Designing AI-Driven SAP Systems for Intelligent Supply Chain Optimization Across Cloud and IoT Platforms
Authors-Hayk Sargsyan
Keyword-Artificial Intelligence, SAP S/4HANA, Supply Chain Optimization, Internet of Things (IoT), Cloud Computing, SAP Business Technology Platform (BTP), Predictive Analytics.
Modern supply chain management has entered an era of permanent volatility, requiring a fundamental transition from reactive automation to proactive, intelligent orchestration. This review article investigates the design and implementation of AI-driven SAP systems that leverage the convergence of Cloud computing and the Internet of Things (IoT) to achieve autonomous optimization. Central to this architecture is the SAP Business Technology Platform (BTP), which serves as a unified data fabric for harmonizing high-velocity telemetry from IoT sensors with the transactional integrity of the S/4HANA digital core. The study evaluates the application of diverse Machine Learning (ML) models, ranging from LSTM-based demand sensing in SAP Integrated Business Planning (IBP) to reinforcement learning for real-time logistics rerouting. A significant focus is placed on the emergence of Agentic AI and the SAP Joule copilot, which move beyond traditional decision support to execute multi-step, self-healing workflows across procurement and warehouse management. Furthermore, the article examines the role of MLOps in managing model drift within dynamic global markets and the strategic importance of a "Clean Core" approach to ensure long-term system agility. By synthesizing implementation best practices with future directions such as quantum-assisted routing and the "Green Ledger" for carbon-aware accounting, this research provides a comprehensive framework for architecting resilient, transparent, and self-optimizing supply chain ecosystems. We conclude that the successful integration of AI, Cloud, and IoT is the primary prerequisite for operational excellence and competitive survival in the contemporary digital economy.
Doi-[http://doi.org/10.5281/zenodo.19470023]