The Impact of Artificial Intelligence on Quality in Smart Grid Requirements Engineering

The growing complexity of the energy sector as a system of systems has led to structured approaches such as the IEC 62559 Use Case Methodology and the Smart Grid Architecture Model (SGAM) to tackle interoperability challenges as major fundament of the smart grid requirements engineering. Smart grid requirements engineering holds a unique role, as it must address the increasing diversity of actors, business models, and technologies, requiring particularly robust and adaptive methods to ensure coherence and interoperability across interconnected, independently managed systems. Artificial Intelligence (AI) is increasingly used in engineering, including requirements engineering, where it improves consistency, completeness, and traceability. This article hypothesizes that AI can similarly enhance the quality of smart grid requirements engineering and deepen system understanding. To investigate this hypothesis, we introduce and evaluateSmart Grid Assisting Artificial Intelligence in Requirements Engineering, a prototypical AI-assisted system designed to support users in creating, analyzing, and validating IEC 62559 Use Cases and SGAM Models. SGAAIRE operationalizes AI-based quality assurance by integrating reasoning and validation mechanisms tailored to the specific structures and semantics of these artifacts. Within a randomized controlled trial, conventional modeling approaches were compared with the AI-assisted method implemented in SGAAIRE. The results demonstrate that integrating AI into the quality assurance process can improve the quality of IEC 62559 Use Cases and SGAM Models, particularly among experienced users. AI can aid especially in the dimensions of correctness, consistency, and traceability – without reducing user awareness for additional, non-obvious errors.