Publikation
Beyond One-Fits-All: A Case Study Approach to AI System Design Methods
Sabine Janzen; Hannah Stein
In: Proceedings of International Workshop on AI Services and Applications (AISA’2024) at 43th International Conference on Conceptual Modeling (ER). International Conference on Conceptual Modeling (ER-2024), October 28-31, Pittsburgh, USA, 10/2024.
Zusammenfassung
Despite the widespread application of artificial intelligence (AI) as universal solution for complex business problems, there remains a significant gap in design methods for AI systems, distinguishing them sharply from traditional software systems. This research aims to address the lack of standardized design methods tailored for AI projects, which are often impeded by unique challenges such as data sensitivity, model performance, and regulatory compliance. Through an exploratory case study of four AI projects, this paper investigates correlations between characteristics of AI projects and the design methods applied, introducing a set of If-This-Then-That (IFTTT) patterns. These patterns are intended to aid in selecting and combining design method components that align with the specific needs of AI projects. Results highlight the importance of understanding project-specific characteristics to enhance the effectiveness of design methods in AI engineering, offering practitioners actionable insights for improving quality and reliability of AI systems through tailored design approaches.
Projekte
SMARF - Smart Farming Welt - Herstellerübergreifende Vernetzung von Maschinen im landwirtschaftlichen Pflanzenbau mithilfe einer Service-Plattform,
SPAICER - Skalierbare adaptive Produktionssysteme durch KI-basierte Resilienzoptimierung,
PAIRS - Privacy-Aware, intelligent and Resilient Crisis Management
SPAICER - Skalierbare adaptive Produktionssysteme durch KI-basierte Resilienzoptimierung,
PAIRS - Privacy-Aware, intelligent and Resilient Crisis Management