Objectives of the service
The objective of the services is to provide bespoke satellite-based analyses to support the management practices of Transmission and Distribution System Operators (TSOs and DSOs, respectively).
A safe, secure and reliable supply of electricity is vital, therefore the electrical grid is a crucial element. TSOs have to control the safety of the infrastructures through regular monitoring and maintenance. In wooded areas, vegetation can quickly become a threat to the network safety: a tree falling on high voltage lines or a pylon, vegetation growing underneath the power lines or branches coming too close to them. Non-authorised activities that occur in the Right Of Way (ROW), i.e. the strip of land immediately below and adjacent to a transmission line, can also pose severe risks to the infrastructures. The cost impact of traditional change detection and vegetation management services is significant. Therefore, the challenge for the TSOs is to guarantee maximum safety while striving on saving costs.
For this purpose, GMV NSL will develop pioneer automatic services to leverage Earth Observation (EO) data and state-of-the-art Machine Learning algorithms to provide measurable and objective information along the power lines.
Users and their needs
GMV NSL services are targeted at Transmission and Distribution System Operators of electricity across Europe.
The services have been designed to meet users' needs, identified through conversations with the stakeholders involved in the project, i.e. nine TSOs belonging to ENTSO-E, the European Network of Transmission System Operators for Electricity.
In particular, users have expressed the need to:
ensure the continuous supply of power;
ensure the safety of infrastructure and operators, through efficient monitoring of the network at a limited cost;
guarantee smart scheduling of maintenance tasks;
decrease the risk of network failure by efficiently managing vegetation along power lines;
be aware of activities that occur in the ROW;
rapidly deal with the impacts posed by disasters;
smoothly integrate new services into their work practices.
The project faces numerous challenges, among which are the highly accurate estimation of vegetation and object parameters, the variety of environmental and meteorological conditions across Europe and the fast track image acquisition, processing and delivery in case of disasters.
The services will hence make use of cutting edge machine learning approaches to extract as much information as possible from satellite imagery and generate easily readable geospatial products.
Service/ system concept
The vision for the services is to provide users with accurate information about what is happening or has happened, and which is the status of the vegetation (trees, shrubs) along the thousands of kilometres of their power network.
The three services aiming to address the needs of the TSOs are:
Vegetation Management: large scale monitoring and prioritisation support for vegetation management (e.g., pruning, clipping) in the ROW;
Change Detection: automatic detection and alert of anomalies (e.g., man-made objects, disposal of vegetation) in the ROW;
Disaster Management: fast provision (few days) of geospatial information in support of post-disaster management activities, immediately following a disaster impacting the energy network infrastructure.
The services strive to contribute to ensuring a secure and reliable supply of electricity, while reducing costs and enhancing management efficiency, by complementing traditional management of the energy network infrastructure with digital mapping, monitoring and change detection based on satellite observations.
Satellite imagery is the core source of data for the generation of geospatial products, and the backend system is a suite of information extraction and enrichment pipelines.
During the project, use cases will be implemented to showcase the services in a real-world scenario.
Space Added Value
The services will make use of multi-source datasets where Earth Observation (EO) data will be the key asset, mainly but not limited to very high resolution (VHR) and high resolution optical imagery.
GMV NSL will use advanced EO and machine learning techniques, as well as in-situ observations of tree height or objects in the ROW, when made available by the customer, to validate the analyses. Some examples of applications are:
Satellite data and change detection analyses to map trees, shrubs and activities that occur in the ROW. The image below shows a vegetation mask map derived from VHR images.
Automatic analysis of satellite-derived data to inform about the risk posed to transmission lines by vegetation or man-made objects in the ROW. Risk thresholds are defined taking into account the TSOs’ management practices. The following image shows an early alert map indicating risk zones.
The project is the demonstration phase of a previous ESA feasibility study, and kicked off at the beginning of February 2022. Since then, the SMARTGRIDS-DEMO team held interviews with single users to gather additional needs if any and consolidate the requirements.
The team has also been busy designing the algorithms pipelines the services rely upon. First activities of developments have also started, with the aim of adapting and evolving any in-house components and developing new ones.
A User Meeting was successfully held at the beginning of May to show the users the first preliminary results coming from these development activities.
The BDR - CDR milestone is scheduled for the end of May 2022.