RESEARCH ARTICLE Juan Carlos JÁUREGUI, Eusebio E. HERNÁNDEZ, Marco CECCARELLI, Carlos LÓPEZ-CAJÚN, Alejandro GARCÍA Kinematic calibration of precise 6-DOF Stewart platform-type positioning systems for radio telescope applications © Higher Education Press and Springer-Verlag Berlin Heidelberg 2013 Abstract The pose accuracy of a parallel robot is a function of the mobile platform posture. Thus, there is no a single value of the robot’s accuracy. In this paper, two novel methods for estimating the accuracy of parallel robots are presented. In the first method, the pose accuracy estimation is calculated by considering the propagation of each error, i.e., error variations are considered as a function of the actuator’s stroke. In the second method, it is considered that each actuator has a constant error at any stroke. Both methods can predict pose accuracy of precise robots at design stages, and/or can reduce calibration time of existing robots. An example of a six degree-of-freedom parallel manipulator is included to show the application of the proposed methods. Keywords pose errors, error estimation, parallel robot, radio telescopes 1 Introduction Parallel robots have become an excellent solution for applications where precise position and orientation are needed. They have different designs and configurations and essentially they consist of a moving platform connected to the ground through a series of kinematic joints. It is called parallel robot because it is formed as a closed chain mechanism, and at a reference position, the moving platform stays parallel to the fixed platform. Their mobility depends on the degrees-of-freedom (DOF) of the moving platform that depends on the number and type of joints. Although their workspace is limited as compared to serial robots, they provide higher accuracy than the open arm robots. The advantages of parallel robots are their dynamic stability, high stiffness, and high pose accuracy. For these reason, they are suited for very precise applications such as machine tools, surgery devices or scientific instruments. For these applications, it is neces- sary to determine the accuracy of the entire workspace. Due to their design, there is no one-to-one relationship between controllable variables and the degrees-of-reedom, and each controllable variable affects all DOFs and vice versa. Besides the complexity of the mechanism kine- matics, the pose error depends mainly on the accuracy of the actuators, the correlation among the controllable variables, the control scheme, and the elastic deformations. Therefore, the algorithms that determine the end-effector- pose accuracy are based on the kinematic model. There are many papers regarding their kinematic and dynamic models. Merlet stated that a parallel robot is able to support heavy loads, move at high velocity operation, and perform at high repeatability [1], but, within a limited workspace [2]. In this way, different indices have been proposed to evaluate their performance such as: dexterity, manipulability or global conditioning index. One of the applications where high accuracy is needed is in machine tools, and there are several publications related to the accuracy of parallel robots applied to machine tools. To be applied as a machine tool, calibration strategies should be clearly defined [3]. In contrast, conventional machine tools consist on three mutually orthogonal axes, Received October 20, 2012; accepted January 24, 2013 Juan Carlos JÁUREGUI (✉) División de Estudios de Posgrado, Facultad de Ingeniería, Universidad Autónoma de Quéretaro Quéretaro, Qro. Mexico E-mail: jc.jauregui@uaq.mx Eusebio E. HERNÁNDEZ National Polytechnic Institute, IPN, Section of Graduate Studies and Research, ESIME-UPT, México D.F., Mexico Marco CECCARELLI Laboratory of Robotics and Mechatronics University of Cassino, Italy Carlos LÓPEZ-CAJÚN Universidad Autónoma de Quéretaro Querétaro, Qro. México Alejandro GARCÍA CIATEQ, A.C. Aguascalientes, Ags. México Front. Mech. Eng. 2013, 8(3): 252–260 DOI 10.1007/s11465-013-0249-7