Toward a New Approach for Human-Robot Interaction

点击翻译按钮获取中文摘要

One remarkable event in the 21th century is the arise of service robotics. The service robots are designed and built for doing a specified work for humans. These robots include the autonomous, mobile or bipedal professional service robots, which are modern robotic platforms equipped with the most advanced sensors, actuators and processing capacities. This new class of robots counts as a new and promising technology that will work side by side with people, collaborating with employees on tasks. MY MAIN OBJECTIVE in the current proposal is making the COOPERATIVE human-robot interaction (HRI) tasks more safe and intuitive. The question of safety issue in HRI is one of the most challenging problems in robotics. Besides the capability of robots to understand and interpret human intentions is another difficulty which is still poorly understood. The HRI has been recognized as a multidisciplinary scientific field in the mid of the 20th century. Nevertheless, the cooperative tasks involving human beings and robots are still limited to few simple tasks. This is partially because of the lack of control algorithms and motion planning strategies that are able to achieve these cooperative tasks appropriately. In the scope of my research program, I plan to develop a human safety model, and an interpreter of human intentions in the context cooperative HRI tasks. These models will be based on the modern motion planning approaches, advanced numerical optimization techniques and human practical reasoning theory. The ultimate goal is to execute the previous models on a real robotic platform interacting cooperatively and, importantly, safely with human beings. The expected results will have valuable impact on HRI and service robotics fields, as well as on other scientific fields, e.g. cognitive science, neuroscience, human-interface design, etc. The future of "co-worker robots" in fields as diverse as the military, medical services, agriculture, rescue and health-care IS REALLY NEAR.