Using satellite data and ML to monitor land surface processes
About the Pilot
Scientific studies have proven the feasibility of monitoring urban structures by satellite data from various sensors. This can help to satisfy their customers’ need to get ready-to-use information about urban dynamics down to the individual property-level, without the hassle of having the experience and the knowledge in processing satellite images.
Accurate information about the current extent of built-up areas and their change over time is not only crucial for sustainable urban development but also relevant for commercial applications in different industries (e.g., real estate). In that context, the start-up ubicube GmbH developed a promising pilot base service: the high-resolution built-up detection with continuous/regular monitoring.
With UDOS, ubicube tends to develop this system further and certify its transferability and its extensibility to larger geographic areas.
By generating the software (YoDS) to import the heterogeneous types of facility-building system datasets in a structured scheme, it is possible to provide interested end-users with all available information collected with the purpose of improving the facility management, the energy management and, in general, the restoration of the Italian public heritage.
The aim of YDMS is to combine data collected in the past on different public buildings and make them searchable according to different criteria: for instance, by ownership, use, location and other peculiarities. Furthermore, the availability of time-series data on the same assets permits to analyse potential changes in terms of investments, technologies adopted and developments in the different sites.
The technological requirements of the Data Lake based platform have been identified. YDMS Data Management System has been implemented, as well as the Import Data Software YoDS. The system was tested and validated, with the aim to benchmark datasets of buildings to use it in the (Building) health system in their software Y0DS.
The overall aim is to produce an urban extent monitoring system with unprecedented temporal and spatial resolutions that can be used for various use cases from the national to the local scale, enabling insights down to the individual property level.
After a mapping exercise, revised service prototypes were created and presented to a set of potential clients to gather their initial feedback.
Implementation sprints through C-SCALE services improved and optimised the application, including design revisions.
The full scale development cycle has been completed, with a final round of feedback ensuring the processing pipeline for the full-scale service is optimised, and the service has been launched. A SAR Coherence processing pipeline is implemented. The overall processing time has been significantly reduced.
WHAT SERVICES WERE USED?
In the pilot the C-SCALE Compute and Data Federation services were used to improve, scale, and validate ubicubes’ satellite data enabled urban dynamics monitoring service.