Soft Barometric Tactile Sensor Utilizing Iterative Pressure Reconstruction

Abstract

Tactile sensors play a pivotal role in enhancing robotic gripping systems across various industries, allowing robots to interact with their environment with precision and adaptability. This study introduces a novel soft barometric tactile sensor designed to capture pressure data beneath the contact pad within an elastomer layer. By employing an iterative fitting process for a parameterized pressure distribution, the sensor eliminates the need for extensive calibration or complex training procedures. A novel metric was introduced to evaluate the accuracy of pressure reconstruction in tactile sensors. This was complemented by quantitative investigations. Moreover, the precision of multiple contact detection and the capabilities of force estimation were examined. Finally, the practical applications of the sensor were demonstrated through in-hand measurement of object orientation. The sensor exhibits promising results, accurately reconstructing pressure distributions for a wide range of indenters, reliably discerning two contacts when their distance exceeds 5 mm and estimating forces with a Mean Absolute Error (MAE) of 1.83 N. Additionally, the in-hand reorientation achieved a Mean Absolute Rotation Error (MARE) of 6.81°, demonstrating the sensor’s potential for real-world applications.

Keywords

• Tactile sensors
• grasping
• robotics and automation