Robot Touch Sensing: Four Companies Begin to Deliver Orders
机器人触觉，订单开始落地的4家公司
The article discusses the progress of tactile sensing in robotics, focusing on companies such as Fule New Materials, Changying Precision, Cixing Shares, and Daren Electronics. It explains the role of tactile sensors in enabling robots to perform precise tasks, the methods of data collection (including data gloves and robotic hands), and the current state of industry adoption and challenges in scaling tactile data for training and deployment.
Source: Chinese robotics — Company news 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
Reported success rate
98%
Reported success rate
26%
Reported success rate
48%
training data hours
≈500,000 hours
View original evidence
国金证券引用的行业估算认为，2026年初全球高质量真实物理交互数据存量约50万小时，而训练通用具身模型可能需要千万小时级的数据。
Open source S20
- paper: not_reported
- code: not_reported
- weights: not_reported
- dataset: not_reported
What remains unknown
Not established in the collected evidence: Environment, Control, Data origin.
Reported performance applies to the described task. It does not establish general autonomy or deployment readiness.
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
10月28日晚，福莱新材披露了一组数据； 截至今年9月底，公司柔性传感器项目累计实现收入466.75万元，占营业收入的0.18%。
Open source S2
这个数字之所以受到关注，是因为此前福莱新材已经宣布，与机器人灵巧手厂商灵心巧手达成了10万套触觉传感器采购合作。
Open source S3
截至8月底，累计交付量超过3万套。
Open source S4
GPT-6 Astra成功完成49次，成功率达到98%。
Open source S10
但在另一组RoboDojo测试中，任务涉及更复杂的物体操作和精细控制，GPT-6 Astra单独运行的成功率只有26%。
Open source S11
研究人员将它与专门训练的机器人操作模型π0.5结合后，成功率提高到48%。
Open source S12
国金证券引用的行业估算认为，2026年初全球高质量真实物理交互数据存量约50万小时，而训练通用具身模型可能需要千万小时级的数据。
Open source S20
Source：Chinese robotics — Company news discovery · sohu.com