Ultrathin Three-Axis FBG Wrist Force Sensor for Collaborative Robots

Collaborative robots equipped with multicomponent force-sensing systems at the human-robot-environment interaction interface provides competitive advantages, such as collaboration, flexibility, and reliability. Most wrist force sensors use the cross-beam structures with electrical strain gauge measurement technologies. This article proposes an approach to develop a new ultrathin three-axis fiber Bragg grating (FBG) force sensor. First, an appropriate elastomer with embedded floating beams in ultrathin dimensions to measure three-axis force is designed using a finite-element method (FEM). A temperature-compensated extreme learning machine (TC-ELM) network is proposed to achieve the characteristics of low coupling and temperature compensation (TC). Calibration and experiment on the UR5 robot are conducted, and results illustrate that the highest sensitivity of the sensor can reach 34.07 pm/N, the repetition error is less than 5%, and the maximum coupling error is less than 0.21%. The maximum temperature drift after compensation is less than 20 pm. The proposed sensor has identical output as a commercial sensor and performs well in dynamic loading and unloading operations. Finally, comparison results from commercial and academic alternatives demonstrate the interest of the proposed approach to develop high-performance and integrated force-sensing solutions.