Automotive Intelligence

1. Definition and Core Connotation

Automotive intelligence refers to the technological evolution that equips vehicles with environmental perception, autonomous decision-making, human-machine interaction, and continuous learning capabilities through the integration of artificial intelligence, sensors, high-compute chips, and vehicle-to-everything (V2X) technologies, transforming them from traditional transport into mobile intelligent terminals. The industry has reached a broad consensus that "electrification is the first half, while intelligence is the second half", with intelligence emerging as the defining battleground for future automotive industry competitiveness.

2. Core Domains

Intelligent Driving

Intelligent driving leverages multi-sensor fusion—including millimetre-wave radars, LiDARs, and cameras—paired with end-to-end large model algorithms to achieve autonomous vehicle control from perception through to decision-making. In China, the new-car penetration rate for Level 2 (L2) driver-assistance systems surged from 3.3% in 2019 to nearly 60% in 2024, and is projected to approach 70% by the end of 2025. The technological approach is shifting from rule-based systems to data-driven, end-to-end large models, making perception fusion and end-to-end architecture the central focus of competition in autonomous driving.

Intelligent Cockpit

The intelligent cockpit integrates multi-modal interaction technologies such as voice recognition, gesture control, eye tracking, and emotion sensing, elevating the in-cabin experience from "passive response" to "proactive service". In 2024, the factory-fitted standard fitment rate of intelligent cockpits in China's passenger car market exceeded 70%, with projections to surpass 80% in 2025. With Large Language Models (LLMs) deployed on board, users can execute complex multi-step commands across navigation, infotainment, and vehicle settings via natural conversation.

Intelligent Chassis and Electronic/Electrical Architecture

Legacy distributed electronic/electrical (E/E) architectures are rapidly giving way to centralized "central computing + zonal control" systems. This consolidation unifies vehicle-wide control and data flow, laying the foundation for deep integration of AI models across the perception-decision-execution loop. Meanwhile, intelligent chassis technologies such as fully active suspension and steer-by-wire deliver a significant leap in dynamic handling and ride comfort.

3. AI-Defined Vehicles: The Paradigm Shift

The sector is progressing from "Software-Defined Vehicles" (SDVs) to "AI-Defined Vehicles" (AIDVs). While SDVs focus primarily on feature digitisation (such as OTA updates), AIDVs centre on cognitive intelligence—leveraging machine learning and closed-loop data pipelines to endow vehicles with autonomous decision-making and predictive situational awareness. AI has now penetrated in-cabin interactions, autonomous driving, chassis control, and smart manufacturing, establishing the motor vehicle as the ultimate physical platform and commercial endpoint for AI technology.

4. Development Challenges

In terms of core technological self-reliance, the domestic production rate of automotive-grade high-compute chips remains modest, driving algorithms still rely heavily on open-source frameworks, and high-end sensor components remain dependent on imports. Regarding data infrastructure, high-quality real-world edge-case data remains scarce, whilst fragmented data formats and interface standards across carmakers continue to create data silos. On the regulatory front, legal frameworks covering accident liability for Level 3 and above autonomous systems, data security, and privacy protection are still evolving, with a unified nationwide road-access policy yet to be enacted. Furthermore, the leap from L2 to L4/L5 continues to face key hurdles, notably corner-case handling and steep commercialisation costs.

5. Future Trends

Intelligent features are rapidly trickling down from high-end flagships across all market segments, with urban NOA (Navigate on Autopilot) functionality already entering the mainstream RM150,000–RM200,000 price bracket, paving the way for advanced driver assistance to become standard equipment across the board. The "Vehicle-Road-Cloud Integration" strategy is championed as China's proprietary blueprint for autonomous mobility, bridging the perception blind spots of standalone vehicle intelligence via roadside sensing and cloud-coordinated telemetry. Concurrently, technological synergies between autonomous driving and emerging frontier sectors, including eVTOL flying cars and embodied AI robotics, are gaining substantial momentum.