EUVIMEDEUVIMED entdecken
Crossref

Multi-Agent Generative AI Framework for Autonomous Exception Management in SAP S/4HANA Manufacturing

Haruto Nakamori, Aiko Fujimori

American Journal of Applied Science and Technology · 22.08.2026 · 6 (08) · 167-174

DOI: 10.37547/ajast/volume06issue08-13 ↗

Kurz gefasst

Fachartikel in der Zeitschrift American Journal of Applied Science and Technology, online erschienen am 22.08.2026. Verfasst von Haruto Nakamori und 1 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Modern manufacturing environments generate large volumes of operational exceptions involving production disruptions, material shortages, quality deviations, foreign-object detection, equipment anomalies, and process interruptions. Enterprise resource planning systems such as SAP S/4HANA Manufacturing provide structured transactional information for production planning and execution, but conventional exception management frequently remains dependent on human monitoring, rule-based alerts, and sequential investigation. This paper proposes a Multi-Agent Generative AI Framework for Autonomous Exception Management in SAP S/4HANA Manufacturing that integrates specialized intelligent agents for exception detection, contextual diagnosis, impact assessment, decision generation, and controlled resolution. The conceptual foundation is derived from the provided literature on machine-vision-based foreign-object detection, anomaly detection, three-dimensional perception, fault early warning, and industrial defect analysis. These studies demonstrate the value of automated perception and intelligent anomaly identification across heterogeneous manufacturing environments. Building upon these principles, the proposed framework organizes manufacturing exceptions into an agent-based decision pipeline connected to SAP production and material processes. The framework emphasizes autonomous reasoning while maintaining transactional controls through confidence thresholds, approval gates, and exception escalation. The resulting model indicates that multi-agent coordination can transform isolated anomaly signals into contextualized manufacturing decisions, thereby reducing response latency and improving operational resilience. However, limitations remain concerning data quality, explainability, integration complexity, model hallucination, and authorization of autonomous actions. The study contributes a research-oriented architecture for integrating generative AI agents with enterprise manufacturing exception management.

Zitieren

APA

Nakamori, H., & Fujimori, A. (2026). Multi-Agent Generative AI Framework for Autonomous Exception Management in SAP S/4HANA Manufacturing. American Journal of Applied Science and Technology, 6(08), 167-174. https://doi.org/10.37547/ajast/volume06issue08-13

Vancouver

Nakamori H, Fujimori A. Multi-Agent Generative AI Framework for Autonomous Exception Management in SAP S/4HANA Manufacturing. American Journal of Applied Science and Technology. 2026;6(08):167-174. doi:10.37547/ajast/volume06issue08-13

BibTeX
@article{Nakamori2026,
  author = {Nakamori, Haruto and Fujimori, Aiko},
  title = {Multi-Agent Generative AI Framework for Autonomous Exception Management in SAP S/4HANA Manufacturing},
  journal = {American Journal of Applied Science and Technology},
  year = {2026},
  volume = {6},
  number = {08},
  pages = {167--174},
  doi = {10.37547/ajast/volume06issue08-13}
}
Kostenlos starten →