Chongqing - Open-source collaboration is emerging as a key solution to long-standing challenges in China’s intelligent vehicle industry, from weak ecosystem coordination to repeated development costs. At a recent industry gathering in Chongqing, automakers, chipmakers, and researchers discussed how shared platforms could accelerate innovation in basic software and automotive chips.

Open source means making software or hardware code publicly available so anyone can use, adapt, and share it under a license. Advocates say it lowers barriers to innovation and encourages wider participation.

Government, industry, and academia representatives gathered at the 2025 Intelligent Vehicle Basic Software Ecosystem Conference & the 4th China Automotive Chip Conference. (Photo/2025 Intelligent Vehicle Basic Software Ecosystem Conference & the 4th China Automotive Chip Conference)

"Today the global auto industry is undergoing profound change with electrification, intelligence, and connectivity," said Fu Bingfeng, executive vice president of the China Association of Automobile Manufacturers (CAAM). "Basic software and automotive chips are the foundation of this shift. Yet the industry still faces weak coordination, the need for core technology breakthroughs, and untested business models."

Fu emphasized the importance of joint innovation across the supply chain.

"Automotive software is now too complex for one company to handle," said Ren Wei, deputy chief engineer at Dongfeng Motor Corporation's R&D Institute. He argued that open-source models allow carmakers to access a broader pool of suppliers, helping reduce costs and improve efficiency.

Despite broad consensus on its potential, speakers acknowledged that building an open-source ecosystem remains difficult.

According to Liang Hao, director at iSOFT Infrastructure Software Strategic Research Institute, the industry faces two major obstacles: vehicle operating systems still struggle to adapt widely to different chips, and no large-scale engineering service system supports open-source implementation.

Li Yan, head of the basic software group at FAW's R&D Institute, said high software complexity, diverse chip types, and long validation cycles make it hard for companies to balance generalization with cost reduction. The absence of unified standards for interfaces, safety, and testing also drives redundant development and higher costs.

Chip development adds further complexity. "From design to mass production, automotive chips require about two years," said Li Wenxiong, vice president of GigaDevice Semiconductor. "But in that period, software may have iterated many times, making the chip incompatible if there was no early open-source collaboration."

Ren Wei noted another issue: most current initiatives stop at releasing code, without transparent processes or open governance, which undermines industry trust.

Looking ahead, Li Keqiang, the Chinese Academy of Engineering academician, said stronger basic research and innovation on key components are essential for intelligent connected vehicles. He called for coordinated mechanisms, development roadmaps, and alignment with both national and international standards.

Li Keqiang, academician of the Chinese Academy of Engineering. (Photo/2025 Intelligent Vehicle Basic Software Ecosystem Conference & The 4th China Automotive Chip Conference)

FAW's Li Yan suggested creating a full-chain mechanism from demand analysis to mass production, with unified standards covering interfaces, communication protocols, and safety requirements. He added that cross-sector cooperation could shorten validation cycles and bring open-source software into mass-market vehicles more quickly.

Chipmakers also see opportunities. "Chip companies should work closely with software developers to build stable and differentiated products," said Xie Wenlu, chairman of Xiaohua Semiconductor.

Several initiatives are already underway. Since 2020, the OpenAtom Foundation has supported open-source projects in operating systems and artificial intelligence. In 2023, CAAM introduced an open-source plan for automotive operating systems.

In 2024, OpenAtom Foundation, CAAM, and major automakers, including FAW and Changan, launched the China Automotive Operating System Open-Source Co-Building Plan, under which China’s first mass-production open-source vehicle control system, Xiaoman, was released.

At this year’s Chongqing conference, iSOFT announced the "Xinhui Plan," the first open-source vehicle operating system project aimed at mass production. Built on Xiaoman, it invites automakers, chipmakers, software firms, testers, and researchers to collaborate on chip compatibility, engineering services, certification, applications, and talent training to create an open and sustainable ecosystem.