Towards explainable and trustworthy autonomous models for power systems: a retrospective case study analysis

Currently, power systems are undergoing a digitalization shift to manage an increasingly heterogeneous and volatile network, incurred by the energy transition, which directly impacts control centres. However, trust in automated infrastructure management through an understanding of the model behaviour is decisive in making operational and scalable decisions. Therefore, there is a need for transformative changes aided by trustworthy, explainable automation, particularly through human-centred design, to empower control room operators. This paper assesses the expected shift towards trustworthy industries based on recent policies and standards development, supported by research. Through a switchgear asset management case study, it evaluates the trustworthy challenges and expected processes to provide recommendations.