Designing Library of Skill-Agents for Hardware-Level Reusability

TL;DR

A necessary and sufficient skill-agent set corresponding to cover all possible actions is defined, and the design principles for these skill agents in the library are considered, and the design principles for these skill agents in the library are considered.

Abstract

To use new robot hardware in a new environment, it is necessary to develop a control program tailored to that specific robot in that environment. Considering the reusability of software among robots is crucial to minimize the effort involved in this process and maximize software reuse across different robots in different environments. This paper proposes a method to remedy this process by considering hardware-level reusability, using Learning-from-observation (LfO) paradigm with a pre-designed skill-agent library. The LfO framework represents the required actions in hardware-independent representations, referred to as task models, from observing human demonstrations, capturing the necessary parameters for the interaction between the environment and the robot. When executing the desired actions from the task models, a set of skill agents is employed to convert the representations into robot commands. This paper focuses on the latter part of the LfO framework, utilizing the set to generate robot actions from the task models, and explores a hardware-independent design approach for these skill agents. These skill agents are described in a hardware-independent manner, considering the relative relationship between the robot's hand position and the environment. As a result, it is possible to execute these actions on robots with different hardware configurations by simply swapping the inverse kinematics solver. This paper, first, defines a necessary and sufficient skill-agent set corresponding to cover all possible actions, and considers the design principles for these skill agents in the library. We provide concrete examples of such skill agents and demonstrate the practicality of using these skill agents by showing that the same representations can be executed on two different robots, Nextage and Fetch, using the proposed skill-agents set.

Figures (49)

• Figure 1: An example of a pick-up task. Object states can be defined by possible directions of motion for the object. The distribution of possible directions of motion can be represented as a region on a Gaussian sphere. When an object sits on a table, the Northern hemisphere represents the possible directions for the object, with the normal direction of the table as the North Pole of the sphere. The pick-up task can be defined as one causing the transition of the region from a hemisphere to a whole sphere.
• Figure 2: Translational states and rotational states. (a) Translational states. For the sake of simplicity, we grouped three partial translational states (i.e., a hemisphere (PC1), a crescent (PC2), and a polygonal shaped state (PCN)) into one PC state, and two one-way prismatic translational states (i.e., a hemi-circle (OT1) and an arc-shaped state (OT2)) into one OT state. (b) Rotational states. See ikeuchi2021semantic.
• Figure 3: Translational tasks and rotational tasks
• Figure 4: State transitions along motion
• Figure 5: State transitions in the dimension orthogonal to the motion