The QPilot MAX 500+ TOPS city navigation solution-built on a World Model and Reinforcement Learning framework-delivers best-in-class performance with an AEB false activation rate well below the industry average

BEIJING: Global autonomous driving leader QCraft held its opening event at the 2026 Beijing International Automotive Exhibition (also known as Auto China 2026) today, where Chairman and CEO Dr. James Yu officially unveiled the QCraft Physical AI Model and announced a full strategic expansion from autonomous driving to the broader frontier of Physical AI.

To a crowd of industry insiders, QCraft revealed for the first time the complete architecture of its Physical AI Model, built on a unified World Model + Reinforcement Learning (RL) framework.

The company also launched QPilot MAX, a 500+ TOPS (Trillion Operations Per Second) city NOA (Navigate on Autopilot) solution; disclosed the latest progress on its L4 Robotaxi and Robovan programs; and announced an upgraded mission and vision.

“If the past decade was about teaching AI to drive, the next decade will see the industry move decisively toward Physical AI-a frontier that is more imaginative, more disruptive, and far more consequential,” said Dr. Yu.

He explained that World Models and RL are the essential bridge between the digital and physical worlds, allowing QCraft to run infinite training cycles in a digital environment and transfer that capability to real-world vehicles.

“This is not a simple algorithm upgrade. It is a fundamental shift in how we approach R&D,” he said.

QCraft’s Physical AI Model operates across two layers. On the cloud side, an upgraded World Model generates rare edge cases-extreme weather, wrong-way cyclists, sudden pedestrian appearances-through natural language commands. On the onboard side, the World Behavior Model integrates a VLA (Vision-Language-Action) model with RL algorithms, achieving full-chain integration from perception to action.

Safety at Scale with QPilot MAX

Remarking on QCraft’s launch of QPilot MAX, a city NOA solution delivering best-in-class performance on a 500+ TOPS computing platform, Dr. Yu shared: “We don’t compete on parameters. We compete on experience.”

Working alongside China's largest OEM automotive manufacturer, QPilot MAX is deployed in 25 production models, with another 50 models expected in 2026. The solution’s Automatic Emergency Braking (AEB) false activation rate sits well below the industry average at one per 500,000 kilometers, helping users avoid approximately 146,000 potential accidents per year.

Dr. Yu pointed to insurance as the real-world measure of the system’s value: “If it is genuinely safer, shouldn’t users pay lower insurance premiums?”

On Robotaxi development, Dr. Yu outlined QCraft’s core thesis: a stronger AI brain, not more sensors: “Humans drive with significant blind spots, yet we do so safely because we have powerful brains. That is exactly QCraft’s approach.”

QCraft’s Robotaxi solution is based entirely on production-grade vehicle configurations, and the company prioritizes careful, steady scaling over speed.

On the logistics side, QCraft also showcased the QC-1 logistics robot, designed to solve the final 100 meters from vehicle to front door.

Dr. Yu closed by announcing QCraft’s new mission statement: Empower a Brighter Future with Safe and Beneficial AI and new vision: Pioneering the Global Frontier of Physical AI.

As QCraft unveiled its Physical AI vision in Beijing, the company’s vehicles were simultaneously undergoing road testing in Munich and Paris. From China to Europe, QCraft is carrying Physical AI and autonomous driving further than ever before.

About QCraft:

QCraft is a global leader in L2++ to L4 autonomous driving (AD) solutions for automakers. Founded in Silicon Valley in 2019, the company has deployed its technology with a range of leading global automakers. Powered by a world-class R&D team and partnerships with leading OEMs and tech companies, QCraft combines proven large-scale adoption with industry-leading safety and efficiency to bring autonomous driving into real life.

Fonte: Business Wire