KBMP: Kubernetes-Orchestrated IoT Online Battery Monitoring Platform

The rise in renewable energy has driven the widespread use of large-scale energy storage batteries, which makes the risk of overheating more threatening. To ensure battery safety, it is essential to build a monitoring system with a comprehensive evaluation of large quantities of batteries. However, existing battery management systems exhibit significant limitations in terms of monitoring scope, analytical precision, and transmission efficiency. As an applicable solution, cloud-edge technology is an advanced integrated method that provides low-latency data access, accurate analysis capabilities, and adjustable monitoring ranges. In this work, the Kubernetes-Orchestrated Battery Monitoring Platform (KBMP), which integrates Kubernetes and cloud-edge technology, is proposed to provide comprehensive battery management. Specifically, Kubernetes is used to ensure low latency in data transmission and analysis, while the K-Means clustering algorithm is applied to provide accurate thermal runaway (TR) warnings. To validate the performance of KBMP, four sets of real battery TR data are fed to test its accuracy and latency. The experimental findings reveal that KBMP is capable of providing battery thermal runaway warnings in advance within 30 minutes. Additionally, the platform concurrently decreases data transmission latency by up to 20% and reduces replica scaling latency by 50% compared to the platform without integrating Kubernetes.