Collaborative robots are envisioned to assist people in an increasing range of domains, from manufacturing to home care;however, due to the variable nature of these fields, such robots will inevitably face unfamiliar ...
详细信息
ISBN:
(数字)9783030503345
ISBN:
(纸本)9783030503338;9783030503345
Collaborative robots are envisioned to assist people in an increasing range of domains, from manufacturing to home care;however, due to the variable nature of these fields, such robots will inevitably face unfamiliar situations and unforeseen task requirements, and must be able to interact with users who possess diverse skill sets, backgrounds, and needs. Presently, robust, autonomous solutions for appropriately handling these vast possibilities and uncertainties are unattainable. end-user robot programming offers an alternative approach that lets endusers provide task specifications and author robot skills to meet their own specific contextual constraints and custom task needs. Contextual information-such as task objects, environmental landmarks, and user preferencesis essential in realizing desirable, flexible, and reliable robot programs. However, most robotprogramming systems at present do not afford intuitive ways of specifying contextual information. In this paper, we draw on our prior work to illustrate the barriers to end-user robot programming when using a state-of-the-art programming interface. We then present two case studies that explore new approaches to providing a robot system with contextual information about the user, task, and environment, and how these methods can help improve task performance and user experience. We discuss our findings and future directions for building effective end-userprogramming tools to bring robotic assistance closer to everyday users.
This paper introduces situated tangible robotprogramming, whereby a robot is programmed by placing specially designed tangible "blocks" in its workspace. These blocks are used for annotating objects, locati...
详细信息
ISBN:
(纸本)9781450343367
This paper introduces situated tangible robotprogramming, whereby a robot is programmed by placing specially designed tangible "blocks" in its workspace. These blocks are used for annotating objects, locations, or regions, and specifying actions and their ordering. The robot compiles a program by detecting blocks and objects in its workspace and grouping them into instructions by solving constraints. We present a proof-of-concept implementation using blocks with unique visual markers in a pick-and-place task domain. Three user studies evaluate the intuitiveness and learnability of situated tangible programming and iterate the block design. We characterize common challenges and gather feedback on how to further improve the design of blocks. Our studies demonstrate that people can interpret, generalize, and create many different situated tangible programs with minimal instruction or with no instruction at all.
暂无评论