ZF Group and Horizon Robotics recently announced their partnership to deliver an advanced driver assistance system (ADAS) solution, supporting up to SAE Level 3 assisted driving and China’s urban Navigate on Autopilot (NOA), for production on a Chinese EV platform in 2026. The solution is powered by the ZF ProAI compute platform and Horizon’s Journey 6P hardware. Dr Christian Brenneke, head of electronics and ADAS at ZF, speaks to AAVI on the collaboration
How do ZF’s automotive-grade hardware and Horizon Robotics’ AI expertise complement each other in developing this Level 3 ADAS system, and what unique advantages does the partnership bring to the Chinese EV market?
ZF’s automotive-grade ProAI compute platform offers a scalable, modular architecture adaptable to all vehicle classes and types. It provides robust processing power and all-scenario adaptability. Horizon Robotics contributes its Journey 6P processing hardware, designed for high-performance perception and AI workloads, including vision language models and transformer-based networks. Together, this synergy enables end-to-end full-scenario driving assistance – from parking lot to parking lot – delivering a safer, more intelligent driving experience. The partnership combines ZF’s global OEM expertise with Horizon’s local innovation, tailored for China’s urban mobility needs.
How could the adoption of Level 3 ADAS influence safety standards and consumer trust in autonomous driving technologies?
The system’s support for SAE Level 3 and urban Navigate on Autopilot (NOA) sets a benchmark for intelligent driving. The system will feature full redundancy in the ADAS domain and the necessary mechanism for minimum risk management. By integrating advanced perception and decision-making capabilities, it enhances safety and reliability in complex traffic environments. This contributes to building consumer trust and may influence future regulatory standards for autonomous driving technologies.
How does the system leverage local data or maps to enhance the performance of urban Navigate on Autopilot in China’s complex traffic environments?
The system is specifically designed for China’s urban traffic scenarios. It leverages local maps and traffic data to optimize NOA performance, enabling precise navigation through multilane intersections, variable traffic rules and dense urban flows. This localization is key to delivering high-value, context-aware driving assistance
How does the system balance processing power (1,000+ TOPS) with energy efficiency for real-world vehicle deployment?
With over 1,000 TOPS of compute capability, the system balances high-performance AI processing with energy efficiency. The ProAI platform is engineered for real-world deployment in electric vehicles, ensuring optimal power consumption without compromising functionality. This makes it suitable for mass-market applications.
Can the system’s end-to-end full-scenario driving capabilities be adapted for international markets with different traffic rules and infrastructure?
The system’s modular architecture and AI-driven design allow for adaptation to international markets. While localization of traffic rules and infrastructure is required, the underlying technology is scalable and configurable. This supports ZF’s global strategy to make intelligent driving inclusive and accessible across diverse regions.
Explore: QNX’s Niko Boeker discusses AI and the future of SDVs in the UK