No abstract

No abstract

This paper was motivated by the need for reliable robotic agents to maintain spacecraft systems in orbit around the moon, Gateway, and other distant locations, that will be uninhabited for long periods. This effort outlines an analysis methodology for determining the manipulation potential required of a robotic gripper based on a set of tasks. The core of the method is to determine all the possible ways for manipulating the task set interfaces (turning a knob) and defining each of the motions (grasping) in terms of taxonomy (contact, prehensile, motion). By eliminating all but the lowest taxonomy, the minimum manipulation potential can be determined. Using the minimum manipulation potential to choose the gripper means it can be the most robust possible, something vital to long-term distant missions. The methodology was applied to a set of interfaces drawn from !SS and expected Gateway operations. The next step was to apply the methodology to each interface and filter the results. Several interfaces with complicated motions were replaced with simpler interfaces that fulfill the same role. The group also agreed that other interfaces, drills, and fire extinguishers should be specially designed robotic tools. After the analysis, a Commercial off-the-shelf gripper’s fingers were specially designed for the operation interfaces. This approach effectively resulted in a co-design of the gripper and the environment. Several successful demonstrations (first using commercial software and then using ROS, Movelt, and NASA software) were performed to ensure the analysis method and gripper were fully capable.

Remotely programming robots to execute tasks often relies on registering objects of interest in the robot's environment. Frequently, these tasks involve articulating objects such as opening or closing a valve. However, existing humanin-the-loop methods for registering objects do not consider articulations and the corresponding impact to the geometry of the object, which can cause the methods to fail. In this work, we present an approach where the registration system attempts to automatically determine the object model, pose, and articulation for user-selected points using a nonlinear iterative closest point algorithm. When the automated fitting is incorrect, the operator can iteratively intervene with corrections after which the system will refit the object. We present an implementation of our fitting procedure and evaluate it with a user study that shows that it can improve user performance, in measures of time on task and task load, ease of use, and usefulness compared to a manual registration approach. We also present a situated example that demonstrates the integration of our method in an end-to-end system for articulating a remote valve.