Efficient, transformable production systems need robots that are flexible and effortlessly repurposed or reconfigured. The present dissertation argues that this can be achieved through the implementation and use of general, object-centered robot skills. In this dissertation, we focus on the design, implementation and uses of a robot programming paradigm, focused on robot skills, that facilitates intuitive and explicit task-level programming by laymen, such as factory workers, as well as ad-hoc task planning in the skill domain. We show how these robot skills can be modeled and implemented, even on different robot systems. Furthermore, we show how laymen can intuitively program tasks on an advanced mobile manipulator, using the skills as the fundamental building blocks. Finally, we demonstrate how the same skills can be used for ad-hoc task planning, where the robot system instead constructs the task autonomously, exactly when it is needed. It is the firm belief of this researcher that industrial robotics need to go in a direction towards what is outlined in this dissertation, both in academia and in the industry. In order for manufacturing companies to remain competitive, robotics is the definite way to go – and current industrial robot systems are plainly too difficult to program for new tasks on a regular basis.
Robot Skills for Transformable Manufacturing Systems
01.01.2015
Pedersen , M R 2015 , Robot Skills for Transformable Manufacturing Systems . Ph.d.-serien for Det Teknisk-Naturvidenskabelige Fakultet, Aalborg Universitet , Aalborg Universitetsforlag . https://doi.org/10.5278/vbn.phd.engsci.00018
Buch
Elektronische Ressource
Englisch
DDC: | 629 |