspects of the product lifecycle. In layman's terms, the digital twin is a simulation of the actual operating space of a number of factories, from product design to production line design, to mechanical design and factory planning and scheduling on the equipment manufacturing side, to final manufacturing execution and product big data.

French software company Dassault Systèmes in the digital twin innovation collaboration and validation, not only pay attention to the digital representation of the product, but also try to realize the interaction between designers and customers through the 3D experience platform.

German software company SAP based on the Leonardo platform in the digital world to create a complete digital twin, in the product test phase of the collection of equipment operation, analysis, the actual performance of the product, and then with the demand for the design of the objectives of the comparison, the formation of product development of the closed-loop system.

In short, the digital twin under Industry 4.0 is more for the manufacturing industry to provide the mapping relationship between the product in the physical space and the virtual space, as well as in the physical world as well as in the digital virtual space to record, simulate, and predict the process of the object's full life cycle trajectory.

Physical world and digital copies

Digital Twin: Spawning Smart Cities 2.0

It is worth noting that the concept of digital twin is not only active in the manufacturing industry of Industry 4.0, but also appears more and more frequently in the field of smart cities. As ICT (information, communication, and technology) has become the main driving force for smart city development, mobile communication, internet, cloud computing, sensors, artificial intelligence, and quantum communication have all been widely used in smart cities. The technological development in various fields such as full-area sensing, digital simulation, and deep learning is also about to reach an inflection point, which has led to the emergence of digital twins for cities.

The development of digital twins in China's smart cities has a long way to go. The digital twin is highly dependent on the data and information collected by the sensors, and at the current state of the art, it is still difficult to obtain a full-area perception of urban data and historical multi-dimensional data at a refined scale. Insufficiently detailed data on the physical entity space will directly lead to the lack of its digital copy. At this stage, the digital twin is still a long way from the imagined sandbox system simulation and derivation, artificial intelligence decision-making and other functions.

The core value of the digital twin in the development and construction of smart cities is that it can comprehensively establish a real-time connection between the physical world and the digital world, and then record, analyze, and predict the changes in the full life cycle of the operating object. The digital twin in smart cities can be divided into four stages, namely

the status quo twin that maps the current state of the city accurately, comprehensively, and dynamically;

the learning twin that learns, analyzes, identifies, summarizes, and discovers the laws of city operation from historical data;

the simulation twin that simulates development scenarios in different environmental contexts with human supervision;

the simulation twin that simulates development scenarios in different environmental contexts under artificial supervision. and finally autonomous twinning through real-time data access and automated decision-making with artificial intelligence.

At the same time, we should also see the huge potential of digital twins in the evolving technologies of sensors, 5g and edge computing. The high-density deployment of sensors and high-precision sensing, combined with the real-time structured computing backhaul of 5g and edge computing, full-area sensing and real-time updating of the city's physical space will be the norm in the 5g era. A brick, a tile, a grass, a tree, a table, a chair, a person, a car, will be updated with different frequencies of location and state information, so as to realize a real "holographic" virtual city.

Urban data: the DNA of the digital twin

In the construction of a smart city, the core of the digital twin lies in the construction of a virtual (information) space that is comprehensively mapped to the physical space of the city. Different from the manufacturing product cycle management by the manufacturer to fully grasp the product informationization data, the city as a huge complex system, which contains the physical space and process, all the time in the generation of multi-dimensional massive big data, which is undoubtedly in the data collection, processing, computing, storage and management of the city's digital twin presented a challenge.

In recent years, the data-centric urban ecosystem has framed the top-level design of the smart city, forming a smart city model of "perception-application-*** enjoyment of information" centered on ****-enjoyment of information and collaborative realization of various industries. At the same time, under the impetus of emerging ICT technologies such as big data, artificial intelligence, cloud computing, and the Internet of Things (IoT), multidimensional massive urban data are gradually being mined and applied in different ways in the research and practice of smart cities.

The electronic and spatial visualization of traditional urban statistics is the first step in the development of urban big data. Based on the outlining of administrative boundaries on the GIS platform and matching them with traditional yearbook statistics, the electronic visualization of traditional data can be realized, and spatial visualization and analysis can be achieved by relying on the spatial analysis function of GIS.

Electronic and spatial visualization of traditional statistical data on Cityeye

The application of Internet big data marks the city's real entry into the era of big data, and Internet big data has also become the "darling" of urban research in recent years, and both the academia and the industry are actively exploring the potential of Internet big data for urban research and development. The Internet big data for urban research and development brought many possibilities.

The biggest advantage of Internet big data is that it breaks the top-down data collection barriers of traditional data, and provides multi-dimensional data at a fine scale in a bottom-up manner, such as point-of-interest (POI) data that records the spatial location and attributes of all geographic entities in the city; social media big data that reflects the topic of hotness and user profiles; and heat maps that show the spatial distribution of the population in real time, etc.

The biggest advantage of Internet big data is that it breaks the top-down data acquisition barrier of traditional data, and provides multi-dimensional data in a bottom-up manner.

These data can be used for a variety of purposes.

And with the advent of smart cities, the advancement and maturity of sensor technology provides another path to data acquisition for urban research.

Through the multi-module integration of sensors within the city, it is possible to realize the fine scale of the urban environment, human and vehicle behavior and other data real-time sensing and collection. For example, the City Grid urban grid data monitoring station can use a multi-module sensor network to monitor the flow of people and vehicles and the quality of the environment, such as wind speed, wind direction, light, temperature and humidity, and pm2.5, etc. City Grid is an Internet of Things (IoT) product for the fine-grained sensing of the urban space, and it is also a classic example of the application of sensor technology in the city's overall sensing, data collection, and even the realization of the city's future microenvironmental and human-vehicle behavior prediction. It is also a classic case of sensor technology application in urban-wide sensing, data collection, and even the realization of the future micro-environment and human-vehicle behavior prediction. Wang Peng's team has also used City Grid to monitor and collect data on Tsinghua University campus and Baita Temple community, and to conduct in-depth analysis of the urban environment and people's behavior.

City Grid urban grid data monitoring station

LBS data (location-based service data), through the collection of uninterrupted signaling data between the user and the base station by the operator, to obtain the relatively accurate real-time spatial location of mobile service users. It is the "ultimate" data for describing the size and spatial distribution of a city's population because of its large number of users and wide coverage.

We digitized ourselves!

Digital Self

While marveling at how digital twins are disrupting the manufacturing industry and the management and operation of cities, some scholars have begun to explore how to create digital twins of individual humans.

Zhang Zhaoxi, a research assistant at Tsinghua University's Long Ying team, recently published a journal paper titled "Application of wearable cameras in studying inpidual behaviors in built environment. "The paper proposes an innovative use of wearable cameras for data collection, analysis and simulation of inpidual behaviors and urban spatial perception.

The research team used a portable camera to record one picture every five minutes directly in front of the wearer, and analyzed the wearer's individual behavioral characteristics, time allocation, path transfer, and place events by means of manual recognition, computer vision analysis, and color recognition analysis. The research results show that the picture data collected by the wearable camera has rich individual behavior and spatio-temporal information, and can effectively describe the behavioral characteristics of individuals in space.

Digitized "life logs"

With the wide application and rapid development of big data in urban research, large-scale studies on built environment level morphological elements (e.g., remote sensing, street view, and POI data) and a variety of Internet data (e.g., microblogs, reviews, and cell phone signaling data) can be used for the purpose of utilizing big data to study large groups of people, and for the purposes of using big data to analyze the behavioral characteristics of individuals. populations have provided a large number of case references and gradually established a theoretical foundation for the use of big data to explain urban problems. However, these big data based on coarse-scale urban physical space, or large-scale groups, are still difficult to be applied to the deeper profiling and research interpretation of individuals.

Wearable cameras provide a new opportunity to collect individual data on a large scale, and digitize personal activity information through recorded image data to form an electronic archive of the "digital self", which compensates for the lack of continuity and richness in the collection of individual behavioral data in the existing research, and this is a shift from data-driven to behavioral informatization of the urban environment. This is also one of the changes from urban environment data to individual behavior informatization. At the same time, the informatization of individual behavior will also promote the innovation of research methods and the intervention of new technologies, from subjective "individual perception" to objective "quantitative research".

From the perspective of digital twin, based on the picture data recorded by the wearable camera, the behavioral characteristics of individuals in the physical space can be stripped out through the collation and analysis, and further digitized in space and time, thus constructing their digital twins in the virtual (information) space. At the same time, a large number of elements of the built environment in the physical space contained in the picture data can likewise be digitized and recorded in the virtual space, thus reflecting the interactions between the individual and the environment in both the physical space and the virtual space.

The rapid changes in technology have not only dramatically changed the way people live, but have also affected every aspect of how cities operate. Undeniably, the rapid development of new technologies has brought new opportunities for urban research and practice, and driven breakthroughs and innovations in urban planning techniques and tools. For example, the data-enhanced design proposed by Long Ying allows planners to make more comprehensive and precise analysis and planning design response to the city with the help of multi-dimensional urban big data.

At the same time, with the help of innovations in information and communication technologies, the refinement of technologies such as data storage, mining, cloud computing, and visualization has provided new perspectives on the study of cities. People's way of thinking has shifted from traditional mechanical thinking to big data thinking, and the cognitive approach has gradually overstepped towards the experience of combining reality and fiction. The concepts of urban digital twin and digital self will also have richer connotations under the technological innovation of the Fourth Industrial Revolution.

References

[1] Wang, P.: Envisioning the future city where everything can be operated | A long overview of smart cities

[2] Wang, P.: Urban data to smart cities

[3] Long, Y. (2019). (New) Urban Science: Studying New Cities with New Data, Methods and Technologies. Landscape Architecture Frontiers, 7(2), 8-21.

[4] Zhang, Z. X ..., & Long, Y. (2019). Application of Wearable Cameras in Studying Inpidual Behaviors in Built Environments. Landscape Architecture Frontiers, 7(2), 22-37.