Nexus2030: Laying the Groundwork for Smart Territorial Innovation
The Nexus2030 project officially commenced on May 30, 2022, with the strategic objective of creating an advanced territorial Digital Twin capable of integrating scientific evidence, urban data, and civic participation to support intelligent and sustainable public policies.
In its first 10 months, activities have primarily focused on methodological setup and the definition of the conceptual and technological architectures that will guide the entire project. Work Package 1 (WP1) established the theoretical and practical foundations for the evidence-informed decision-making process enabled by the Digital Twin, a crucial process for making decisions based on concrete evidence. This process included the adoption of international best practices and an initial classification of different types of evidence, ranging from scientific evidence and best practices (usually transmitted informally among domain experts) to all information produced by citizens, thus citizen-based. This classification made it possible to model the knowledge space, a tool designed to manage all information, configuring itself as a space where the user, regardless of the type of evidence, can make informed choices by comparing the available data. This is not simply a knowledge base, but an advanced tool integrated with artificial intelligence models capable of answering questions posed by users in natural language. Over the course of the project, this tool will be implemented and tested with real data to evaluate its effectiveness. In parallel, a study on ontologies and taxonomies was conducted to identify a smart and lightweight solution for representing the urban environment. The adopted solution is to transform a complete ontology into a leaner structure, such as a hierarchical taxonomy, while maintaining ‘parent-child’ relationships. This approach allows the tags obtained to be used within the designed knowledge management tool, offering a simpler method for modeling and retrieving it.
Finally, to evaluate the effectiveness of this tool, studies have been initiated to monitor the results obtained through the application of this methodology.
In parallel, with WP2, work has been done on defining a Citizen Science framework that integrates with the concept of evidence described above and that, in general, involves citizens in the collection of urban data. This involvement occurs in various ways, starting with the design of collaboration tools, such as a simple application that allows users to answer questionnaires, up to developing dedicated mobile applications that allow citizens to express their opinions on the city or to have a direct view of the urban context. In this way, citizens become actual data producers, similar to sensors.
The data produced will then be useful for potential analyses and predictions. Alongside these environments, co-design spaces allow users to create knowledge. Through these tools, the most experienced can collaborate to generate new information based on debates, events, evidence, and so on. All of this must be accompanied by participatory motivation mechanisms to ensure continuous and fruitful collaboration between society and science.
With WP3, advanced research activities on Artificial Intelligence technologies have begun, with a particular focus on AutoML and Explainable AI (XAI), to develop predictive models that are also accessible to non-expert users and, at the same time, ensure transparency and traceability in automated decision-making processes. This process will lead, in the near future, to the creation of a system capable of providing explanations about the predicted results, along with a kind of guide that allows even users without particular experience to understand the results. Several models and algorithms have been identified and tested through targeted experiments to analyze the gaps and strengths of each. The main objective is to make the system multi-domain, so that a user can use it for a specific purpose, but also for other applications. In addition to this, it is also necessary to think about how more experienced users can exploit the system in an advanced way to get the most out of the potential offered by these technologies.
In this phase, the following have also been initiated:
- the analysis of simulation models with System Dynamics and physical models (WP4)
- the study for the design of the Digital Twin architecture (WP5)
- the mapping of interactive and multidimensional data visualization technologies (WP6).