UBTECH wins first prize in Guangdong Science and Technology Progress Awards for robot collaboration project
瞄准高危、高精度机器人协同作业，优必选联合项目拿下省科技进步一等奖
UBTECH participated in a joint research project titled 'Key Technologies and Applications of Robot Collaborative Perception and Decision-Making in Complex Dynamic Environments', which won the first prize of the Guangdong Province Science and Technology Progress Award for 2025. The project team will continue to refine collaborative perception and decision-making technologies, conducting test iterations in actual scenarios such as nuclear power inspection, intelligent warehousing logistics, and intelligent manufacturing. UBTECH will play a leading role in industrialization, accelerating the application of technical achievements in real business scenarios.
Source: Chinese robotics — Deployment discovery · Read original article ↗
English translation unavailable. Choose the original article to read it.
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
Reported numbers
training hours
30,000 hours
View original evidence
模型基于3 万小时异构数据训练，配套Data Factory 数据基础设施，Motion Probe 消融实验也证实，多模态表征是真正的深度耦合，并不是多路信息简单拼在一起。
Open source S20standard count
73
View original evidence
与此同时， 优必选也积极推动伺服驱动器、灵巧手、端侧芯片等核心零部件产业链发展， 牵头及参编的首批六项人形机器人国家标准已于近期正式发布， 覆盖感知、决策、运动控制、仿真测试等关键方向，目前已累计参与起草发布 73 项各类标准，积极推动行业规范化建设。
Open source S22
- paper: not_reported
- code: not_reported
- weights: not_reported
- dataset: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
模型基于3 万小时异构数据训练，配套Data Factory 数据基础设施，Motion Probe 消融实验也证实，多模态表征是真正的深度耦合，并不是多路信息简单拼在一起。
Open source S20
沙利文研究报告显示， 2026 年上半年，优必选位列全球人形机器人收入榜榜首， 是唯一同时在整体人形机器人、全尺寸具身智能人形机器人两大赛道收入市占率排名第一的企业。
Open source S23
与此同时， 优必选也积极推动伺服驱动器、灵巧手、端侧芯片等核心零部件产业链发展， 牵头及参编的首批六项人形机器人国家标准已于近期正式发布， 覆盖感知、决策、运动控制、仿真测试等关键方向，目前已累计参与起草发布 73 项各类标准，积极推动行业规范化建设。
Open source S22
其二，优必选自研具身世界模型 Thinker-WM，在 WorldOlympiad、EWMBench、WorldArena1.0 Track-1 三项行业权威基准评测全部取得第一。
Open source S18
平台以Real2Sim2Real为核心，搭建起真实场景数字化、规模化仿真数据合成、机器人能力评测的一站式体系，用来降低仿真环境搭建成本，提升机器人算法训练与验证效率。
Open source S15
人形机器人在高危、复杂现场反复开展真机测试成本高昂，部分场景还存在安全风险。
Open source S16
Source：Chinese robotics — Deployment discovery · gkzhan.com