Geometrical calibration and uncertainty estimation methodology for a novel self-propelled miniature robotic machine tool. (February 2018)

• Record Type: Journal Article
• Authors: Olarra, Aitor; Axinte, Dragos; Kortaberria, Gorka
• Is Part Of: Robotics and computer-integrated manufacturing. Volume 49(2018)
• Page Start: 204
• Page End: 214
• Publication Date: 2018-02
• Subjects: Kinematical calibration -- Uncertainty -- Observability index -- Parallel kinematic machine -- Hexapod -- Parameter identification; Robots, Industrial -- Periodicals; Computer integrated manufacturing systems -- Periodicals; Robotics -- Periodicals; Robots industriels -- Périodiques; Productique -- Périodiques; Robotique -- Périodiques; 670.285
• DOI: 10.1016/j.rcim.2017.06.011
• Languages: English
• ISSNs: 0736-5845

Abstract

Highlights: Novel calibration solution for a Stewart–Gough platform. New robotic developments such as the self-propelled hexapod considered in this paper require autonomous calibration solutions to accurately identify the kinematics of the robot before each use, this implies on-board metrology means, no operator intervention and fully automated solution. Such a solution is presented in the paper. Novel calibration configuration evaluation methodology in design phase. A novel and powerful analytical methodology that supports the design of calibration solutions is presented. This methodology is a step forward from the well-known observability index methodology that has been reported in the robot calibration literature. Besides enabling the comparison of different calibration configurations, it is possible now to estimate the uncertainties of the identified parameters. The new methodology is compared to observability indexes and is also experimentally validated. Abstract: This paper reports on a novel calibration method which enables completely automatic identification of the kinematics of a walking hexapod robotic machine tool. The method uses three on-board cameras and relies on a coupled model that combines kinematics and photogrammetry. Both the mathematical modelling and the actual implementation are detailed. Besides the calibration method, the paper proposes an analytical methodology to estimate the uncertainties of the identified kinematical parameters. The methodology is validated against both experimental results and against previously reported observability indexes. This methodology enables moving from qualitative indexes, observability indexes, to quantitative estimations. The methodology is applied to guaranty a calibration configuration that allows estimating the robot parameters with an uncertainty of 0.1 mm due to non-repeatability of the measurements.