Robotics Data Collection: Why Physical Interactions Are Hard to Replicate
机器人会翻跟斗，却拧不开瓶盖：卡住它的不是手，是没有人教。
The article explores the challenges of training humanoid robots by highlighting the lack of available data and the difficulty of replicating physical interactions across different robot bodies and environments. It discusses the shift from data abundance to data scarcity, the rise of data factories, and the need for new data infrastructure and standards to support embodied AI development.
The article explains the challenges of data collection for embodied AI, emphasizing the difficulty of replicating physical interactions across different robot bodies and environments. It highlights the shift from data abundance to data scarcity and the need for new data infrastructure and standards.
Source: Chinese robotics — Dataset and collection 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
required data hours
10,000,000 hours
data hours needed
10,000,000 hours
data hours needed for practical operation
10,000,000 hours
- paper: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
2025年，整个具身智能行业积累的有效数据，大约只有十万小时量级。
Open source S15
国金证券今年7月的报告估算，要训练一个具备实际作业能力的具身智能模型，行业至少需要1000万小时的多模态交互数据。
Open source S18
而到报告发表那时为止，全球积累的数据不到这个需求的5%。
Open source S19
中间差着两个数量级（按行业测算口径，缺口超过99%）。
Open source S17
而业内普遍认为，要让机器人迎来自己的「ChatGPT时刻」，至少需要千万小时的数据打底。
Open source S16
Source：Chinese robotics — Dataset and collection discovery · huxiu.com