PIA Automation Releases Embodied Intelligence Industrial Validation Roadmap and Data Flywheel

Product / Tech
โดย 证券日报·CN·Read original
Summary · why it matters

PIA Automation, marking 2026 as the first year of embodied intelligence delivery, has released an industrial validation roadmap covering high-end manufacturing, retail, and general services, along with the Primus Ego head-mounted data collection device and the Primus Forge data platform. Leveraging the group's global network of one hundred factories, the company has built a complete closed loop from single-machine capability testing to production-line-scale deployment. In a joint research lab with Zhiyuan, it achieved a 99 percent success rate in flexible assembly for retractor housing loading, with a fastest cycle time of 12.97 seconds, and reached a proof-of-concept collaboration with an overseas electric vehicle giant on steering mechanism assembly. At the 2026 World Artificial Intelligence Conference, the company showcased the world's first pilot production line for BMS controllers featuring six self-developed G2 series embodied robots working collaboratively. After more than ten thousand hours of real-machine testing, the line achieved a comprehensive stability rate of 99.9 percent, with single-process cycle times kept within 70 seconds. On the data side, Primus Ego captures workshop process data from a 270-degree first-person perspective, while Primus Forge uses a VLM large model to replace 80 percent of manual annotation. The company has also partnered with Boden Intelligence and the MINT Lab at Shanghai Jiao Tong University to open-source the RW-RL-Dataset, with an initial batch exceeding one thousand hours and plans to expand it to three thousand hours by the end of 2026.

Impact on stocks 1

Others · 1 stocks

Theme Impact 3

Off-coverage companies 1

智元机器人Private▲ Positive
Technologyrelevance

Zhiyuan collaborates with PIA in joint lab achieving 99% success in flexible assembly.

Related news

2

Caterpillar Expands Autonomous Hauling to Two More Virginia Quarries

Luck Stone announced in mid-September 2026 that it had expanded its collaboration with Caterpillar to roll out autonomous hauling technology to two additional Virginia quarries, building on a site where autonomous Cat trucks have already moved more than 3.50 billion tons without reported injuries. The expansion includes the first-ever deployment of Caterpillar's autonomous haulage on Cat 775 trucks, and the company is pairing the automation with workforce skill development to address quarry safety and productivity challenges. The move reinforces Caterpillar's broader push into autonomy and AI, which analysts tie to higher quality recurring revenue, and follows the company's August update highlighting record backlog and heavy investment in digital and automation. Caterpillar's narrative projects $94.5 billion revenue and $17.4 billion earnings by 2029, with a $970.37 fair value implying 20% upside, while some of the most optimistic analysts already assumed revenues above US$112,200,000,000 and earnings near US$20,700,000,000 by 2029. Investors are still weighing rising tariffs and pricing pressure against the pace at which digital and service income can scale.
Simply Wall St·13hRead more →

Citrini Research Finds Humanoid Robots Advancing Without a Single ChatGPT Moment

Humanoid robots are making meaningful progress in autonomy, dexterity and repetitive real-world tasks, but widespread commercial deployment remains a work in progress, according to a report published Wednesday by Citrini Research. After a week meeting robotics labs, engineers and investors across the San Francisco Bay Area, the firm found most humanoid robots remain in development, training and tele-operated demonstrations, with real-world deployments closer to pilots than clear return-on-investment opportunities. The report examined whether robotics is approaching its own ChatGPT moment and highlighted advances from GeneralistAI and Skild AI, along with Figure AI's demonstration involving 250,000 packages, suggesting the field may advance through a mosaic of signals rather than one defining breakthrough. Basic humanoid locomotion is effectively a solved problem while complex dexterity remains more difficult, and simple repetitive tasks are already trainable. Investment in the broader robotics infrastructure is keeping pace: Tesla Inc has been preparing production lines for Optimus, XPeng Inc has launched an automated production line for its IRON humanoid robot, and Figure AI is securing access to up to 100,000 Vera Rubin GPUs from NVIDIA Corp to train the models powering its humanoid robots. The report concluded robotic autonomy is developing along a curve rather than arriving through one decisive breakthrough, with broad generalizability remaining the longer-term goal.
Yahoo Finance·19hRead more →

Spirit AI aims to deliver humanoid robots that respond to verbal commands as early as next year

Gao Yang, co-founder and chief scientist of Chinese startup Spirit AI, told Reuters that humanoid robots capable of handling most basic tasks on verbal command could become a reality as early as next year. "By mid-2027, you will be able to speak to a robot in natural language and it will complete a series of actions to carry out a task," he said, expressing confidence in the pace of software advances. The company has already achieved a roughly 90 percent success rate on simple tasks, but challenges remain with fine movements such as loosening a bottle cap and with handling unexpected situations, and adoption in ordinary households is still expected to take time. About 1,000 contract workers at the company wear data-collection devices at home and in factories, gathering movement data to train the robots. Spirit AI was founded in 2024 and has about 300 full-time employees. It has raised more than 670 million dollars to date and has deployed dozens of humanoid robots at battery giant CATL and online retail giant JD.com.
ロイター·22hRead more →