Action representations in robotics: A taxonomy and systematic classification

Abstract

Understanding and defining the meaning of "action" is substantial for robotics research. This becomes utterly evident when aiming at equipping autonomous robots with robust manipulation skills for action execution. Unfortunately, to this day we still lack both a clear understanding of the concept of an action and a set of established criteria that ultimately characterize an action. In this survey we thus first review existing ideas and theories on the notion and meaning of action. Subsequently we discuss the role of action in robotics and attempt to give a seminal definition of action in accordance with its use in robotics research. Given this definition we then introduce a taxonomy for categorizing action representations in robotics along various dimensions. Finally, we provide a systematic literature survey on action representations in robotics where we categorize relevant literature along our taxonomy. After discussing the current state of the art we conclude with an outlook towards promising research directions.

Toward next-generation learned robot manipulation

“…However, in robot manipulation, action primitives are robot dependent. As Zech et al (66) pointed out, a representation of action in robotics is tied to perception, embodiment, and actuation of the robot. For example, action primitives can be drawn from a robot's mobility, sensing, and control primitives.…”

Robot Multimodal Object Perception and Recognition: Synthetic Maturation of Sensorimotor Learning in Embodied Systems

“…Nevertheless, the success of this 10% of attempts is not guaranteed, as the reaching related maps are not fully populated, causing non-refined reaches. In turn, that also affects the system's ability to grasp, a reflection of the latter's dependency to the reaching movement [30]. Starting by randomly generated coordinates of the next target position to reach k x,y,z and by knowing the current hand position h x,y,z within the reach space K, the algorithm iteratively tries to minimise the Euclidean distance until no more movement is possible.…”

Intrinsically Motivated Open-Ended Multi-Task Learning Using Transfer Learning to Discover Task Hierarchy

“…A procedure is a combination of outcomes (ω 1 , ω 2 ) ∈ Ω 2 . Carrying out a procedure (ω 1 , ω 2 ) means executing sequentially each component π 1,2 of the action sequence, where π i is an action that reaches ω i ∀i ∈ [1,2].…”

“…Likewise, as a conceptual illustration, an unbounded sequence of actions is needed to control the interactive table and play tunes of any length in a setup like in Figure 1. In this work, we consider that actions of unbounded complexity can be expressed as action primitives and unbounded sequences of action primitives, also named in [2] respectively micro and compound actions. The agent thus needs to estimate the complexity of the task and deploy actions of the corresponding complexity.…”