Abstractions for design-by-humans of heterogeneous behaviors

This paper will discuss the use of movement observation, taxonomy, and expert knowledge, as found in Laban/Bartenieff Movement Studies to facilitate the production of diverse robotic behaviors that are easy to construct for a wide range of people. A ‘behavior’ will be defined by a set of movement primitives that are scaled and sequenced differently for distinct behaviors. These methods will be applied to real robotic platforms, in the context of manufacturing, motivating the fundamental value of high-level abstractions that produce a wide array of behavior. In particular, the integration of methods for designing an alternate set of “knobs” for motion control, abstracting sequencing behaviors and interactions of teams, and controlling robots via a web-based platform will be outlined in a language accessible to an interdisciplinary audience.