DTUP – Digital Twin Urban Pilot

The urban environment we live in is more and more complex. Interrelations and dependencies as well as the effects of changes are becoming increasingly difficult to assess. In this framework, the Digital Twin Urban Pilot (DTUP) concept is of key importance: It allows to create an image of the reality on which one can effectively test and simulate the effects of new solutions, plans or system changes in the digital image first, before they are then implemented – or showing an effect – in the real world. The DTUP goals are: i) to develop a system that allows to create, visualise and explore pilot 4D digital twins (DTs) generated from drone and street-view imagery; and ii) to showcase their high potential for integrated and advanced analyses once combined with different types of spatiotemporal data and by means of state-of-the-art machine- and deep-learning (ML/DL) techniques. DTUP builds on the experience acquired during the Artificial Intelligence for Smart Cities (AI4SC) project, which concluded in July 2020. Specifically, its main objective was the generation of a set of indicators at global scale to track the effects of widespread urbanization processes and, concurrently, a set of indicators to help addressing key challenges at local scale. In this latter framework, from constructing exchanges with the project users, it clearly emerged the need for more detailed 4D information which allows to characterize in high detail the morphology of the urban environment and, alongside, the possibility of integrating any spatiotemporal georeferenced dataset for advanced analyses. These requirements represented the basis of the DTUP activity.

The project study areas include the Frascati town center and the ESRIN establishment. The envisaged DTs consist in modern and responsive platforms which allow to explore detailed textured 3D models of ESRIN and Frascati generated from drone imagery, along with different EO-based products and ancillary datsets. Specifically, the DTs are organized in 3 different components, namely: a browser web application, an Android application for Smartphone, and an Android application for Wearable Devices. In this framework, use cases are being designed where the integration of satellite-based data, as well as mobility records and in-situ weather/air quality sensors allow to effectively support different thematic applications (including end-to-end decision-support ‘what-if’ scenarios). Here, two different approaches are considered: