Monitoring of Autonomous Threaded Fastening based on Curve Fitting and LSM Estimation

M. Klingajay (Thailand) and N.I. Giannoccaro (Italy)


Automated Threaded Fastening, Screw insertions, Parameter estimation, Curve Fitting, Least Square Method.


The thread fastenings have been known and used for decades with the purpose of joining one component to another. Thread fastenings are a common assembly method that have been accounting for over a quarter of all assembly operations. This operation is widely used and popularly applied, because it is permitted easy to disassembly, reassembly, and relocation for maintenance, repairing. There is very little research on automating threaded fastenings, and most research on automated assembly focus on the peg-in-hole assembly problem. Screw insertions are typically carried out manually with same purposes of threaded fastening in joining one and another component that is carried out manually. Prime monitoring based-on parameter estimation is the essential parts in the automated manufacturing systems. Information about process conditions enables operations to improve the quality of products. This paper presents a successful monitoring framework and technique for a modern identification system for the manufacturing systems. A modern end-of-line electronics manufacturing environment is a combination of individual cells, designated to complex assembly and material processing tasks. The success of such a system is greatly depended on the stable operation of every task. Deviations from normal situation decrease productivity. Continuous monitoring is therefore required to maintain the functionality and quality of the production. Curve-fitting technique reasoning and basic filter methods are combined to reduce the noise from the online process. Its signal has used to identify changes in normal model residuals. Estimation model are used for prediction of two unknown parameters that is required for the process in selected time series signal. The online outputs after filtering are then applied to monitoring task. Principles of the monitoring method are briefly discussed and demonstrated with the experimental tests. Modelling results indicate that the proposed method can handle noise in experiment data. Generalisation ability of the normal model was also notified. Based on experiments, presented monitoring approach was verified to have potential features to be implemented as an online application.

Important Links:

Go Back