The energy turnaround and the shift towards sustainable mobility threaten the stability of European energy distribution grids due to substantially increasing load fluctuations and power demand. These challenges can critically impact assets in the distribution grid—e.g., switchgears—intensifying the need to plan, conduct, and manage the maintenance of such assets. Predictive maintenance strategies that analyze assets' current and historical condition data have been discussed as promising approaches toward that end. However, the extant research focuses on designing and improving analytical algorithms or information technology (IT) artifacts while not considering how a maintenance service is cocreated by companies with IT. This research article posits that IT and service must be aligned closely, presenting an ensemble artifact comprising a digital industrial platform and a smart service system for predictive maintenance on the distribution grid. The artifact is evaluated by conducting a willingness-to-pay analysis with asset operators, documenting their demand for condition monitoring and predictive maintenance as an integrated solution, although they still struggle with even getting the condition data of their assets. Building on these results, we formalize the knowledge in the form of design principles and implications for managing the maintenance of critical assets in the distribution grid.

In recent times, both geopolitical challenges and the need to counteract climate change have led to an increase in generated renewable energy as well as an increased demand for clean electrical energy. The resulting variability of electricity production and demand as well as an overall demand increase, put additional stress on the existing grid infrastructure. This leads to strongly increased maintenance demands for distribution system operators (DSOs). Today, condition monitoring is used to address these challenges. Researchers have already explored solutions for monitoring critical assets like switchgear and circuit breakers. However, with a shrinking knowledgeable technical workforce and increasing maintenance requirements, mere monitoring is insufficient. Already today, DSOs ask for actionable recommendations, optimization strategies, and prioritization methods to manage the growing task backlog effectively. In this paper we propose a vision of a grid-level cognitive assistance system that translates the outcome of diagnosis and prognosis systems into actionable work tasks for the grid operator. The solution is highly interdisciplinary and based on empirical studies of real-world requirements. We also describe the related work relevant to the multi-disciplinary aspects and summarize the research gaps that need to be closed over the next years.

Traditional dyadic customer-provider interactions are being shifted to polyadic interactions involving diverse participants in digital service ecosystems. Especially, artificial intelligence (AI) is increasingly integrated into these ecosystems, so that they comprise non-human participants (e.g., AI-based chatbots)—fundamentally altering the nature of value (co-)creation. While existing literature examines human-to-human interactions, knowledge of service interactions between human actors and AI-based systems is still underexplored. To address this research gap, we develop a taxonomy, comprising six iterations, that explores the peculiarities of AI as either a resource or a (non-human) agent in digital service ecosystems. We evaluate our taxonomy using a multiple case study and derive the four archetypes of AI in digital service ecosystems: (1) discriminative experience enhancer, (2) protective ecosystem orchestrator, (3) ecosystem innovation companion, and (4) personalized service composer. Our results extend the knowledge on service science by showing how AI-based systems—discriminative or generative, and focusing on the interaction in the ecosystem or the individual service encounter—assume the role of resources and non-human agents. Researchers and practitioners can utilize our results to augment their ecosystems with AI.

Zusammenfassung: Verteilnetzbetreiber in Deutschland stehen vor großen Herausforderungen bei dem Management ihres unternehmensspezifischen Wissens: Mitarbeiterengpässe durch den demographischen Wandel, Wissen ist nur implizit vorhanden und nicht in Wissensmanagementsystemen digitalisiert, teilweise gibt es gar keine Wissensmanagementsysteme oder Konzepte und das Verteilnetz wird immer komplexer. Verbunden mit zunehmender Belastung von zentralen Komponenten im Verteilnetz durch die Energiewende bedarf es neuer Lösungen, besonders für die wissensintensiven Wartungs- und Instandhaltungsprozesse. Generative Artificial Intelligence als aufstrebende Technologie, insb. durch Large Language Models, zeigt hier erste Erfolge für die Anleitung, Entscheidungsunterstützung und den Wissenstransfer. Aufbauend auf dem Design Science Research Forschungsparadigma wird in diesem Beitrag ein ganzheitlicher Ansatz des Wissensmanagements konzipiert, welcher als zentrale Komponente auf einem Assistenzsystem basiert. Ein Large Language Model generiert Hilfestellungen für Netzmonteure während der Wartung und Instandhaltung auf Basis von Anleitungen. Neben der Konzeption zeigt dieser Beitrag auch die erarbeitete Strategie zur Demonstration und zukünftigen Evaluation der Ergebnisse. Der Beitrag liefert ein für Verteilnetzbetreiber neuartiges Konzept Large Language Model basierter Assistenzsysteme zum Wissensmanagement und zeigt zudem nachgelagerte Schritte auf, die vor einer Markteinführung notwendig sind.