Development of highly sensitive multimodal tactile sensor

The sense of touch is crucial for interpreting exteroceptive stimuli, and for moderating physical interactions with one’s environment during object grasping and manipulation tasks. For years, tactile researchers have sought a method that will allow robots to achieve the same tactile sensing capabilities as humans, but the solution has remained elusive. This is a problem for people in the medical and robotics communities, as prosthetic and robotic limbs provide little or no force feedback during contact with objects. During object manipulation tasks, the inability to control the force (applied by the prosthetic or robotic hand to the object) frequently results in damage to the object. Moreover, amputees must compensate for the lack of tactility by paying continuous visual attention to the task at hand, making even the simplest task a frustrating and time-consuming endeavor. We believe that these challenges of object manipulation might best be addressed by a closed feedback loop with a tactile sensory system that is capable of detecting multiple stimuli. To this end, the goal of our research is the development of a tactile sensor that mimics the human sensory apparatus as closely as possible. Thus far, tactile sensors have been unable to match the human sensory apparatus in terms of simultaneous multimodality, high resolution, and broad sensitivity. In particular, previous sensors have typically been able to sense either a wide range of forces, or very low forces, but never both at the same time; and they are designed for either static or dynamic sensing, rather than multimodality. These restrictions have left them unsuited to the needs of robotic applications. Capacitance-based sensors represent the most promising approach, but they too must overcome many limitations. Although recent innovations in the touch screen industry have resolved the issue of processing complexity, through the replacement of clunky processing circuits with new integrated circuits (ICs), most capacitive sensors still remain limited by hysteresis and narrow ranges of sensitivity, due to the properties of their dielectrics. In this thesis, we present the design of a new capacitive tactile sensor that is capable of making highly accurate measurements at low force levels, while also being sensitive to a wide range of forces. Our sensor is not limited to the detection of either low forces or broad sensitivity, because the improved soft dielectric that we have constructed allows it to do both at the same time. To construct the base of the dielectric, we used a geometrically modified silicone material. To create this material, we used a soft-lithography process to construct microfeatures that enhance the silicone’s compressibility under pressure. Moreover, the silicone was doped with high-permittivity ceramic nanoparticles, thereby enhancing the capacitive response of the sensor. Our dielectric features a two-stage microstructure, which makes it very sensitive to low forces, while still able to measure a wide range of forces. Despite these steps, and the complexity of the dielectric’s structure, we were still able to fabricate the dielectric using a relatively simple process. In addition, our sensor is not limited to either static or dynamic sensing; unlike previous sensors, it is capable of doing both simultaneously. This multimodality allows our sensor to detect fluctuating forces, even at very low force levels. Whereas past researchers have used separate technologies for static and dynamic sensing, our dynamic sensing unit is formed with same capacitive technology as the static one. This was possible because of the high sensitivity of our dielectric. We used the entire surface area effectively, by integrating the single dynamic sensing taxel on the same layer as the static sensing taxels. Essentially, the dynamic taxel takes the shape of the lines of a grid, filling in the spaces between the individual static taxels. For further optimization, the geometry of the dynamic taxel has been redesigned by fringing miniature traces of the dynamic taxel within the static taxels. In this way, the entire surface of the sensor is sensitive to both dynamic and static events. While this design slightly reduces the area that is covered by the static taxels, the trade-off is justified, as the capacitive behavior is boosted by the edge effect of the capacitor. The fusion of an innovative dielectric with a capacitive sensing IC has produced a highly sensitive tactile sensor that meets our goals regarding resolution, noise immunity, and overall performance. It is sensitive to forces ranging from 1 mN to 15 N. We verified the functionality of our sensor by mounting it on several of the most popular mechanical hands. Our grasp assessment experiments delivered promising results, and showed how our sensor might be further refined so that it can be used to accurately estimate the outcome of an attempted grasp. In future, we believe that combining an advanced robotic hand with the sensor we have developed will allow us to meet the demand for human-like tactile sensing abilities.