ZhiXing Tech’s Zhang Wei: Navigating Global ADAS Expansion Requires Tackling Diverse Scenarios—Mapless Solutions Aren’t the Only Way

At the 9th Intelligent Driving and Global Expansion Summit held on June 18, 2026, Zhang Wei, Senior System Director at ZhiXing Technology, highlighted that Chinese intelligent driving solutions face multifaceted localization challenges when expanding overseas, spanning driving and parking scenarios, perception systems, and human-machine interaction. Regarding on-road driving, European road environments differ significantly from those in China. For instance, certain German highways have no speed limits, placing higher demands on vehicle control algorithms and vehicle interface integration. Additionally, unconventional vehicles commonly seen abroad—such as trailers and cargo bikes—as well as ambiguous or temporary lane markings (e.g., gravel-based road markings or yellow-and-white construction lines) complicate perception model recognition. Moreover, varying traffic signs and speed limit regulations across countries necessitate continuous system optimization using localized data and map information. In parking scenarios, overseas parking spaces exhibit far greater diversity in size, boundary markings, and layout compared to the highly standardized spaces in China. Directly applying domestic perception models can lead to unstable recognition or even collision risks. To address this, ZhiXing has built a database covering thousands of parking scenarios across dozens of countries and leverages large-model generative technology to achieve scenario generalization with limited real-world data collection, thereby reducing overseas validation costs. Furthermore, to accommodate overseas users’ preference for “one-shot perpendicular parking,” the system must deliver higher precision in early parking-space release and trajectory planning. On human-machine interaction, developing multilingual HMI interfaces and adapting systems for both left- and right-hand drive vehicles are key priorities. These efforts require AI-assisted validation and strict compliance with data regulations such as GDPR. Regarding "map-free" solutions, Zhang clarified that current EU GSR-compliant map-free approaches specifically refer to systems that do not rely on map data and determine speed limits solely through visual perception. However, such systems may experience performance degradation in specific conditions like poor lighting. He emphasized that automakers can opt to integrate map data to enhance user experience, noting that map-free is not the only viable technical path. Zhang also revealed that ZhiXing has established an R&D center in Europe and a manufacturing facility in Malaysia, currently serving overseas OEMs such as Proton. Although overseas testing and validation entail significant investment, the data and experience accumulated during early-stage projects help reduce R&D costs for subsequent initiatives, positioning global expansion as a critical long-term profit driver for the company.