Description
|

Across many application domains, robots are expected to work in human environments, side by side with people. Interactions between robots and their users will take many forms, from a trained operator supervising several industrial robots, to an older adult receiving care from a rehabilitation robot, to a child safely practicing social, cognitive, or emotional skills with a readily available socially assistive robot. Robotic products are expected to be intuitive, easy to use, and responsive to the needs and states of their users, making human-robot interaction (HRI) a key area of research.
Some areas of robot control deal with very fast interactions with the environment, but HRI is unique in requiring a broad spectrum of temporal dynamics: interactions that happen very quickly (a wink or a twitch of the mouth), interactions that happen very slowly (gradually getting used to a pattern of behavior), and interactions that change unexpectedly (due to context or intent inaccessible to the robot).
The purpose of the PhD theme is to follow this direction in order to deal with the multiple open challenges that characterize the field of HRI. More in detail, the work considers an alternative control algorithm that does not strictly rely on pure motion planning in order to navigate and move in a realistic environment, but that instead allows to specify a desired behavior for the robot, that would be able to accomplish a series of tasks. This is done through the use of optimization-based techniques and Hierarchical Quadratic Programming (HQP), which permits to define a Stack of Tasks (SoT) to be performed by the robot in hierarchical manner. In this way, it possible to account also for human intentions and movements, adjusting the robot behavior on the fly. The aim is thus to improve collaboration between human and robot, increase confidence level and trust in the machine, and ease the human from the most physically and mentally stressful tasks and operations.
|