To address the significant safety hazard of brake air source failure caused by potential faults in the piston-type air compressor of new electric heavy-duty trucks, this paper designs and implements an online health monitoring and early warning system based on multi-source information fusion. The core innovation lies in constructing a health assessment model that combines weighted scoring and trend prediction, fusing multi-dimensional parameters into a quantified Health Index (HI). Based on the HI value and its degradation trend, the system executes a multi-level warning strategy, ranging from "Reminder" to "Emergency," and can intervene with the vehicle control system to limit speed and activate auxiliary braking in critical situations. This study details the system architecture, the construction method of the health model, and the workflow, demonstrating that the system transforms passive alarms into active predictions, significantly enhancing the active safety and operational reliability of commercial vehicles.