CI/CD-Driven EOL Calibration Automation for Battery Electric Commercial Vehicles

Abstract

End-of-Line (EOL) calibration and verification processes for battery electric commercial vehicles (BEVs) have become increasingly complex as heavy-duty vehicle platforms transition from mechanically dominated architectures to software-defined and electronically coordinated propulsion systems. Manual EOL procedures used in conventional commercial vehicle manufacturing environments are difficult to scale for modern Class 8 BEV platforms because calibration activities now span multiple interdependent electronic control units (ECUs), including the Vehicle Control Unit (VCU), inverter, Battery Management System (BMS), Electric Vehicle Communication Controller (EVCC), DC-DC converter, and Electric Power Take-Off (E-PTO). Manual execution introduces risks associated with sequencing errors, inconsistent traceability, incomplete audit evidence, and configuration mismatches across vehicle variants. This paper presents a Continuous Integration and Continuous Delivery (CI/CD)-driven EOL calibration automation framework for heavy-duty battery electric commercial vehicles integrating Python-based orchestration with Jenkins pipeline management for structured execution, automated reporting, and audit-ready traceability. The framework supports calibration and verification of multiple ECUs through SAE J1939/CAN and Unified Diagnostic Services (UDS) communication over standard diagnostic interfaces. Structured JSON logging, Git-based evidence archiving, and CI/CD dashboard reporting are incorporated to provide production-level traceability and compliance support. Results from deployment within a representative Class 8 BEV EOL environment indicate approximately 20% reduction in manual calibration effort, improved configuration consistency, reduced error escape rates, and enhanced audit traceability compared with conventional manual workflows. The paper contributes the first openly described CI/CD-driven EOL calibration automation framework integrating Python orchestration, Jenkins pipeline management, J1939/CAN and UDS diagnostic communication, and structured audit traceability specifically for multi-ECU Class 8 BEV commercial vehicle production environments.