Human Robot Workflow - Modelling and Control of Collaborative Tasks

点击翻译按钮获取中文摘要

This research addresses the challenges of Human-Robot collaboration from experimental, modeling and analytical perspectives; and advances methods for safe, effective and efficient human-robot interaction design. My research has provided novel methods for, and demonstrations of, closed-loop human-robot interaction based on physiologic, haptic and gesture based cues applied to building block' scenarios: turn-taking, object passing, resource sharing, command issuance. These human-in-the-loop communication and control methods support the creation and experimental evaluation of human-robot interactions applications in teleoperation, product assembly and collaborative carrying. ***To advance this area of research and create new applications for the numerous robot assistant platforms coming out in the market, a robust framework of knowledge, methods, and design principles for complex interactions is required; one that allows interactions to shift smoothly and safely between different modes of collaboration. My team will develop this framework through investigations of interactions along two tracks: i) temporal interactions, that is, the fluency and rhythm of collaborative human-robot work; and ii) spatial interactions, that is human-robot proxemics. These two tracks directly map onto the range of interactions in shared tasks: from distal, low frequency, supportive interactions to close proximity, high frequency cooperative activity. Through a process of hypothesis generation, prototyping, experimental work, and analysis we will explore each of these tracks, and develop interactions methods that support human-robot collaboration over the full range of work sharing modalities. ***Integrating these two tracks, my team will develop online tools for assessing, communicating, and mediating between the overall physical system state (human, robot/devices, and task), the user's intentions, and the robot controller actions. Further, we will continue to investigate and improve methods and protocols for modelling and testing human-robot interactions so that we can better understand how people and robots work together. ***The study of human-robot interactions across modes of work-sharing, and the development of advanced tools for modelling and design of human-robot teamwork, provides engineers, system designers, integrators the ability to effectively design and implement robotic devices in a wide range of emerging manufacturing and consumer product scenarios. Training of graduate students in this area will advance the field and provide the talent and technology needed to create new market opportunities in the collaborative robotics space. For industry, effective human-robot collaboration, utilizing new robotic systems, will increase productivity and quality while giving companies an edge in competitive global markets.