Sign up for our weekly emails to stay up to date on the latest news!
Pony.ai, a frontrunner in autonomous driving technology, has reached a significant milestone with its Level 4 (L4) automotive-grade domain controller, surpassing 2 million kilometers of on-road testing in its seventh-generation Robotaxi fleet. This achievement underscores the system’s reliability, durability, and adaptability, meeting the stringent requirements of fully autonomous driving. Unveiled at the Shanghai Auto Show, this production-ready solution marks a leap forward in performance and cost-efficiency, positioning Pony.ai as a leader in the autonomous vehicle industry.
Key Highlights
- 2 million kilometers of on-road testing completed.
- 80% cost reduction in domain controller compared to previous generation.
- Utilizes four NVIDIA OrinX chips, delivering 1,016 TOPS of computing power.
- Features full-vehicle redundancy for enhanced safety.
- Supports liquid and passive cooling for operational reliability.
- Mass production underway, targeting 1,000 autonomous vehicles in 2025.
Advanced Technology for L4 Autonomy
Pony.ai’s seventh-generation autonomous driving system is built entirely with automotive-grade components, designed for a 10-year lifecycle or 600,000 kilometers. The in-house developed domain controller integrates three NVIDIA OrinX chips in the main unit, with a fourth dedicated to redundancy, ensuring robust computing power for L4 algorithms. This architecture achieves a 50% to 80% reduction in size, weight, power consumption, and cost, optimizing hardware expenses while maintaining high performance. The controller serves as the central hub, managing sensors, system power, interactive displays, hazard lights, gateways, and GNSS, streamlining vehicle design and reducing manufacturing costs.
Safety and Redundancy at the Core
The domain controller’s multi-layered safety architecture ensures “Fail Operational” capability, allowing seamless transitions to redundant systems or the Minimum Risk Condition Controller (MRCC) during failures. Even if the main system’s power or chassis communication fails, the redundant system maintains critical perception, including blind spot coverage, enabling safe navigation and pull-over. This safety-first approach, combined with Pony.ai’s PonyWorld foundation model and reinforcement learning, delivers driving performance that surpasses human capabilities, continuously improving in complex real-world scenarios.
Cost Efficiency and Scalability
Through advanced algorithm optimization and system engineering, Pony.ai has slashed the domain controller’s cost by 80% compared to its predecessor, while extending its lifespan. The integration of components reduces wiring complexity and potential failure points, simplifying assembly and enabling faster mass production. With liquid and passive cooling options, the system remains reliable even if the primary cooling system fails, ensuring safe operation. These advancements make Pony.ai’s solution both cost-effective and scalable, critical for large-scale Robotaxi deployment.
Mass Production and Future Outlook
Pony.ai’s seventh-generation Robotaxis are now in mass production, with plans to deploy 1,000 autonomous vehicles in 2025. This milestone marks the company’s first year of large-scale production, paving the way for economies of scale and enhanced user mobility. The robust, cost-efficient domain controller is set to expand Pony.ai’s driverless fleet, offering a safe and reliable solution for autonomous transportation.
For more details, visit Pony.ai’s official website.
Sign up for our weekly emails to stay up to date on the latest news!
