Sun Huanbo, Spiers Adam, Lee Hyosang, Fiene Jonathan, Martius Georg
Max Planck Institute for Intelligent Systems, Tübingen, Germany.
Yale University, New Haven, CT, USA.
Nat Commun. 2025 Aug 28;16(1):8031. doi: 10.1038/s41467-025-63230-7.
Robots can benefit from touch perception for enhanced interaction. Interaction involves tactile sensing devices, contact objects, and complex directional force motions (normal and shear) in between. We introduce a comprehensive theory unifying them to advance sensor design, explain shear-induced performance drops, and suggest application scenarios. Our theory, based on sensor isolines, achieves superresolution sensing with sparse units, avoiding dense layouts. Through structural analysis of the sensor perception field, force sensitivity, and contact object effects, we also explore the force direction influences: normal, tangential shear, and radial shear forces. The model predicts an inherent accuracy reduction under shear forces compared to pure normal forces. Validation used Barodome, a 3D sensor predicting contact locations and decoupling shear/normal forces. Its performance confirmed the significant impact of shear forces, with observed drops (0.5 mm) closely matching theoretical predictions (0.33 mm). This theory provides valuable guidance for future tactile sensor design and advanced robotic touch systems.
机器人可以通过触觉感知来增强交互。交互涉及触觉传感设备、接触物体以及它们之间复杂的定向力运动（法向和切向）。我们引入了一个综合理论来统一这些要素，以推动传感器设计、解释剪切力导致的性能下降，并提出应用场景。我们基于传感器等值线的理论，通过稀疏单元实现了超分辨率传感，避免了密集布局。通过对传感器感知场、力灵敏度和接触物体效应的结构分析，我们还探讨了力的方向影响：法向力、切向剪切力和径向剪切力。该模型预测，与纯法向力相比，剪切力作用下固有精度会降低。验证使用了Barodome，这是一种预测接触位置并解耦剪切力/法向力的3D传感器。其性能证实了剪切力的显著影响，观察到的下降（0.5毫米）与理论预测（0.33毫米）非常接近。该理论为未来触觉传感器设计和先进的机器人触摸系统提供了有价值的指导。
