The concept of Citizen-Centric Digital Twins (CCDT) is gaining traction among academics due to its predictive, simulative, and visualizing capabilities, fostering public engagement. CCDT has the potential to address various city-level issues through citizen participation, such as disaster reporting, feedback for planning, and environmental concerns. However, there is a lack of focus on cutting-edge technologies for CCDT development.
Study: Empowering Cities: Citizen-Centric Digital Twins for Enhanced Governance. Image credit: Blue Planet Studio/Shutterstock
In a recent paper published in the journal Telematics and Informatics, researchers examined existing technologies enabling CCDT development.
Background
Recent reports reveal limited success in leveraging community involvement in the public sector and commercial initiatives. Utilizing digital platforms for user engagement in the public sector offers potential benefits for democracy, regulations, and city governance. Despite infrastructure governance's importance, it is underrepresented in literature.
The current study reviews infrastructure governance, focusing on automated data collection to enhance citizen-supported governance. As citizens frequently interact with urban infrastructure, they play a vital role in its operational efficiency. City digital twins allow real-time monitoring and citizen engagement. While this technology improves openness and trust, not all digital twins emphasize citizen engagement.
Data acquisition and model development for urban monitoring
The researchers conducted a systematic literature review using iterative keyword searches to select articles. Gathering 635 papers from 2011-2022, 210 were chosen after the title and abstract screening for the CCDT study. The critical appraisal skills program assessment led to 75 selected articles, supplemented by two papers and ten web pages. The articles are divided into citizen engagement with dynamic city digital twins and enhancing capabilities. Findings focus on CCDT's data acquisition, machine learning, and application programming interfaces (APIs) used for infrastructure governance. Content analysis addresses research questions, mapping, classification, and argument development to extract insights.
The literature content analysis revealed diverse definitions and concepts associated with digital twins. The content was categorized into three main groups: a) Representation - describing the interface’s characteristics (e.g., virtual, mirroring the urban system); b) Purpose - specifying expectations from the digital twin (e.g., urban development, decision-making support); and c) Output - highlighting capabilities such as simulation and visualization.
While many publications focus on representation and purpose, end-user engagement is underemphasized, potentially crucial for infrastructure projects. Definitions of CCDT are still evolving due to its novelty, but it involves combining collective intelligence into a digital twin. CCDT is a digital miniature of a city with physical-virtual connectivity, updated by citizens, enhancing city monitoring and governance.
Promoting citizen engagement
Various mechanisms were explored regarding data acquisition methods for CCDT to promote citizen engagement. Open-source platforms, remote sensors, crowdsourcing, and IoT sensors were prominent. Social sensing and VGI play significant roles in citizen engagement. Challenges include data quality and interoperability, particularly for integrating crowd-sourced data. Crowd-sourcing could supplement sensor networks, integrating citizens' localized information into CCDT for improved decision-making.
Machine learning algorithms for CCDT encompass object detection and tracking, which is valuable for real-time urban monitoring. Change detection algorithms, especially with mobile images, could enhance CCDT's assessment of urban developments. Integrating natural language processing (NLP) tools for categorizing text-based feedback and filtering spam could enrich citizen engagement.
In the realm of digital twins, artificial intelligence techniques analyze remote sensor-generated images and point clouds. Image and point cloud segmentation is crucial for CCDT development, extracting semantic data for environment building. This data includes land cover, vegetation, buildings, and even people. Though segmentation algorithms are used in city twin development, their application for citizen input processing is limited. Yet, a promising CCDT approach involves integrating convolutional neural network (CNN) algorithms known for image comprehension. This enhancement can bolster infrastructure insights and foster citizen engagement.
Application programming interfaces (APIs) such as Cesium, WebGL, and WorldWind facilitate 3D virtual applications for CCDT. WebGL, Cesium, and WorldWind were compared for their specifications, features, and contributions to CCDT. These APIs enable citizen participation through commenting, requesting changes, and assessing environmental impacts.
Integration of infrastructure digital twins within CCDT remains an area for development, with the current platforms offering limited scalability compared to manufacturing digital twins. CCDT's potential lies in merging infrastructure with human dynamics, spatiotemporal knowledge flow, and the continuous updating abilities of citizens.
Conclusion
In summary, researchers’ contributions encompass data acquisition, processing methods, and interface development for effective CCDT implementation. Key findings include diverse data sources, vital algorithms, and APIs like WebGL, Cesium, and WorldWind. The research offers a roadmap for CCDT technologies, enhancing urban management with citizen engagement.
This study lays the groundwork for researchers, outlining current developments and prospects for CCDT-enabling technologies. It proposed future research directions for CCDT. Future directions include overcoming data interoperability challenges through semantic web advancements; enhancing crowd-sourced data quality with noise correction methods; advancing geometric models using artificial intelligence such as computer vision; leveraging open-source libraries for 3D geospatial visualization; harnessing NLP to enhance CCDT capabilities; and utilizing participatory sensing data for change detection while addressing mobile device limitations.
Enhancing UAV Safety with Machine Learning