The Art of Testing Cars and Delighting Drivers
Welcome to the era of rolling autonomous death machines, also known as the era of the software-defined autonomous vehicle. As companies like Tesla have proven, modern car buyers value software driven experiences at a growing rate. No longer are cars just about horsepower but are now really about personalization, entertainment, and comfort. They are also all becoming at least semi-autonomous with level 4 autonomy coming within the decade. In this new world, continuous deployment and deep learning AI will change the user's experience with the modern vehicle. Even today, all modern cars receive software updates multiple times a year, sometimes to fix bugs and sometimes to ship improved features like automated lane change for autonomous mode. Ken will share how Ford invests in big upfront testing but still leverages Minimum Viable Quality with Testing in Production as methods to safely release and measure the quality of all updates. Complexities of extreme heat and cold, continuous vibration, and dealing with regulatory compliance, along with the risk of injury and actual death, play complex roles in test automation and regression testing strategies. Combine this with a big auto company challenging itself to compete with smaller, faster startup competitors like Tesla, Lucid, and Rivian and you will find established companies like Ford going through massive skill and culture change. Autonomous vehicles are coming quickly. Continuous Deployment and Continuous Integration are required to compete in the future. Getting software quality right is the next big challenge!
Ken Johnston is a Founder of the AiGovOps Foundation, a nonprofit practitioner community advancing AI Governance as Code — the discipline of embedding governance controls into AI systems through engineering rigor, automation, and operational feedback loops rather than policy documents and good intentions. With over 30 years of engineering and data leadership, Ken brings rare depth across the full stack of responsible AI deployment: software quality, cloud platforms, telemetry, observability, and applied AI at scale. As a former GM and Principal Data Science Manager at Microsoft, he helped build foundational systems for A/B experimentation, product telemetry, and the M360 Business Intelligence Graph. He went on to serve as VP of Cloud Platforms and Telematics at Ford and CEO of Autonomic.ai, where he led scalable data-driven innovations in connected vehicle technology. Ken is coauthor of How We Test Software at Microsoft. His upcoming book, The Lean AI Handbook, Pierson summer 2026, captures best practices for operationalizing AI at scale. He currently serves as Advisor and VP of AI at Envorso, helping enterprise clients architect responsible, scalable AI solutions that drive measurable impact.