Aviation Masters
In this episode of Aviation Masters, Mike Busch sits down with Dr. John Sipple — machine learning professor at George Washington University, former Boeing Phantom Works engineer, Google AI veteran, and Diamond DA40 owner — for one of the most forward-looking conversations in the show's history. From tracking ballistic missiles with Kalman filters to training neural networks at Google to building AI-powered diagnostics for GA aircraft, John brings a uniquely deep and practical perspective on where artificial intelligence is headed in aviation. The conversation covers everything from how to certify ML models with the FAA, to why technicians struggle with troubleshooting, to the Diamond DA40 oil pressure failures that inspired John to found Vyzerion AI. If you own, maintain, or fly a piston aircraft, and especially if you've ever wondered what AI can actually do for GA safety, this episode is essential listening. ⏱ Chapters 00:49 – Introducing Dr. John Sipple 05:30 – Boeing Phantom Works: Missile Tracking & Kalman Filters 09:30 – From Big Data to Machine Learning: The Career Pivot 11:45 – AI vs. Machine Learning vs. Deep Learning: Defining the Terms 14:00 – How Neural Networks Work & Why 2012 Was a Turning Point 18:00 – Applying ML to Telemetry: From Missiles to Aircraft Engines 26:00 – What Is Explainable AI (XAI) and Why It Matters in Aviation 36:00 – Anomaly Detection in Piston Engines: How It Works 47:00 – Project Spark: Savvy Aviation's AI Diagnostics Initiative 53:00 – Training vs. Inference: Why the FAA Draws a Hard Line 55:00 – How You Certify an ML Model for Aircraft Use 59:00 – Why Tesla FSD Couldn't Happen in Certificated Aviation 1:00:00 – Experimental Aircraft as the Path to AI in the Cockpit 1:05:00 – Three Lines of Defense for AI in Mission-Critical Systems 1:09:00 – The Vision for Self-Diagnosing Airplanes 1:13:00 – Why Mechanics Struggle with Troubleshooting (And How AI Fixes It) 1:18:00 – Project Squawk: Training an LLM on Aviation Maintenance Data 1:19:00 – The Data Funnel: Small Models → Explainability → LLM Reasoning 1:24:00 – The Cadillac Index & Why GA Affordability Is Getting Worse 1:25:00 – Autonomous Aircraft vs. Autonomous Cars: Why Is GA So Far Behind? 1:27:00 – Vyzerion AI: Smarter Diagnostics, Healthier Aircraft, Safer Flying 1:28:00 – The Diamond DA40 Service Bulletin & In-Flight Oil Pressure Failures 1:35:00 – Broader AI Fears, Regulation & the Elevator Analogy 1:48:00 – Multimodal AI & the Future of Richer Diagnostics 1:49:00 – Using LLMs to Query the NTSB Accident Database 1:51:00 – Annual Inspections: Are They Helping or Hurting Safety? 📍 Subscribe to Aviation Masters for more long-form conversations with the people shaping the future of GA. 🔗 RESOURCES MENTIONED IN THIS EPISODE Vyzerion AI —https://vyzerionai.com/ [https://vyzerionai.com/] Savvy Aviation — https://savvyaviation.comsavvyaviation.com [http://savvyaviation.com] NTSB Aviation Accident Database - https://carol.ntsb.gov/ [https://carol.ntsb.gov/]
6 episodios
Comentarios
0Sé la primera persona en comentar
¡Regístrate ahora y forma parte de la comunidad de Aviation Masters!