Hello Launches HR1 Robotaxi, Sets Sights on 1 Million Robotaxis in Operation...

AsianFin -- Hello, one of China's leading mobility technology firms backed by Ant Group and CATL, has officially launched operations for its first mass-produced Robotaxi, the HR1.

Less than three months after announcing its entry into the Robotaxi sector, Hello unveiled the vehicle and laid out an ambitious roadmap to scale its fleet to over 50,000 vehicles globally by 2027, with mass production scheduled to begin in 2026.

The company’s rapid commercialization reflects the convergence of mature AI infrastructure and declining computing costs, according to Hello CEO Yang Lei.

Yang said the timing was deliberate, emphasizing that 2025 marks the beginning of commercial operations in four to five Chinese cities. “The AI and hardware environment has matured, and computing costs have decreased sufficiently to make this a viable business,” Yang noted. Over the next eight years, Hello aims to scale its Robotaxi fleet to one million vehicles on the platform.

Adding to the company’s technical leadership, former SAIC SAIKE Intelligent CTO Yu Qiankun has joined Hello Autonomous Driving as co-founder. Yu previously spearheaded L4 autonomous driving research and development at IM Motors and played a key role in commercial rollouts at SAIC’s wholly owned SAIKE Intelligent subsidiary, a critical pillar of the automaker’s “New Four Modernizations” strategy.

Hello initially considered an aggregation model, where it would integrate fleets from existing Robotaxi operators into a single platform. “We did consider this approach, but it comes with inherent problems,” Yang explained. Trust barriers between autonomous driving developers and operators often prevent full data sharing, limiting operational effectiveness.

Instead, Hello opted to pursue full-stack in-house development while collaborating with Horizon Robotics on L4-level vehicles. According to Yu, the HR1 departs from traditional modular Robotaxi architectures, which separate perception, prediction, planning, and control. Instead, Hello employs an end-to-end framework that maps multi-sensor inputs directly to vehicle control commands, improving system responsiveness and decision-making consistency.

Supporting this architecture, Hello has established a comprehensive technology foundation encompassing big data, high-performance computing, and large AI models. The company plans to create a dataset with over 10 million clips by 2026, developed through automated, compliant production lines. Partnerships with Alibaba Cloud enable Hello to establish a computing cluster comprising tens of thousands of GPUs to train its autonomous driving models. The company also collaborated with Alibaba Cloud’s Qwen team to develop the “Daoyu” large model with billions of parameters, supporting three core cloud-based capabilities: data mining, World Model simulation, and the VLA Cloud Driver.

The data mining model identifies over 500 types of long-tail scenarios for efficient data labeling. The World Model simulation model enables high-fidelity scenario testing, including collisions and extreme weather. The VLA Cloud Driver, with more than 50 billion parameters, supports distillation training for in-vehicle models and can issue extrication commands via a cloud cockpit system to replace safety operators when needed, laying the groundwork for fully unmanned operations.

Additionally, Hello is developing a “Cockpit AI Agent” using multimodal recognition, retrieval-augmented generation (RAG), and knowledge-base technologies to provide intelligent in-vehicle assistants, travel planning tools, and smart agent functionalities.

Hello’s first factory-installed Robotaxi, the HR1, is based on a Dongfeng Venucia model and features a comprehensive sensor suite: four medium- and long-range radars, four blind spot radars, six millimeter-wave radars, twelve ultrasonic radars, and multiple cameras, delivering full 360-degree multimodal coverage. The HR1 also incorporates an elastic wave sensor array and an in-cabin OMS camera to monitor structural integrity and passenger behavior in real time.

For computing, the HR1 employs dual automotive-grade domain controllers with a combined 2,200 TOPS of processing power. Its positioning system uses dual combined inertial navigation for precise and stable vehicle localization.

Yang Lei emphasized that Hello will not adopt a heavy-asset operational model for its Robotaxi fleet. He outlined potential arrangements, including joint ownership with local government investment entities or state-owned taxi platforms, and asset securitization through bonds to manage financial pressure as fleet size grows. According to Yang, by 2026, Robotaxi platforms with fewer than 10,000 vehicles will struggle to compete, highlighting the importance of scale and operational efficiency.

Yang stressed that as autonomous driving technology matures, the true competitive edge will shift from vehicle hardware to operational capabilities. “Optimizing operational efficiency in even the smallest details can deliver significant cost advantages,” he said. Hello’s extensive offline operations network spans more than 500 cities in China, with dedicated teams managing local business efficiency and maintaining communication with transportation authorities and urban management departments.

Hello’s experience in the shared mobility sector has provided a strong foundation for Robotaxi operations, giving it an edge over competitors in fleet deployment, local operations management, and government liaison.

With its full-stack development approach, robust AI infrastructure, and an operational strategy built on scale and efficiency, Hello is positioning itself as a major contender in China’s rapidly evolving autonomous mobility market. The company’s ambitious plan to deploy over 50,000 Robotaxis by 2027—and eventually scale to one million vehicles—signals its intention to set the pace for the country’s emerging Robotaxi industry.

By combining in-house technology development, strategic partnerships, and operational excellence, Hello is betting that its integrated approach will allow it to outpace competitors and establish a dominant presence in both Chinese and potentially global autonomous vehicle markets.