AAA: AGL Assessment Automation Project

AAA: AGL Assessment Automation Project

Overview

AGL Assessment Automation (AAA) is an initiative launched as part of the Automotive Grade Linux (AGL) CIAT-EG activities. The goal of AAA is to build a reference system and best practices for applying SBOM (Software Bill of Materials)-driven cybersecurity approaches, developed in the OpenSSF project, to the automotive and embedded industries.

While OpenSSF’s SBOM practices are currently optimized for cloud-based tech industries, direct application to automotive and embedded domains presents unique challenges. AGL leverages the widely adopted Yocto build system and a Jenkins-based CI/CD pipeline, making it possible to develop reusable practices and reference systems tailored to embedded environments.

image-20251109-032343.png

 

Key Activities

AAA focuses on the following two pillars:

  1. Establishing SBOM Workflows Utilizing Yocto’s SPDX 3.0 Support

    • Standardizing workflows for SBOM verification and publication using SPDX 3.0 SBOMs generated by the Yocto build system.

    • Identifying current challenges in Yocto and contributing improvements upstream.

  2. Automating Workflows through Policy Based Automation (PBA)

    • Leveraging generated SBOMs and technologies such as Open Policy Agent (OPA) to automate workflows for license compliance and security assurance.

    • Employing advanced validation and automation techniques, including SHACL/SPARQL, to enhance quality assurance.

Materials

 

Project Members

  • Takashi Ninjoji - Project Lead, Honda Motor

  • Taiki Kawamura - Honda Motor

  • Masanori Itoh - Toyota Corporation

  • Hiroyuki Ishii - Panasonic Automotive Systems

  • Jan-Simon Möller - Linux Foundation

 

Regular Meetings

English: Every 2 weeks (even) on Wednesday, 13:00 - 14:00 UTC (ICAL)

Japanese: Every 2 weeks (odd) on Wednesday, 17:00 - 18:00 JST (LFX Meetings)