Encoding and presentation of surface textures using a mechanotactile display

Tactile displays can present virtual tactile sensations to the user by stimulating tactile receptors, which are distributed spatially within the skin and are classified into four types based on their characteristic temporal pattern of impulses. The mechanotactile displays can stimulate different types of the receptors selectively by changing the displacement and frequency of the skin deformation. In our previous work, we developed an array of large-displacement microactuators composed of hydraulic amplification mechanism (HDAM) and piezoelectric actuators, which is capable of stimulating all the four types of the tactile receptors. We controlled the driving parameters that were the driving voltage, vibration frequency, and actuation patterns of the actuators, and successfully presented various surface textures. However, in practical applications, the control parameters are to be tuned according to the surface textures that we want to present. In addition, the input signals to the piezoelectric actuators would consist of multiple frequency components and are difficult to create through trial and errors. To solve this inverse problem, in this work, we encoded the sample surface textures to voltage signals by sliding the sample over the HDAM and using the piezoelectric actuators as sensors. First, we attempted to reproduce the surface textures. The encoded signals were amplified to drive the mechanotactile display. The perception tests indicated that this approach was effective. Secondly, we attempted to correlate the encoded signals with hardness of the samples. The encoded signals were investigated among the samples of different hardness. In addition, we fabricated moicropatterned tactile samples that have the same physical properties except Young's modulus to isolate hardness from other parameters. The encoded signals were compared to extract the characteristic signals to determine the hardness. These signals were used to drive the tactile display in the perception tests, which verified the effectiveness of the approach. These two approaches that were proposed and experimentally verified in this paper are readily applicable to solve the inverse problem in tactile display applications.