Baotong leans on Unitree to build industrial robots, yet its R&D spend fell after going all-in on AI

On 29 August, Baotong Technology, a firm built on games and industrial internet, released its half-year report. With both businesses recovering, revenue and profit rose together, but the longer arc is uneven: gaming is volatile, and Baotong’s two main lines have little in common.

Baotong Mine-Rhino quadruped inspection robot
Baotong Mine-Rhino (Kuangxi) quadruped inspection robot built on a Unitree base. (Source: OFweek Robotics)

Two main businesses up, but still short of the peak

For the first half, Baotong posted revenue of RMB 1.966 billion, up 22.3 per cent, and net profit of RMB 192 million, up 37.4 per cent. Mobile games contributed RMB 857 million, up 22.09 per cent, driven by licensed overseas releases of Bloodline and The Cozy Florist. On the industrial side, Baotong signed a global framework deal with BHP, moving from selling product to product plus technology plus lifecycle service; its Thai base achieved mass production of ultra-high-strength conveyor belts, and Australia and Bangladesh awarded large mine integration and replacement projects.

Yet profit still trails the 2020 and 2021 peaks of RMB 230 million to RMB 240 million. The cause is structural: lower-margin industrial-internet revenue now weighs more, and game margin fell from 74.39 per cent to 66.17 per cent. As a licensing publisher, Baotong hits a natural ceiling and depends heavily on hits, so pushing revenue up means paying for user acquisition in a crowded field.

Partnering Unitree, going all-in on AI, yet R&D fell

Against slipping margins, Baotong is extending into AI and industrial robots. As early as 2021 it embedded AI defect detection in conveyor monitoring. In March 2025 it signed a strategic deal with Unitree: buying Unitree locomotion bases and adding its own secondary development, Baotong launched the Mine-Rhino (Kuangxi) quadruped inspection robot for high-risk industrial inspection. In 2025 it ran small pilots and batches but no scale revenue. In 2026 it published an AI-landing master plan, formally going all-in on AI across industrial AI, content AI and frontier physical AI.

Under industrial AI it deploys inspection, cleaning and conveyor-monitoring robots. In content AI it secured exclusive overseas operating rights to efun’s AniShort full-chain AIGC platform. In physical AI it formed a 70-per-cent-owned JV, Wuxi Baoguang Physical AI, focused on industrial simulation, synthetic data and robot evaluation, the AI base layer for industrial robots.

The puzzle: despite declaring an all-in pivot, Baotong’s first-half R&D was only RMB 59.48 million, down more than RMB 16 million from the RMB 76.19 million of the first half of 2024. By contrast, Kunlun Tech, also all-in on AI, spent RMB 789 million in the first half, more than double its 2023 level. Baotong has not broken out AI revenue separately, and the profit lift so far comes mostly from the legacy businesses.

Whether AI can carry Baotong’s next growth curve remains unproven. The Unitree-based Mine-Rhino is a real product with a real buyer profile in hazardous inspection, but a falling R&D line sits awkwardly next to an all-in slogan.

More images from the source report:

Baotong industrial AI robot lineup
Baotong’s industrial AI and inspection robot lineup. (Source: OFweek Robotics)
Baotong conveyor-belt AI monitoring system
Baotong smart conveyor-belt AI monitoring system. (Source: OFweek Robotics)
Baotong mine conveyor replacement project
Baotong integrated mine conveyor replacement project. (Source: OFweek Robotics)
Baotong physical AI joint venture
Baotong Wuxi Physical AI joint venture for robot evaluation. (Source: OFweek Robotics)
Baotong AIGC content platform
Baotong AniShort AIGC content platform. (Source: OFweek Robotics)

Editor’s note: This is an adapted translation of the original OFweek Robotics report. It has been trimmed and restructured for readability for an international business audience.

Translated and adapted from OFweek Robotics (https://robot.ofweek.com/2026-09/ART-8321202-12003-30702348.html).

Leave a comment