ELIV-2025 – Efficient Integration of Cross-Platform Functions onto Service-Oriented Architectures

Abstract

The automotive industry is rapidly evolving its Electric/Electronic (E/E) and software architectures in response to growing vehicle complexity. This transformation has led to the coexistence of multiple platforms, including:

• AUTOSAR Classic
• AUTOSAR Adaptive
• ROS 2™ (Robot Operating System)

To address integration and development challenges across these heterogeneous environments, this project presents a concept for building applications as Software as a Product (SaaP). The approach focuses on:

• Designing applications in a platform-agnostic manner
• Standardizing application interfaces
• Describing application and middleware aspects in machine-readable formats (JSON)
• Providing tools for semi-automated development and integration

A demonstration application has been successfully integrated onto AUTOSAR Adaptive and ROS 2, validating the approach and providing efficiency metrics.

---

Mission

Our objectives are to:

• Develop applications agnostic to hardware and software platforms
• Use JSON to describe application and platform aspects
• Enable tool-based integration leveraging these descriptions
• Standardize data interfaces using Vehicle Signal Specification (VSS)

---

Code of Conduct

Please review our Code of Conduct.

---

Features

• Function Model: Pydantic models generate JSON schemas to describe application information independently of the underlying platform.
• Integration Model: Pydantic models describe integration details required for specific platforms (e.g., AUTOSAR Adaptive, ROS 2).

---

Project Language

• Language: Python (tested with v3.10+)

---

Workflow

Below is a sample workflow for the AUTOSAR Adaptive platform using Pydantic models and JSON:

Note: Function and integration models are currently generated manually. Future releases aim to automate JSON generation based on function and platform data.

---

Getting Started

Prerequisites

• Python 3.10 or higher

Installation

pip install pydantic>=2.0 typing-extensions>=4.0

Repository Structure

• src/function_model: Function Pydantic model, example function model, and JSON for an exemplary ADAS application.
• src/integration_model: Integration Pydantic models for AUTOSAR Adaptive and ROS 2 platforms, with example integration models and JSON.
• src/example_adas_application.vspec: Vehicle Signal Specification (VSS) for the exemplary ADAS application.

---

Build and Test

Function Model

• Generate function model schema:

  python function_model.py  # Outputs: function_json_schema.json

• Generate function model for core ACC:

  python function_model_core_acc.py  # Outputs: function_model_core_acc.json

Integration Model – AUTOSAR Adaptive

• Generate integration schema:

  python integration_model_adaptive_autosar.py  # Outputs: integration_schema_adaptive_autosar.json

• Generate integration model for core ACC:

  python integration_model_AA_core_acc.py  # Outputs: integration_model_AA_core_acc.json

Integration Model – ROS 2

• Generate integration schema for ROS 2:

  python integration_model_ros2.py  # Outputs: integration_schema_ros2.json

• Generate integration model for core ACC:

  python integration_model_ros2_core_acc.py  # Outputs: integration_model_ros2_core_acc.json

---

Contribution

We welcome contributions to improve code quality, add features, and expand platform support. Please see CONTRIBUTING.md for guidelines.

TODO: Further instructions for contributing will be added soon.

---

References to other projects

• The Vehicle Signal Specification (VSS) file created is in compliance with the COVESA VSS repo.
• The example Vehicle Signal Specification (.vspec) file is generated using the VSS GUI tool.

---

Publication

For full project details, refer to the upcoming publication.

TODO: Add the link to the publication when available.

---

Contact & Support

For questions or support, please open an issue or contact the maintainers via GitHub.