SERVICE DEVELOPMENT PROJECTS

MAFIS – Multiple Actors Forest Information Service

Description:

The MAFIS project will develop and demonstrate at pre-operational level a service aimed at providing consistent and statistically reliable information on forests, by combining multi-spectral satellite data and SAR data, to monitor natural and planted forests growth and forest losses due to human activities or to natural phenomena.

The project will develop a specific methodology and will use Artificial Intelligence to routinely produce, at affordable customer price, high accuracy information consistent at local and supranational level. The resulting MAFIS service will provide information useful to different end-users both in the public domain (European, national and local level) and in the industrial domain (forest owners/producers, forestry companies and wood-forestry value chain).

The project combines multi-spectral and radar data, develops Artificial Intelligence algorithms to speed-up data analysis and information extraction and will implement a software procedure suitable to be run on a cloud infrastructure. The MAFIS service will be accessible through a delivery platform already developed by GMATICS that will be adapted to enable effective visualization of the forest related information for different users.

The achievement of the project objectives will be demonstrated over two test areas in Italy through comparison of the results obtained from the MAFIS methodology and software procedure with in-situ survey data. Validation of the results include the participation of end-users.

The MAFIS project will be implemented by a team that includes GMATICS (Prime), GEO-K, RDM Progetti and University of Molise.

DYDAS – Dynamic Data Analytics Services

Description:

The aim of the DYDAS project is to develop a collaborative platform in order to offer data, algorithms, processing and analysis services as well as business transactions with high added value by exploiting HPC and Big Data technologies, including machine learning. The platform will match demand and offer among those who own intellectual properties on data/algorithms/services and those who need to exploit them.

The platform will offer services for a wide range of users (e.g. municipalities, industry, product developers, research centres, etc.) exploiting heterogeneous data, open and proprietary algorithms, service providers, supercomputing services. The software platform ensures a secure collaboration environment, through NDA between players, and IT security to protect user case development (confidentiality) and attract more data, algorithms and service providers.

The project envisages three use cases:

Maritime – to forecast detailed marine status (waves, winds, tides, etc.) to provide regular information for public and private entities as well as to citizens involved in maritime activities.

Energy – to assess energy demand and renewable energy potential production (PV and wind) for energy planning (electricity industry sector) and energy security (government) purposes.

Mobility –  to use data streaming (videos + other data) from car-sharing fleets and car black boxes for acquiring updated information about the current status of the urban environment and critical situations.

NetMoA – Network Monitoring and Alert (2)

Description:

The second phase of NetMoA development aims to realize the evolution of the prototype monitoring service developed under Copernicus Incubation. to better respond to the information needs of large infrastructure operators (such as electricity, transport, energy, telecommunications) that must monitor, protect and manage networks that unfold for several tens of thousands of linear kilometers and that are subject to interference from human activities vegetation and natural disasters.

This project envisages the purchase and the use of a workstation with high data processing capacity, equipped with a high-performance NVIDIA professional GPU, specifically designed for data processing through artificial intelligence and deep learning. The project implements a workflow with a more complex logic decision tree that combines algorithms for data processing with more stringent multi-temporal analyses and various AI techniques.

The results of the new processing chain will be extensively tested, also through in-situ surveys, in test areas selected with end-customers to validate the project outcome.

The project also includes the promotion of the service to potential customers abroad to start testing the market potential outside Italy.

NetMoA – Network Monitoring and Alert (1)

Description:

The NetMoA project is aimed at providing regularly updated information to network operators about critical situations (anthropic activities and growth of trees) occurring in proximity of their infrastructures, to improve the protection of their assets and to reduce the cost of current surveys.

NetMoA will use both Sentinel data and VHR data to detect and characterize various risk situation that can affect the integrity of the infrastructures and the continuity of the related service. NetMoA will exploit EO processing algorithms and various AI techniques to process and integrate different types of satellite data to provide regular updated analyses about the environment surrounding the various infrastructures. Possible service adaptations to include specific features for different types of infrastructures will be tackled progressively.

NetMoA service will provide customers with the following benefits:

• Reduction of network damages and service interruptions;
• Cost savings on the aerial/helicopter/vehicle surveys, by limiting them when and where it is really needed;
• Shorter intervention time and optimization of field workforce operations;
• Faster and more cost-effectively planning of new infrastructures.