Gasgoo Daily brings you important automotive news from China every day. For those we report on, the article title contains a hyperlink with more information.
Pony.ai and ATBB partner on Robotaxi services for premium business travel scenarios
Who: Pony.ai (self-driving company); ATBB (mobility service provider).
what: The two companies have entered into a strategic partnership to jointly develop and operate a fully driverless robotaxi fleet targeting premium business and aviation-related travel scenarios.
when: Announced on January 28th.
where: Initially introduced in first-class cities in China, focusing on airports, train stations, and urban centers.
how:
The fleet is based on Pony.ai’s 7th generation robotaxis model.
Initial vehicles have received multiple regulatory approvals, including Level 4 driverless testing and pilot permission, enabling public ride-hailing services.
The vehicles will be integrated into Pony.ai’s own ride-hailing platform and selected third-party platforms.
role:
Pony.ai: Provides self-driving technology, vehicles, and operational networks using existing unmanned fleets in Beijing, Shanghai, Guangzhou, and Shenzhen.
ATBB: Contributes to premium mobility service capabilities through brands such as Xinghui Mobility and has a strong presence in the business and air travel markets.
Reason/strategic significance:
The partnership aims to expand Robotaxis services into the premium mobility segment, increasing service availability, reducing waiting times and supporting scenarios such as airport transfers and business trips.
Pony.ai builds on previous deployments in major transportation hubs such as Beijing Daxing International Airport, Guangzhou Baiyun International Airport, Guangzhou South Railway Station, and Shenzhen Bao’an International Airport.
WeRide launches GENESIS simulation platform to accelerate global autonomous driving development
Who: WeRide (self-driving company).
what: Launched GENESIS, a general-purpose autonomous driving simulation platform developed in-house.
when: Announced on January 28th.
Purpose/reason:
To address the limitations of traditional road tests in covering different regulations, road conditions, climates, and rare or extreme scenarios.
Accelerate the large-scale development and commercialization of autonomous driving systems.
How/Core Features:
It combines physical and generative AI to bridge real-world physics and virtual simulation.
Use generative AI to quickly create realistic urban environments and simulate long-tail cases such as extreme weather, traffic jams, and emergency situations.
Platform architecture——Built around a closed-loop system with four AI modules.
AI scenarios: Generate everyday risky driving situations based on billions of kilometers of driving data and over 8 years of real-world operations.
AI agents: Simulate realistic behavior of vehicles, pedestrians, and cyclists through everyday and high-risk interactions.
AI metrics: Quantify performance across safety, regulatory compliance, comfort, and efficiency.
AI diagnostics: Automate root cause analysis and generate optimization recommendations.
Scope and impact:
Supports vehicle platforms from ADAS to Level 4 autonomous driving.
Enables a “one platform, global training” approach and reduces market-specific redevelopment.
Compress millions of kilometers of physical testing into days of virtual simulation.
Company background: WeRide operates in more than 40 cities in 11 countries and has self-driving permits in eight markets.
Black Sesame Technologies Announces FAD 2.0 Open Platform
Who: Black Sesame Technologies (AI computing platform developer).
what: Launch of FAD 2.0 open platform for full scenario assisted driving.
when: Announced on January 28th
Regulatory background: The platform is built around the Huashan A2000 driver assistance chip, which has received regulatory clearance from the US Department of Commerce and the US Department of Defense. Black Sesame Technologies is currently the only Chinese company with such approval for driver assistance chips.
Platform configuration; integrates A2000 high-performance computing hardware, comprehensive SDK, AI toolchain, reference models for end-to-end and VLA architectures, and third-party algorithms. It is designed as a near-production-ready open platform, with a formal release scheduled for Q1 2026.
Technical ability: Supports both in-vehicle installation and prototype verification. Scalable coreboard architecture reduces integration costs and shortens development cycles. Supports over 24 camera channels and 4 10Gb Ethernet links. AI toolchains based on MLIR enable optimization of CNN, BEV, end-to-end, and VLA models.
Safety and Compliance: A2000 complies with ISO 26262 ASIL-D functional safety standards. Integrate hardware security modules for automotive-grade safety.
Demonstration and progress: The first model was deployed in 10 days. At CES 2026, Black Sesame Technologies exhibited a VLM and point-to-point (P2P) demonstration based on the FAD 2.0 platform.
Strategic significance: We position FAD 2.0 as a full-stack, open hardware and software platform to reduce development complexity and accelerate global deployment of advanced driver assistance systems.
Unauthorized reproduction is prohibited. Please do not reproduce, copy or use the edited content without permission. Contact: autonews@gasgoo.com
