Model-based robot control in human-in-the-loop cyber physical systems

Abstract

It is evident that over the last decade, complex problems and challenges in a diverse set of applications have led to the emergence of innovative approaches and research in human-in-the-loop cyber physical systems (HiLCPS). Robots play an integral role as physical agents within a HiLCPS and enable safer and more effective interactions within dirty, dull, and dangerous environments. Previous approaches and methods do not extend well this new class of systems due to an inability to incorporate human input and scale with the complexity and variety of components. As a result, there is a need to investigate, design, implement and validate novel shared control techniques that integrate humans and robots in a heterogeneous team to enable reliable, robust, and sufficiently agile interaction, communication, and operation. We posit that there are certain tasks humans are (and will be) superior to robots such as perception, intuitive control, and high-level decision-making, on the other hand, there are tasks robots can (or should) perform such as precise low-level motion planning, solving an optimization problem, and operating in hazardous situations. Therefore, the investigation of new control interfaces and shared control methods that can effectively delegate tasks and blend the control between the robot and human operator will enable us to field robot systems that act in direct support of humans.