Motional Vision Unveils Next-Gen Integrated Driving & Parking Domain Controller Powered by AXERA AI M57 Chip

At the 2026 Beijing Auto Show, MOTRIS Intelligence officially launched its new-generation integrated driving and parking domain controller, built on the AXERA M57 series automotive SoC chips. Yu Zhenghua, Founder and Chairman of MOTRIS Intelligence, and Qiu Xiaoshen, Founder and Chairman of AXERA Intelligence, attended the launch ceremony. The domain controller features a lightweight, high-performance BEV-based driving and parking neural network, optimized computing resource allocation, and a highly versatile platform architecture. It enables seamless switching between driving and parking functions, operates automatically around the clock without manual setup, and delivers an out-of-the-box, universally compatible solution—aiming to make L2+ intelligent driving features standard across vehicle models. For driving functions, the product comes standard with ten core ADAS capabilities: Forward Collision Warning (FCW), Lane Departure Warning (LDW), Adaptive Cruise Control (ACC), Automatic Emergency Braking (AEB), Lane Keeping Assist (LKA), Emergency Lane Keeping (ELK), Integrated Cruise Assist (ICA), Traffic Sign Recognition (TSR), Lane Centering Control (LCC), and Traffic Jam Assist (TJA). The parking module supports fully automated parking into all common scenarios—including parallel, perpendicular, diagonal, nose-in, ultra-narrow spaces, wall-limited U-turns, and mechanical parking spots—and includes 11 practical features such as trace reversing and remote parking, covering both everyday and extreme use cases to lower user barriers. Developed entirely in-house by MOTRIS Intelligence, the solution spans the full stack—from hardware architecture, foundational software, middleware, 4D BEV perception algorithms, to planning and control modules. Combined with the AXERA M57 chip’s high effective computing power, low power consumption, and full automotive-grade safety certification, it achieves an optimal balance of performance, efficiency, and cost. The two companies have deeply collaborated throughout the entire process—from chip definition and algorithm adaptation to mass production—overcoming key technical challenges such as lightweight BEV deployment and dynamic compute allocation. The product’s universal architecture complies with both domestic and international regulations and road conditions, significantly shortening OEM development cycles and reducing integration costs. Moreover, the solution leverages an AI large-model-driven data closed-loop system that covers the entire workflow—from data governance and intelligent annotation to model training, simulation validation, and mass-production iteration—continuously enhancing performance in long-tail scenarios and elevating the standard intelligent driving experience.