Neolix Autonomous Vehicles Showcase Cutting-Edge Embodied AI At 2025 Yangtze River Delta Competition

(Source: China Automotive News, August 12, 2025)

Kunshan, China – August 12, 2025 – The 2025 Yangtze River Delta (Kunshan) Embodied Intelligence Scenario Application Competition, a landmark event spotlighting the embodied AI industry in China, kicked off this week in Kunshan, Jiangsu Province. Among the key participants in the smart, connected, new energy vehicle (SC-NEV) field, Neolix stood out. Their autonomous vehicles demonstrated advanced L4 self-driving technology and proven commercial deployment capabilities, successfully navigating complex professional test scenarios at the Maxxis Proving Ground. This performance underscored the leading capabilities of Chinese companies in the rapidly evolving embodied AI sector.

Neolix's competition vehicles are powered by a sophisticated L4 autonomous driving system. This system integrates perception, decision-making, and execution into a cohesive intelligent framework. Hardware-wise, the vehicles are equipped with a robust suite of sensors: 12 high-definition cameras, 1 LiDAR unit, and the powerful Orin-X automotive-grade computing chip. This combination delivers an impressive 254 TOPS of AI processing power and enables 360-degree environmental awareness within a 120-meter radius.

Neolix employs a multi-sensor fusion approach, significantly enhancing object recognition accuracy even in challenging weather conditions like heavy rain and dense fog. Crucially, their "cloud-edge-device" automated data loop system allows for continuous, real-time optimization of driving algorithms, ensuring the system learns and improves constantly.

The company also leverages a dual-redundancy architecture: combining robust "single-vehicle intelligence" with Vehicle-to-Everything (V2X) communication. By exchanging real-time data with the cloud – such as traffic light phasing and road construction updates – Neolix vehicles can proactively plan the most efficient routes. This enables nuanced controls like smoothly decelerating for upcoming red lights and accelerating through greens. This cloud integration effectively addresses decision-making gaps inherent in purely vehicle-based systems in dynamic environments, substantially boosting response speeds.

During the competition, Neolix focused intensely on the unmanned logistics scenarios. Their vehicles adeptly handled complex, simulated factory and campus environments, performing tasks like autonomous mapping, dynamic route planning, and precise obstacle avoidance. The Orin-X computing platform, coupled with their proprietary "4D One Model" perception algorithm, underpinned the reliable 360-degree, 120-meter environmental awareness. Impressively, dynamic route replanning occurred within one second, meeting the demanding efficiency requirements of real-world logistics operations.

In comprehensive scenario testing, Neolix vehicles confronted unexpected situations like pedestrians jaywalking and active construction zones. Their integrated "vehicle-road-cloud" data system proved vital here. By instantly accessing feeds from roadside cameras and high-definition map data, the "smart vehicles" and the "intelligent road infrastructure" worked in concert. For instance, when faced with sudden road changes, a vehicle could receive updated instructions via the 5G-Advanced network, swiftly replan its route, and maintain compliant driving – showcasing emergency response capabilities far exceeding those of human drivers.

Neolix's competition participation wasn't just a technology showcase; it highlighted significant shifts within the smart, connected vehicle industry.

• Technical Evolution: The synergy between their "4D One Model" perception system and V2X technology offers the industry a compelling path towards lower-cost, higher-efficiency algorithm optimization. Notably, Neolix has achieved substantial reductions in the hardware costs of their vehicles compared to similar-level competitors. This integrated hardware-software innovation presents a replicable model for the viable commercial deployment of L4 autonomy.
• Ecosystem Strength: The Kunshan competition leveraged the city's mature SC-NEV ecosystem, encompassing dedicated testing facilities, supportive policies, and a comprehensive supply chain. This created a powerful "R&D - Validation - Commercialization" loop. Neolix's collaborations with local partners on critical components like automotive-grade chips and by-wire chassis further strengthened the Yangtze River Delta region's cluster advantage in this sector. As Zhang Mengheng, General Manager of Kunshan High-Tech Group, emphasized: "The vitality of technology lies in its application. Kunshan is accelerating the journey of embodied AI from the lab to the industrial front line through scenario-driven initiatives."
• Transformative Value: The large-scale application of Neolix vehicles in logistics delivery and urban services promises to reshape traditional transportation models. Take their innovative X3 model as an example. Departing from fully enclosed designs, the X3 features a unique "back cage" concept – a detachable container integrated into the vehicle. This design streamlines loading/unloading processes, drastically cuts labor requirements, and significantly boosts delivery efficiency. It represents a practical, innovative solution tackling the persistent challenge of "last-mile" delivery.

The successful 2025 Yangtze River Delta (Kunshan) Embodied Intelligence Scenario Application Competition marks China's entry into a new era of scenario-driven innovation for smart, connected vehicles. Through this platform, Neolix convincingly demonstrated the reliability and scalability of its L4 technology in demanding real-world simulations. Their breakthroughs in the "cloud-edge-device" data system, the dual-redundancy V2X architecture, and cost-effective hardware integration provide clear, valuable directions for the entire industry's technological advancement. Neolix exemplifies how embodied AI is moving beyond concept into tangible, commercially viable solutions.