This thesis, conducted in collaboration with Metatron S.p.A., presents a model-based approach for expanding and redefining an API layer for automotive Electronic Control Units (ECUs), focussing on two key aspects: memory management and on-board diagnostics (OBD). The first part of this work addresses memory management, where a robust, multi-module strategy was developed and implemented to efficiently manage the non-volatile memory (NVRAM). The found solution ensures the secure storage and retrieval of safety-critical data, essential for the reliable operation of automotive systems. The second section introduces a novel OBD strategy, designed to detect, validate, store, and communicate faults in compliance with current emission regulations and heavy-duty vehicle protocols.