Zero Touch Predictive Orchestration: Automating Time-Series Models for the Cloud-Edge Continuum

Distributed resources in the Cloud-Edge Continuum require proactive management due to extreme volatility, making Zero Touch Predictive Orchestration essential for latency-critical applications. A significant challenge arises when newly discovered nodes lack historical data to train localized predictive models, while generalized models are insufficient. Researchers propose automating time-series models to address this "cold start" problem, enabling efficient forecasting and proactive management. By leveraging automated time-series models, orchestrators can effectively manage the Cloud-Edge Continuum without requiring extensive historical data. The proposed approach focuses on distributing resources to the far edge, ensuring low latency and high performance for critical applications. This development has significant implications for cloud-edge continuum management, as it enables proactive and efficient decision-making¹. So what matters to practitioners is that automated time-series models can enhance the reliability and performance of latency-critical applications in the Cloud-Edge Continuum.