Back

Haptic Intelligence Autonomous Learning Members Publications

Insight: a Haptic Sensor Powered by Vision and Machine Learning

Insight v2
We introduce a new soft haptic sensor that uses vision and learning to accurately estimate where and how it is being contacted. The sensor constantly records images from the inside using a camera. Feeding these images and a reference image into a trained deep neural network makes it possible to estimate the directional force distribution all over the sensing surface.

Members

Thumb ticker sm 20220401 huanbo sun 2 min
Autonomous Learning
  • Doctoral Researcher
Thumb ticker sm thumb ticker kjk 2024
Haptic Intelligence
Director
Thumb ticker sm georg 2018 crop small
Empirical Inference, Autonomous Learning
Senior Research Scientist

Publications

Autonomous Learning Haptic Intelligence Article A Soft Thumb-Sized Vision-Based Sensor with Accurate All-Round Force Perception Sun, H., Kuchenbecker, K. J., Martius, G. Nature Machine Intelligence, 4(2):135-145, February 2022 (Published)
Vision-based haptic sensors have emerged as a promising approach to robotic touch due to affordable high-resolution cameras and successful computer-vision techniques. However, their physical design and the information they provide do not yet meet the requirements of real applications. We present a robust, soft, low-cost, vision-based, thumb-sized 3D haptic sensor named Insight: it continually provides a directional force-distribution map over its entire conical sensing surface. Constructed around an internal monocular camera, the sensor has only a single layer of elastomer over-molded on a stiff frame to guarantee sensitivity, robustness, and soft contact. Furthermore, Insight is the first system to combine photometric stereo and structured light using a collimator to detect the 3D deformation of its easily replaceable flexible outer shell. The force information is inferred by a deep neural network that maps images to the spatial distribution of 3D contact force (normal and shear). Insight has an overall spatial resolution of 0.4 mm, force magnitude accuracy around 0.03 N, and force direction accuracy around 5 degrees over a range of 0.03--2 N for numerous distinct contacts with varying contact area. The presented hardware and software design concepts can be transferred to a wide variety of robot parts.
DOI URL BibTeX

Autonomous Learning Haptic Intelligence Patent Method for force inference, method for training a feed-forward neural network, force inference module, and sensor arrangement Sun, H., Martius, G., Kuchenbecker, K. J. (PCT/EP2021/050231), Max Planck Institute for Intelligent Systems, Max Planck Ring 4, January 2021
The invention relates to a method for force inference of a sensor arrangement for sensing forces, to a method for training a feed-forward neural network, to a force inference module, and to a sensor arrangement.
BibTeX

Autonomous Learning Haptic Intelligence Patent Sensor Arrangement for Sensing Forces and Methods for Fabricating a Sensor Arrangement and Parts Thereof Sun, H., Martius, G., Kuchenbecker, K. J. (PCT/EP2021/050230), Max Planck Institute for Intelligent Systems, Max Planck Ring 4, January 2021
The invention relates to a vision-based haptic sensor arrangement for sensing forces, to a method for fabricating a top portion of a sensor arrangement, and to a method for fabricating a sensor arrangement.
BibTeX