Digital Transformation

We have a strong focus on sensing our assets and the data available, on capturing the value of this information to improve our stakeholders' experience and on the efficiency of asset and operations management. The provision of a consistent and secure data platform, as well as new intuitive tools for better access and exploration of information and new advanced analytical models for decision support. 

We use several advanced analytics techniques to find solutions to complex problems and volumes of data impossible to process on a human scale – Data Mining, Process Mining, Neural Networks, Artificial Intelligence, among others. 

Our central position in the electricity sector and the information that smart grids enable nowadays also allow us to define a repository of relevant information so that any entity can explore our platform – see the 'Open Data' area for available datasets.  

How do we do it? 

• Increasing fire risk makes vegetation management and its impact on the continuity of essential services increasingly complex. We have created analytical models that combine Machine Learning and Computer Vision capabilities to analyse vegetation proximity to lines, predict vegetation growth and determine intervention needs, instructing field trips and cut-off operations based on the severity of the situations.
• Climate change induces increasingly frequent extreme events, with significant impacts on the electricity grid. Portugal is particularly affected given its large extension of the overhead network most exposed to these phenomena. To this end, we have created Machine Learning models to predict network incidents, affected customers and degree of severity by geographical area. These have improved our resource allocation plans and the information made available to the various stakeholders. The PREDICTIVE GRID project won the BEST FUTURE OF INTELLIGENCE PROJECT and BEST ENERGY & UTILITIES PROJECT awards at the Portugal Digital Awards 2020.
• E-REDES manages a total of 230,979 km of network, with critical assets scattered throughout the country. Our investment and maintenance plans for Network Assets are based on information resulting from a combination of artificial intelligence capabilities such as Big Data, Natural Language Processing (NLP), Computer Vision and Machine Learning to extend their useful life and minimize environmental impact.
• Around 69% of our High and Medium Voltage assets are already managed using analytical models.
• Machine Learning models calculate the probability of failure in critical assets (e.g. the probability of a circuit breaker failing at the next manoeuvre).
• Big Data algorithms that calculate the health and risk index of critical assets and remaining lifetime based on the Common Network Asset Indices Methodology (CNAIM).
• Natural Language Processing (NLP) models analyse comments written by Service Providers on breakdown orders, inferring from them the most affected assets, the damaged elements and their exact location.

Computer Vision models analyse thousands of photographs collected in the field to automatically identify anomalies in the inspected assets.