A Decentralised Relational Data Model for Reliability Studies of Medium-Voltage Cables

The distribution grid is evolving into a multilateral cyber-physical system where efficient data sharing and management across stakeholders are fundamental. Consequently, many applications benefit from collecting data from decentralised sources, as in the case of data-driven reliability prediction of medium-voltage cables, where distribution grid operators expect more accurate results through data sharing and joint model development. However, challenges arise when transitioning from centrally stored data to distributed data handled by decentralised peers. Current data collection processes often involve manual and static data requests, requiring additional efforts in data combination or harmonisation. To address these challenges, this study proposes a relational database model tailored for the reliability prediction of medium-voltage cables, and discusses its evolution into a decentralised database design. Eventually, such a setup may not only facilitate the development of ML applications but also guide the way for more standardised and dynamic data sharing among distribution grid operators.