Zeekr Technology Europe and Intellias are demonstrating an AI-driven quality assurance system for in-vehicle infotainment at CES 2026. The solution combines vision-language models with computer vision to automate UI testing across navigation, media, connectivity, and comfort features.
The technology generates test cases directly from written requirements and UX design guidelines, then executes automated validation without manual scripting. According to the companies, the system interprets screens similarly to human testers while detecting layout issues, validating user flows, and monitoring performance in real time.
Highlights
- AI visual testing achieves 4x faster validation cycles than manual testing with 95%+ accuracy across thousands of tests
- Vision-language models automate end-to-end UI testing for in-vehicle infotainment, eliminating manual scripting requirements
- System adapts to environment changes autonomously, handling variations in navigation routes and dynamic content without script updates
- Live demonstration at CES 2026 features direct screen capture and external camera-based recognition scenarios
Technical Capabilities and Performance Metrics
The AI-driven approach addresses limitations inherent in manual quality assurance processes. Traditional IVI testing requires extensive scripting and human oversight, creating bottlenecks in software development cycles.
The automated system delivers several measurable improvements over conventional methods. Testing cycles complete up to four times faster than manual validation, enabling earlier feedback loops during development. The platform maintains accuracy above 95% across thousands of test iterations without the quality degradation that affects repetitive human testing.
“This innovation changes how we approach EV software development,” said Christian Hering, Head of Intelligent Software at Zeekr Technology Europe. “By leveraging the power of vision-language models and computer vision, we are transforming quality assurance: what once required hundreds of hours of manual effort is now automated with even greater precision.”
CI/CD Integration and Adaptive Testing
The solution operates autonomously within standard CI/CD workflows. When environment changes occur, the AI handles minor variations without requiring script updates—a significant advantage for dynamic testing scenarios.
Navigation testing illustrates this flexibility. The system manages dynamic route changes while validating route-specific properties including charging stations, points of interest, and navigation data. This occurs without modifications to underlying test logic.
“We believe that this AI-powered quality assurance solution for visual testing will help Zeekr accelerate development, strengthen product quality, and deliver exceptional digital cockpit experience to its drivers,” said Juergen Friese, SVP Mobility Business at Intellias.
CES 2026 Demonstration Details
Attendees can view the technology at Intellias Meeting Space #3377 in the LVCC West Hall, Vehicle Tech & Advanced Mobility section. The demonstration showcases two primary testing approaches: direct screen capture and external camera-based recognition.
Both scenarios illustrate how AI visually analyzes and validates IVI performance in real-world conditions.
Company Background
Intellias provides AI-enabled product engineering and digital solutions, supporting clients including HERE Technologies, Zeekr, Elmos Semiconductor, DKV Mobility, TomTom, and Renesas across more than 20 years of operation.
Zeekr Technology Europe, headquartered in Gothenburg, Sweden, was established in 2013 as CEVT. The company developed the Compact Modular Architecture platform, now deployed in over 4 million vehicles globally. More recently, the team created the SEA-M architecture designed specifically for shared autonomous applications. Zeekr and Waymo launched the first vehicle on this platform in 2022.
Sign up for our popular daily email to catch all the latest EV news!
