Gasgoo Munich- "Getting different restaurant recommendations in the cockpit is one thing; but in vehicle diagnostics, AI must guarantee deterministic correctness."Speaking at the AI-Defined Vehicle Forum 2026 on September 17, Sonatus co-founder and CTO Fang Yu argued that while current industry discussions on AI-defined cars focus heavily on intelligent driving and human-machine interaction, truly AI-native cars are missing "the other half"—extending the same intelligence across the entire lifecycle from R&D, testing, and production to after-sales.In Fang's view, a core problem with traditional automotive development is that information is fragmented across departments, tools, and data systems. R&D, testing, production, and after-sales each operate with their own software, databases, and workflows, leaving vast amounts of knowledge trapped in source code, models, Jira tickets, historical test records, and vehicle data, making seamless flow difficult."On the surface, it looks like a technology choice, but the root cause lies in human limitations." Fang explained that an individual can only understand and process so much information, and time is equally finite, making it hard to fully compile and transmit all knowledge accumulated in one phase to the next team.AI offers another possibility. By connecting R&D materials, test data, historical records, and vehicle data, AI can understand and correlate information across different tools, further assisting in generating problem analyses, test cases, and potential solutions.Fang cited a real-world case at Nissan's Technical Center, where combining R&D phase information with vehicle test data and using AI for analysis slashed the time engineers needed to pinpoint vehicle faults from roughly two weeks to just two days.This information flow shouldn't be one-way. Insights accumulated during R&D can be passed downstream to help test teams design more targeted schemes, and serve production line End-of-Line tests and after-sales diagnostics. Conversely, batch issues discovered in after-sales should be quickly fed back into the R&D system.Sonatus' current Fastlane product series is built around this logic, including Fastlane Collector for vehicle data acquisition, Fastlane Edge for on-board edge computing, and Fastlane Insight for cloud-based AI Agent analysis. These three products can be used independently or combined to form a continuous "Observe-Analyze-Act" loop.However, Fang emphasized that AI doesn't mean dumping all raw vehicle data directly into large models. The vehicle first needs flexible data acquisition and processing capabilities to extract valuable information from the sea of raw data before handing it to AI for further analysis. At the same time, in scenarios like vehicle diagnostics and safety, AI output must maintain sufficient determinism."In the end, only two things truly matter: time and risk control."Fang believes that once information across R&D, testing, production, and after-sales is truly connected by AI, the value AI brings will no longer be limited to the in-car experience, but will extend to development efficiency, quality control, and total lifecycle cost management.
+5 others
![Continuing post-Labor Day Green Deals hub - EVs, power stations, tools, smart devices, more [Updated]](https://cdn.topcarnews.net/thumb/id/aHR0cHM6Ly9pMC53cC5jb20vZWxlY3Ry-ZWsuY28vd3AtY29udGVudC91cGxvYWRz-L3NpdGVzLzMvMjAyNi8wOS9MZWN0cmlj-LUVjb0Zsb3ctQW5rZXItZXVmeS1TZWd3-YXktTmF2aW1vdy1MYWJvci1EYXktSHVi-LUZJLmpwZz93PTE1MDAmYW1wO3F1YWxp-dHk9ODImYW1wO3N0cmlwPWFsbCZhbXA7-c3NsPTE/16c16d62c32c03f486c198130cd2a06e.jpg?t=20260918&post_id=61223)
![[Updated] Continuing Labor Day Green Deals hub - EVs, power stations, tools, smart devices, more](https://cdn.topcarnews.net/thumb/id/aHR0cHM6Ly9pMC53cC5jb20vZWxlY3Ry-ZWsuY28vd3AtY29udGVudC91cGxvYWRz-L3NpdGVzLzMvMjAyNi8wOS9MZWN0cmlj-LUVjb0Zsb3ctQW5rZXItZXVmeS1TZWd3-YXktTmF2aW1vdy1MYWJvci1EYXktSHVi-LUZJLmpwZz93PTE1MDAmYW1wO3F1YWxp-dHk9ODImYW1wO3N0cmlwPWFsbCZhbXA7-c3NsPTE/16c16d62c32c03f486c198130cd2a06e.jpg?t=20260918&post_id=60924)
