MEDUNATECH

SOVD/MCP Server

Bridging Vehicle Diagnostics with Agentic Software Systems

Overview

The SOVD/MCP Server is a gateway solution that connects Service-Oriented Vehicle Diagnostics (SOVD) with the Model Context Protocol (MCP), enabling vehicles to seamlessly integrate into modern, agent-based software ecosystems. It exposes vehicle diagnostic capabilities as discoverable, secure, and callable tools that can be consumed by cloud services, backend systems, and AI-driven agents.

This solution is designed for software-defined vehicle (SDV) architectures and supports both current and future automotive platforms.

 

The Solution

The SOVD/MCP Server acts as an intelligent middleware that:

  • Uses SOVD (ASAM standard) to model and execute vehicle diagnostics

  • Uses MCP to expose these capabilities as agent-consumable tools

  • Preserves full backward compatibility with UDS

Diagnostic services defined via DEXT are automatically mapped to MCP tools, enabling external agents to discover and invoke vehicle functions without ECU-specific knowledge.

Overview of the SOVD/MCP Architecture

Key Capabilities

  • Online Capability Discovery
    Vehicle diagnostics are discoverable at runtime via MCP, reflecting the actual vehicle configuration and software state.

  • Agent-Ready Interface
    Vehicle functions are exposed as MCP tools, ready for AI, LLMs, and autonomous agents.

  • Secure and Controlled Access
    Authentication and authorization are enforced through MCP, mapped to SOVD diagnostic security levels.

  • UDS Compatibility
    Existing UDS-based diagnostics are reused as the execution backend.

  • SDV-Ready Architecture
    Decouples diagnostics from hardware topology, supporting centralized and zonal vehicle architectures.

 

The SOVD/MCP Server provides a practical and standards-based path to integrate vehicles into modern agentic software environments. By combining ASAM SOVD with the MCP protocol, it enables a true software-defined diagnostic interface that connects vehicles with today’s state-of-the-art AI and cloud ecosystems.