Large Workspace Robots

A new class of robots is presented in this research, with the denomination of Large Workspace Robots (LWR). A definition based on the workspace extension and the mass of the robot is provided, to allow a proper classification of existing and new systems. The aim of the present study is to produce a coherent and organic framework to allow the design and analysis of this newly defined class of robotic systems. Several aspects were explored, and the research was partitioned in three parts, the first being topological efficiency methods for Repetitive Workspace Robots (RWR), which are systems where an arbitrarily small conventional robot is used to cover a large surface by moving it in a grid-like fashion. Three algorithms are defined to compute an especially defined index: the Genetic Covering Algorithm (GCA), the Uniform Expansion Covering Algorithm (UECA) and finally the Corrected Inertial Ellipsoid Covering Algorithm (CIECA). The second part regards the topic of controlling robots that are intended to work upon very large surfaces; to this end, a trajectory- and speed profile- planning methodology is defined. The main field of application is robotized painting of photo-realistic images on large surfaces like the faҫade of buildings. Experimental investigations were performed on the spray-cone of paint. The third part is related to hardware and design aspects of LWRs. Different types of actuators are studied that could carry advantages in this context. In particular, an actuator based on Storable Tubular Extendible Members (STEMs) is presented. The design and evaluation of a prototype shows remarkable structural characteristics, very large stroke and a limited bulk. Exploiting this actuator, a planar, parallel, over-actuated robot is presented and analysed in detail. Results show that the over-actuation helps with limiting the singularities of the mechanism and the design proves to be classifiable as a LWR, in light of its extremely large workspace and lightness of the structure. Along the same lines, linear actuators applied to the design of an overhead crane are presented that exploit the mechanisms known as Variable Radius Drum Mechanisms (VRDM). An analytical synthesis methodology is presented to allow these VRD to be designed. The VRD is a drum with a continuously variable radius; this feature allows non-linear coupling between the winding (or unwinding) of the cable and the rotation of the drum itself. Moreover, a modular Cable Driven Parallel Robot (CDPR) is presented for space applications: inspection and light manipulation of the Moon and Mars and other celestial bodies. The system comprises an serial manipulator-equipped rover, three modules and an end-effector. These last components are assembled together by the rover and ultimately constitute the CDPR. In this study, the robot is analysed, particularly for what concerns the influence of the stability of the modules to the morphology of the workspace of the robot. Several applications are described. Finally, several possible applications are presented that encompass multiple of the discussed aspects belonging to LWRs; technological improvement on the CDPR for space applications are discussed as well, particularly regarding actuation and stabilization. The conclusions state the purpose of the definition of the class of LWRs and how the various discussed aspect influence the design and analysis process of this kind of robots.