The workshop was held on Tuesday, May 30 2023 in cooperation with our sister project DigiBUILD.
Here below we report the programme and the minutes of the event.
Session 3.2- Towards the digital built environment: high-quality measurements to deliver data-driven services- Part III
Chairs: Gian Marco Revel, Università Politecnica delle Marche, Italy; Serena Serroni, Università Politecnica delle Marche, Italy
Round Table- Partecipants:
DigiBUILD: Gian Marco Revel, Serena Serroni, Nicole Morresi, Vittoria Cipollone [UNIVPM]; Tancredi Testasecca, Lorenzo Cristofori [Università degli Studi di Palermo- ENG]; Elissaios Sarmas [NTUA]; Jose L. Hernandez [CARTIF].
MODERATE: Philipp Mascherbauer [TU Vienna]
Philipp provided an introduction to the MODERATE project, offering a general overview of data-driven services, the methodology involved, and the generation and utilization of synthetic data.
Points of Discussion:
- Lack of awareness of data by the data provider: There exists a gap between users and technical developers, with technical partners often struggling to interpret the provided data. As a result, the data analysis process becomes time-consuming and requires significant effort.
- Synthetic data generation: To generate synthetic data, it is necessary to collect real data as a basis for training the model. Synthetic data can be generated to reflect daily, monthly, or seasonal patterns.
- Overcoming privacy concerns: Synthetic data can address privacy issues as it is impossible to trace the origin of real data from synthetic data sets.
- Application of synthetic data in Digital Twin: Synthetic data can be employed to create preliminary scenarios within the Digital Twin framework and compare them with the Digital Twin implemented using real-time data.
- Incorporating noise in synthetic data: Synthetic data should include noise since real data is inherently noisy. Algorithms cannot be fed with perfect data.
- Exploitation involving buyers and data providers: The process of data exploitation involves collaboration between both buyers and data providers.
Potential Collaboration between DigiBUILD and MODERATE and the Next Steps:
Considering that VEOLIA is a partner in both projects (DigiBUILD and MODERATE), we could:
- Using the VEOLIA Pilot as a use case, the collaboration aims to leverage real data from DigiBUILD to obtain synthetic data from MODERATE for District Heating Network (DHN) optimization and forecasting.
- It is proposed to schedule a meeting with the project coordinators before the summer break to discuss and plan further collaboration and project integration.