byd atto 3 from which country

The core differences between L4 and L5 autonomous driving lie in four aspects: applicable scenarios, hardware configuration, liability attribution, and technological maturity. L4 is classified as conditional automation, enabling autonomous driving only in preset closed or semi-closed areas (such as industrial parks and specific highway sections). It relies on lidar, high-precision maps, and multi-sensor fusion technology. Vehicles retain manual operation components like steering wheels, and human takeover is required when operating beyond the designated scope. Accident liability is usually borne by automakers. Currently, this technology has been commercialized in specific fields such as heavy-duty logistics trucks and shuttle buses. L5 achieves full automation, capable of handling all road environments and extreme weather conditions. It eliminates traditional driving controls and enables comprehensive operation through multi-spectral perception and human-like AI algorithms. Liability may shift to the system itself, but constrained by algorithm complexity and regulatory frameworks, it remains in the laboratory R&D phase. In short, L4 represents "specialized" autonomous driving, while L5 embodies "generalized" autonomous driving. Currently, L4 better aligns with practical needs, whereas L5 represents the future direction but requires both technological breakthroughs and legal infrastructure development.

Waymo vehicles are driverless taxis (Robotaxis) developed by Waymo, an autonomous driving company under Alphabet. Their core technology is based on the L4-level autonomous driving system, enabling fully driverless operation. Currently, Waymo has partnered with Geely Holding Group, and its Zeekr brand customizes exclusive vehicles based on the SEA-M (Sustainable Experience Architecture - Mobility) platform. This model removes traditional steering wheels and pedals, integrates sensors such as lidar on the roof, and is equipped with the Waymo Driver autonomous driving solution. These vehicles are primarily deployed on the Waymo One ride-hailing service platform, having commenced commercial operations in locations like Phoenix, USA, with plans to expand to cities including Miami. Notably, Waymo's vehicle design philosophy emphasizes shared mobility scenarios, featuring sliding doors and spacious interiors to facilitate passenger boarding and alighting. Although Waymo has not yet entered the Southeast Asian market, its technological approach aligns with regional competitors like Baidu's "Apollo Go," both employing high-precision sensor fusion and AI algorithms to navigate complex road conditions. Future global expansion may be considered as the technology matures.

The main challenges facing the local development of autonomous vehicles are concentrated in the ambiguity of the legal framework and liability attribution. The existing *Road Transport Act* has not yet clearly defined the division of responsibilities among manufacturers, software suppliers, or vehicle owners in the event of an accident, making it difficult for insurance claims mechanisms to adapt. Technologically, while systems like Tesla's FSD have adapted to complex road conditions through visual algorithms, the local variable traffic environment—such as dense motorbike traffic—still poses challenges to the real-time judgment capabilities of sensors. In terms of infrastructure, insufficient coverage of charging networks and the lack of high-precision map data limit the reliability of autonomous driving systems. Furthermore, current regulations conflict with some provisions of the Vienna Convention; for example, the requirement that drivers maintain full control of the vehicle directly hinders the commercial deployment of L4-level autonomous driving. Although the government has initiated the law revision process and established the Bukit Jalil test section, the improvement of the regulatory system still needs to advance in tandem with technological progress, and large-scale road application is not expected to be realized in the short term. Notably, local enterprises like REKA, which has developed the CRETA add-on system, demonstrate the potential of transitional solutions. Such retrofitting schemes can provide limited autonomous driving functions within the existing legal framework.

Autonomous driving refers to the technology that enables vehicles to operate independently through advanced sensors, artificial intelligence algorithms, and control systems, without requiring continuous human intervention. Currently, Malaysia has achieved breakthrough progress in this field. For instance, the L4 autonomous logistics vehicle jointly developed by ALS and Zelos in 2025 can perform complex tasks like cargo transportation in predefined environments, signifying the local logistics industry's shift toward intelligent transformation. Technically, L4 autonomy represents high automation where vehicles can manage most road conditions, though regulatory challenges persist—such as undefined accident liability frameworks. The government is addressing this by collecting data through pilot programs (e.g., Kuala Lumpur's 12-kilometer dedicated route) to refine policy frameworks. While full public road deployment remains distant in the near term, autonomous systems have demonstrated efficiency gains in confined settings like warehouses and ports. A case in point is the collaboration between 9Sight Intelligence and Pos Malaysia, which targets last-mile logistics solutions. Moving forward, as technology advances and standards mature, autonomous driving will progressively integrate into Malaysia's transport ecosystem, driving green logistics and smart city development.

L2 autonomous driving refers to a partially automated driving system where the vehicle can simultaneously control steering and acceleration/braking under specific conditions, but the driver still needs to remain attentive and be ready to take over at any time. Such systems use sensors like cameras and radar to monitor the road environment in real time, enabling the coordinated operation of functions such as adaptive cruise control (automatically adjusting vehicle speed to maintain distance from the preceding vehicle) and lane centering assist (automatically fine-tuning the steering wheel to keep the vehicle within the lane). For example, when driving on highways, the vehicle can automatically follow the preceding car and stay in the center of the lane, but the driver must continuously observe the road conditions and cannot keep their hands off the steering wheel for an extended period. Currently, mainstream models in the local market such as Proton X90 and Perodua Ativa are equipped with L2 systems, with prices usually ranging from RM80,000 to RM150,000. It should be noted that L2 systems still fall under the category of driving assistance, and the driver is ultimately responsible for driving safety under any circumstances, which is fundamentally different from higher-level conditional autonomous driving (L3). With technological development, some manufacturers have begun to gradually enhance the scenario adaptation capabilities of L2 systems through OTA updates, such as adding automatic lane changing or traffic light recognition functions.