(Yicai) Aug. 14 -- Factories in China still encounter technical challenges when adopting androids despite many domestic robot makers seeing a surge in orders this year, with a balance between improving plants' operational efficiency and enhancing robots' capabilities still needing to be struck, according to industry insiders.

Although overall orders in the robotics sector have surged this year from previous years, robots have not necessarily successfully begun working in factories, a source from the embodied intelligence industry told Yicai. Even if they did, it does not mean developers can obtain valuable data to train and upgrade embodied brain models, the person added.

Agibot, Unitree Robotics, and UBTech Robotics lead by the number of orders received so far this year, including Agibot securing a CNY78 million  (USD10.9 million) order from a unit of mobile network operator China Mobile at the end of June, Unitree receiving a CNY46.1 million order from the same firm, and UBTech winning a CNY90.5 million project from a car exporter last month.

Contact-based operations like bolt screwing are still difficult for industrial robots, with many technical problems needing to be overcome, noted Gao Feng, a professor at Shanghai Jiao Tong University. Work mistakes by robots, quality issues, and other accidents are likely to occur if the issues are not resolved, Gao pointed out.

UBTech expects to deliver about 500 industrial androids this year, mainly used in handling, sorting, and quality inspection, Vice President Jiao Jichao said at the World Robot Conference. The Shenzhen-based firm counts auto, 3C (computers, communications, and consumer), and semiconductor companies among its clients, he added.

Although embodied intelligence tech, including large language models and multimodal vision language models, is rapidly developing and vision-language-action models are being explored, almost all robots carry only the simplest operations currently, Liu Yunhui, a professor at the Chinese University of Hong Kong's engineering faculty, said to Yicai. For example, precise grasping is often difficult to achieve, Liu noted.

These challenges cannot prevent this from being "the first year of robot mass production," according to the insiders. However, robot makers and their clients must find an appropriate balance between improving efficiency and robot performance, they added.

Clients hope robots can help production lines complete more complex tasks without modifying their infrastructure, Xie Tiandi, marketing director of RoboSense Technology, told Yicai, adding that the embodied intelligence sector and corporate customers must seek a balance between improving factory operational efficiency and robot investment costs.

For example, adjusting the degrees of freedom of a robot's dexterous hand to an appropriate level that allows it to complete basic tasks in factories is sufficient for now, Xie pointed out.

More and more manufacturers are no longer solely pursuing the technical capabilities of robots but are focusing on their cost-effectiveness, Ma Yang, general manager of Tashi Technology, said to Yicai. They stopped having excessive requirements for the dexterous hand's freedom because they know that higher degrees demand more computing power, Ma noted.

Editors: Tang Shihua, Martin Kadiev