SBIR Phase II: Artificial Characterization of Objects Relating to Human Tactile Perception

这个项目更广泛的影响/商业潜力是为目前依赖于专家感官小组（相当于专业品酒师的触觉）的定性数据的行业提供一种量化触觉的新标准。如果机器人要像人类一样行动，推进对触觉感知在感知和操纵中的作用和功能的理解也是必不可少的。研究表明，由于永久性疾病或暂时麻醉而失去知觉的人在精细操作任务中表现不佳（类似于没有触摸的最好的机器人系统）。在这个项目中提出的研究是将触觉感知和感觉运动智能引入下一代机器人的下一步。与人类指尖感知相似的触觉传感器的成功演示将标志着远程操作领域的重大进展，使世界距离远程触觉感知又近了一步。这个小型企业创新研究（SBIR）第二阶段项目旨在开发世界上第一个人类触觉感知标准。触觉识别是一个比视觉和听觉更困难的问题，不仅需要智能的感觉处理，还需要智能算法来选择和控制运动，这对感知的内容有巨大的影响。模仿人类指尖的机械特性和灵敏度的人工传感器直到最近才出现。本文提出的研究将测试一个假设，即一个受生物学启发的机器人系统可以测量与人类通过触摸表征物体的主观感知、描述符和联想相对应的特性。

The broader impact/commercial potential of this project is to provide a new standard of quantifying touch for industries currently relying on qualitative data from expert sensory panels (the tactile equivalent of professional wine tasters). Advancing the understanding of the role and function of tactile sensing in perception and manipulation is also essential if robots are to behave like humans. Studies have demonstrated that humans who cannot feel due to permanent disease or temporary anesthesia perform poorly in fine manipulation tasks (similar to even the best robotic systems without touch). The research proposed in this project is the next step to bring tactile sensing and sensory-motor intelligence to the next generation of robotics. The successful demonstration of a tactile sensor with perceptual similarity to the human fingertip would mark substantial progress in the field of telemanipulation, bringing the world one step closer to remote haptic perception. This Small Business Innovation Research (SBIR) Phase 2 project seeks to develop the world's first standard of human tactile perception. It has been proposed that tactile recognition presents a more difficult problem than vision and hearing, requiring not only intelligent sensory processing, but also intelligent algorithms to select and control movements, which have a tremendous influence on what is sensed. Artificial sensors that mimic the mechanical properties and sensitivity of the human fingertip have not existed until recently. The research proposed herein will test hypotheses that a biologically inspired robotic system can measure properties that correspond to subjective percepts, descriptors and associations that humans use to characterize objects by touch.