4.0: Architecture Principles

Key Terms

Key acronyms used in this chapter:

  • ASPICE: Automotive SPICE (Software Process Improvement and Capability dEtermination)
  • AI: Artificial Intelligence
  • CI/CD: Continuous Integration/Continuous Deployment — automated pipelines that build and test code on every change
  • IDE: Integrated Development Environment — the editor and tooling a developer uses to write code (e.g., VS Code, CLion)
  • HITL: Human-in-the-Loop (human oversight pattern)
  • LLM: Large Language Model — the type of AI model underlying tools like Claude or GitHub Copilot

What You'll Learn

After reading this chapter, you will be able to:

  • Apply the "source code as single source of truth" principle
  • Design technology-agnostic processes
  • Adopt the AI augmentation mindset

Chapter Overview

This chapter presents three guiding principles that inform the practical application of ASPICE and AI throughout the book:

  1. Source code as single source of truth — the actual code in version control is the authoritative record of what the software does, not Word documents or PDF exports. Everything else is derived from it.
  2. Technology-agnostic process design — processes are defined in terms of outcomes, not specific tools. When a better tool arrives, you adopt it without redesigning the process.
  3. AI augmentation mindset — AI assists engineers; it does not replace their judgment. The human remains accountable.

Architecture Principles

The Core Insight

These three principles share a common insight:

Focus on what matters, not what is fashionable.

  • Source code matters; specific editors do not
  • Process outcomes matter; specific tools do not
  • Human expertise matters; AI tools serve it

This perspective enables organizations to:

  • Survive technology changes
  • Adapt to new tools without disruption
  • Scale AI adoption thoughtfully
  • Maintain quality regardless of tooling

Principle Interconnection

The three principles reinforce each other:

Architecture Relationships

When tools change:

  • Code remains (source of truth)
  • Processes adapt (technology-agnostic)
  • AI role evolves (augmentation mindset)

Practical Implications

For Development Teams

Principle Practical Impact
Source of truth Use any editor; CI/CD validates
Technology-agnostic Change tools without changing process
AI augmentation Adopt AI at appropriate pace

For Organizations

Principle Practical Impact
Source of truth No vendor lock-in to IDEs — because the process depends on source files, not a specific editor's proprietary format
Technology-agnostic Process investment persists
AI augmentation Controlled AI rollout

For AI Integration

Principle Practical Impact
Source of truth AI works on source, not artifacts
Technology-agnostic AI tools can be swapped
AI augmentation AI enhances, does not replace

Chapter Sections

The following sections provide detailed coverage:

  • 04.01: Source code as single source of truth
  • 04.02: Technology-agnostic process design
  • 04.03: The AI augmentation mindset
Previous
Tool Qualification for AI
Next
Source Code as Single Source of Truth