On January 7, 2025, at the CES 2025, Professor Amnon Shashua, CEO of Mobileye, delivered the keynote address titled "Now, Next, and Beyond." This presentation showcased the company's latest advancements and future roadmap in autonomous driving technology.

This article will provide an in-depth look at Mobileye's two fundamental dimensions for autonomous driving and its five cornerstone strategies for product commercialization, offering readers a comprehensive understanding of Mobileye's technology and strategy.

Two Key Dimensions for Autonomous Driving

In his speech, Shashua highlighted the origins of Mobileye's product solutions and emphasized the two crucial dimensions necessary for achieving full autonomous driving:

Precision (Safety): Ensuring a sufficiently high Mean-Time-Between-Failures (MTBF) for hands-off or driverless operation.

• Objective: To ensure driving safety by minimizing errors such as false positives or false negatives.
• Key Indicator: MTBF
• Specific Requirement: Achieving an MTBF that supports hands-off or driverless driving.

Recall (Availability): Ensuring the system can handle a wide range of scenarios, geographical environments, and conditions, covering as many driving situations as possible.

• Operational Design Domain (ODD): The conditions under which an autonomous vehicle can operate; a broader ODD leads to higher recall.
• Geographic Scalability: Out-of-the-box performance in new locations.
• Cost: Lower costs for the system and its maintenance.

These two dimensions represent the cornerstone of Mobileye's product technology, prioritizing safety while ensuring the implementability, scalability, and availability of its products. However, for the Chinese market, there may be a need to infuse more appeal and exciting elements into the offerings.

Five Cornerstones for Revolutionizing Transportation's Strategic Framework

Mobileye's five key cornerstones are its product solutions, designed to convince B-end customers to adopt their offerings and boost shareholder confidence.

Safety: To transition from human-supervised driving to full autonomy, Mobileye employs four key mechanisms:

• RSS (Responsibility-Sensitive Safety) Regulation Algorithm: Ensures "zero" unreasonable regulations.
• Fail-safe, MRM (Maximize Failure Detection and Safe Stop on the Shoulder) Safeguard Mechanism: Maximizes fault detection and safe stopping through built-in redundancy.
• Transparency: Eliminates unreasonableness, distinguishing between "reasonable" and "unreasonable" risks.
• PGF (Primary-Guardian-Fallback) Perception Fusion Redundancy:
  • Primary: Provides preliminary prediction results, making initial judgments and analyses of the vehicle's driving state and surroundings.
  • Guardian: Conducts comprehensive verification of the primary system's results to ensure decision accuracy.
  • Fallback: Provides decision-making solutions when the primary system fails, ensuring system reliability and safety.

Technology Stack: Ensures reliable and efficient autonomous operation through:

• CAIS Architecture: A centralized AI system for processing and decision-making.
• EyeQ6H: A high-performance chip for real-time computing.
• Imaging Radar: Complements cameras and LiDAR to enhance perception capabilities.

Scalability: Achieving geographic and ODD expansion through:

• REM (Road Experience Management): Utilizes crowdsourced mapping to enhance the system's understanding and adaptability to different environments, suitable for global expansion.

Productization/Execution: Transitioning from demos to mass production by:

• Ensuring reliability, regulatory compliance, and integration into various vehicle models.
• Collaborating with over 50 OEMs to leverage its strong position in the ADAS market.

Cost: Balancing autonomy and affordability through:

• Efficient Hardware Design: Reducing system costs through optimized hardware.
• Scalable Software Solutions: Reducing maintenance costs through software scalability.
• Economies of Scale: Reducing costs through large-scale production and deployment.

Regarding pricing, Mobileye's offerings in China might seem less competitive, targeting more toward international investors. The L2 solution costs $700-800 (approximately RMB 5,000), competitive with domestic L2++ solutions, while their Supervision L2++ pricing is around RMB 15,000, similar to China's L2++ prices from a few years ago. However, their steady (stable and reliable) approach might still hold sway.

Summary

By leveraging these five cornerstones, Mobileye is committed to continuous improvement in precision and recall, aiming for scalable autonomous driving solutions. Through the PGF fusion method, CAIS architecture, EyeQ6H chip, and imaging radar technology, Mobileye ensures high system reliability and safety. Additionally, through REM technology and collaborations with OEMs, Mobileye achieves geographic and ODD expansion, ensuring broad applicability. Finally, through efficient hardware design and scalable software solutions, Mobileye maintains autonomy while keeping technology affordable. However, in the Chinese market, Mobileye's L2++ solutions may face competition from high-end providers like Huawei, Momenta, and Horizon Robotics, as well as lower-end solutions from Bosch and Renesas.

*Reproduction and excerpts are strictly prohibited without permission - References:

Mobileye CES 2025 Now Next Beyond.pdf and video Mobileye-CES2025 Speech Video