In this paper, we introduce the concept of a human-inspired pain sensing system, imitating the sensing location of mechanical nociceptors in human skin (superficial somatic pain) and skeletal muscle (deep somatic pain) for personal care robot safety application. We expect that the system will not only measure the pain for personal care robot safety evaluation, but also show the difference between these two kinds of pain due to various contact conditions. In this study, we developed the arm prototype which imitates human nociceptor mechanism, and verified our pain sensing concept by conducting experiment with various contact situations.