Episode Description

Welcome to the Embedded AI Podcast! In our inaugural episode, Ryan and Luca introduce the podcast and dive into what it means to use AI in embedded systems.

We discuss:

• AI on embedded devices: Traditional machine learning, edge computing, and predictive maintenance (including a cool example of acoustic monitoring)
• AI for developers: Using LLMs and AI tools in embedded development workflows
• Real-world applications: From aerospace conferences to rocket launch coordination
• Practical challenges: Getting LLMs to write STM32 code, the importance of TDD, and staying in control of AI-generated code

This is a conversation about learning together - expect frank discussions about what works, what doesn’t, and plenty of mistakes to learn from.

Listen

Topics Covered

• The dual focus of the podcast: AI on devices AND AI in development tools
• Traditional AI/ML vs LLM applications in embedded systems
• Real examples from the German Aerospace Conference
• Contrail detection using AI in satellite imagery
• LLM-assisted rocket launch operations and air traffic control
• Sound-based predictive maintenance on industrial machines
• The challenges of “vibe coding” and LLM-assisted development
• Why test-driven development becomes even more important with AI
• How AI is changing the role of embedded engineers
• The importance of slowing down LLMs and staying in control

Hosts

Ryan Torvik - Software engineer with over two decades in cybersecurity and embedded systems. Former principal engineer at Raytheon Intelligence & Space, now founder of Tulip Tree Technology, building CodeForge - an AI-powered edge-case discovery tool for embedded systems.

Luca Ingianni - Aerospace engineer working across embedded systems in aerospace, automotive, medical, and industrial sectors. Agile and DevOps practitioner since 2009, now coaching teams on AI integration in embedded development. Co-host of the Agile Embedded Podcast and TÜV-certified in AI.