Flexible Tactile Sensor Array for Slippage and Grooved Surface Recognition in Sliding Movement

Abstract

Flexible tactile sensor with contact force sensing and surface texture recognition abilities is crucial for robotic dexterous grasping and manipulation in daily usage. Different from force sensing, surface texture discrimination is more challenging in the development of tactile sensors because of limited discriminative information. This paper presents a novel method using the finite element modeling (FEM) and phase delay algorithm to investigate the flexible tactile sensor array for slippage and grooved surfaces discrimination when sliding over an object. For FEM modeling, a 3 × 3 tactile sensor array with a multi-layer structure is utilized. For sensor array sliding over a plate surface, the initial slippage occurrence can be identified by sudden changes in normal forces based on wavelet transform analysis. For the sensor array sliding over pre-defined grooved surfaces, an algorithm based on phase delay between different sensing units is established and then utilized to discriminate between periodic roughness and the inclined angle of the grooved surfaces. Results show that the proposed tactile sensor array and surface texture recognition method is anticipated to be useful in applications involving human-robotic interactions.

Keywords: finite element modeling; grooved surface; inclined angle; spectral analysis; surface texture; tactile sensor array; wavelet transform.

Figure 1

(a,b) Structure of the flexible tactile sensor array; (c) Fabricated tactile sensor array.

Figure 2

(a) 3D finite element modeling (FEM) model of flexible tactile sensor array; (b) Close-up view of sensing unit.

Figure 3

Measured nominal stress versus strain curves of rubber, room temperature vulcanizable (RTV) and polydimethylsiloxane (PDMS) materials.

Figure 4

Experimental setup for the validation tests.

Figure 5

The stress distribution in the cross-section view of the sensing unit at the end of (a) compress and (b) sliding. A-A cross-section at the end of (c) compress and (d) sliding.

Figure 6

(a) Simulated normal force extracted from the left and right area of the patterned electrodes in the sensing unit (b) Measured voltages of R₁ and R₃ when sliding along a flat surface, DSWT analysis of the simulated normal force in left area (c) and right area (e) DSWT analysis of the measured voltages in R₁ (d) and R₃ (f).

Figure 7

(a) Force-time curves of No. 1–3 units, (b) schematic view for the calculation of inclined angle α.

Figure 8

Flow chart and procedure to calculate the inclined angle and spatial period value for the grooved surfaces.

Figure 9

(a) Three grooved surfaces with different spatial periods of 0.9, 1.2 and 1.5 mm; (b) Measured voltage; (c) Spectrum analysis for surface texture recognition.

Figure 10

Simulated normal force curves when the sensor array is sliding along the grooved surface with an inclined angle α equal to (a) 0°, (c) 30° and (e) 60°. Measured voltages from the validation test for inclined angle calculation when α equals (b) 0°, (d) 30°, (f) 60°.