The study of sensor structure to distinguish a surface morphology for psychological tactile sensor

Abstract

Artificial tactile sensors have been recently studied by various research groups mimicking the human sensing mechanism. Tactile sense takes an important part in the human’s daily life such as griping and, perceiving objects based on information, their shape, size, temperature and texture. To get the tactile information, the human skin is packed with various tactile receptors corresponding to each of these stimuli. Through these receptors and nervous system, humans can feel various tactile sensations at the same time without limitation. Additionally, the human perceive diverse psychological sense such as roughness, hardness, warm, cool or even the pain by touching some external objects. However, researches about these psychological feeling have not been done so far because the degree to which each person feels about the same material is different. Therefore, there are a lot of issues as to which physical elements are most important for generating psychological sensations. If robots or artificial systems can generate this psychological feeling, more creative or human-like work will be possible. To give these psychological feelings to robot or artificial system, detecting surface characteristics of object should be a starting point among various tactile parameters. Many researchers are approaching the study of surface characteristics in various aspects such as friction, vibration, strain, softness/hardness, even the viscosity. Among them, the study of surface morphology is a very important parameter. There are some information related to surface morphology such as texture shape, height and width of the object, surface flexion and so on. Various sensing device such as optical device, piezo-photonic sensor and commercialized products like as surface profiler and AFM can be used to get the information of surface morphology. However, these things require additional equipment like light source and detector, also it is difficult to attach to the artificial skin, and robot arms. Therefore, in this paper, new artificial tactile sensor design based on a PVDF-TrFE were suggested and studied for detecting surface morphology. This piezoelectric material has many good advantages such as a flexibility, an ability to sense dynamic stimuli, an easy-fabrication and a self-power generation. Although, sensor structure should be deformed well according to depth profile of surface to obtain depth information of surface morphology, general pressure sensors based on 2-dimensional design fabricated by micro-fabrication process have been shown a poor deformation behavior to millimeter level curvature. To solve this problem, 3-dimensional probing tactile sensor design was suggested. A spring was mounted on the cell of sensor mimicking the deformation of human skin. Sharp tip was designed and fabricated to get spatial resolution, which positioned on the top of the spring. This structure allows for more accurate scanning. Also, flat structure is fabricated for transmitting the force well. So, the spring is compressed according to surface morphology. These spring deformations contain information about the surface and can be transmitted and interpreted through the sensor layer. The tactile sensor with spring structure can detect the deformation of spring through simple and easy calculation and signal processing. In the Integrate & Add graph, we can find out the peak of the signals mean that the spring is compressed to the maximum. The tactile sensor was pressed at four different depth (1, 2, 3 and 4 mm) under the condition of 10 mm/s. The depth which is the same with sensor is compressed can be obtained accurately in consideration of common error. Also, it is able to restore the width data of object surface by using sliding experiment. This experiment was carried out at three different speeds of 5, 10, and 20 mm/s for reliability. The restored information is highly accurate compared to the real structure. Also, using a bidirectional scanning method with a spring mounted sensor, we can obtain the perfect width information of the surface morphology with just a single cell without using an array structure. The information from the additional opposite scanning is used to compensated for the deficient part of the morphology. And, as the height of some morphology of object surface increases, the degree of deformation of PVDF film also increase, so the area of integrate & add also shows a tendency to increase. The ability of this sensor and spring mounted structure will be helpful for next generation artificial skin which is sense physical and psychological sensations. ⓒ 2017 DGIST