Great Wall Motor’s She Shidong Outlines In-House ADAS Strategy: Phased Rollout, Gradual Integration, No Full Supplier Replacement

During the 2026 High-Level Forum on Intelligent Electric Vehicle Development, She Shidong, Deputy General Manager of Intelligent Products at Great Wall Motor, stated that vehicle intelligence is evolving along two major trends: First, diverse powertrain demands will become widespread—not limited solely to the rapid growth of new energy vehicles. Second, entire vehicles will evolve toward becoming "intelligent agents," with user-vehicle interaction achieved exclusively through natural conversation and proactive services. Great Wall has already launched driver intelligent agents and cabin intelligent agents and plans to further clarify its whole-vehicle intelligent-agent strategy within the next two to three years. The user adoption rate of its intelligent driving features has risen from single digits to over 30%. Regarding its in-house intelligent driving solutions, She explained that Great Wall employs a "layered in-house development + open co-creation" strategy: Flagship models will undergo deep in-house R&D, as end-to-end experience requires integrating actuators like steering and chassis into model training—a field still in its early industry stages. Mass-market models will fully adopt in-house solutions to ensure high-speed scenarios work well and basic scenarios remain functional. Mid-tier models will continue incorporating high-quality external solutions, complementing in-house efforts. Overall, in-house development does not aim to fully replace suppliers but rather to gradually penetrate the ecosystem. She also outlined Great Wall’s three-stage intelligent transformation journey: The first stage focused on platformization—launching Coffee OS 3.0 to unify software interfaces and operational logic across all brands, enabling 2–3 OTA updates per year on average. The second stage shifted toward deep user-centric operations, completing a consumer-facing (to C) transformation based on over 1,000 user interviews and more than 60 rounds of 48-hour field research. This resulted in OTA user satisfaction exceeding 90%, though it also revealed significant variations in personalized needs that product managers struggle to cover comprehensively. The third stage is now advancing the implementation of intelligent agents, leveraging AI to deliver truly personalized ("thousands of faces for thousands of users") services. However, the company still faces challenges due to insufficient industry partners and must independently accumulate data and build vertical-specific models.