Zelostech, an innovative Chinese autonomous driving startup, is turbo-charging its overseas push with homegrown advanced Level 4 self-driving technologies, said its top executive in an interview with China Daily.

"We are striving to build an algorithm that can run not just across China but around the world. An algorithm — by deeply understanding and generalizing diverse traffic environments — can learn from Beijing traffic and still handle a rainy day in Singapore, or even a complicated road of Berlin," said Zhuang Li, co-founder of Zelostech.

He made the comments as Zelostech announced completion of nearly $300 million in Series B fundraising earlier this week. Following this round, Zelostech plans to focus on next-generation product development, establishing its own supply chain, expanding domestic and international markets, building urban operational networks, and fostering an industrial ecosystem alliance.

According to Zhuang, the Chinese company has already gained some momentum in overseas markets. In early 2024, it helped Singapore draft its first driverless logistics standards, filling a regulatory gap. It then passed the M1 test to secure Singapore's inaugural unmanned logistics vehicle license.

"The company has also rolled out its first autonomous delivery vehicle in South Korea, followed by deployments in Japan in November last year and Europe in January this year," he noted.

This month Zelostech marks 100 days since it began delivering its 2025 L4 urban delivery series. The new models ditch the mechanical lidar "hat" in favor of a sleek flattop design that hides solid-state lidar and high-resolution cameras.

According to Zhuang, the streamlined profile not only sharpens aesthetics, but also enables adjustable cargo box heights to tackle diverse delivery tasks, from last-mile e-commerce drops to heavy industrial transports.

Under the hood, the 2025 series pairs dual-camera 3D perception with its in-house Jupiter inference engine, achieving decision latencies below 50 milliseconds.

Such performance leaps are made possible by the startup's full-stack research and development, covering sensor fusion, artificial intelligence decision-making and edge computing. "This is the first large-scale rollout of a solid-state lidar solution on an L4 production vehicle," Zhuang said.

Zelostech's rapid commercial roll-out has translated into eye-popping growth. Last year, its monthly deployment rate surged above 45 percent for five straight months. Its footprint now spans six countries and more than 200 regions, with operating cities up 387 percent year-on-year.

Beyond city streets, Zelostech's vehicles are carving out niche roles in specialized zones. In Inner Mongolia autonomous region's grasslands, a fleet embarks on daily 100-kilometer loops between county hubs and remote villages, slashing rural delivery times and costs by roughly 40 percent. "These use cases highlight the versatility of a one-algorithm platform," said Zhuang.

With solid-state lidar lowering sensor costs and weight, the 2025 models are poised to drive down total-ownership costs, enabling wider adoption in smaller cities and secondary markets.

"We're not just selling vehicles," Zhuang added. "We're delivering a plug-and-play logistics solution that can be replicated anywhere."

According to market consultancy, the global autonomous vehicle market was valued at $207.38 billion last year and is forecast to skyrocket at a compound annual growth rate of 36.3 percent from 2025 to 2034.

As global logistics players seek high-efficiency, low-carbon delivery methods, Zelostech's homegrown technology and lightened vehicle architecture could serve as a template for scaling Level 4 autonomy.

"We believe the era of universal intelligent delivery is within reach," Zhuang said. "Our mission is to make advanced driving accessible, equitable and profitable for operators everywhere."