Annotating Everyday Grasps In Action

Grasping has been well studied in the robotics and human subjects literature, and numerous taxonomies have been developed to capture the range of grasps employed in work settings or everyday life. But how completely do these taxonomies capture grasping actions that we see every day? In a study to classify all actions during a typical day, we found that single entries in an existing grasp taxonomy were insufficient, apparently capturing not one grasp, but many. When we investigated, we found that these seemingly different grasps could be distinguished by features related to the grasp in action, such as the intended motion, force, and stiffness. In collaboration with our subjects, we developed an annotation scheme for untrained annotators to use, which captured the differences we observed between grasping actions. This chapter describes our annotation scheme. We discuss parallels to and differences from Laban Movement Analysis, which has been long developed to capture motion and action, but does not focus on grasping. We also discuss parallels to impedance or operational space control, with the goal of moving from annotations to actionable robot control.