Drill bit state-oriented drilling process classification with time-series data for wheeled drilling rigs

Abstract

Abstract This paper represents an analysis of the wheeled drilling rig’s drilling process. Thanks to data from the onboard measurement unit of the machine, the characteristics of the drilling process regarding state of the drill bit are identified and calculated. The aim of the work is to provide a comparison between different drill qualities and process classification using Threshold-based segmentation with feed pressure levels and duration of single hole drilling. Second methodology is hierarchical clustering to create cluster analysis. Thanks to these approaches, it is possible to detect the time when the drill bit should be changed. The obtained results state that the average drill time for a new drill bit is shorter approximately by 50% than for the worn-out bit in terms of average drilling duration. Moreover, these changes are visible in the subsystem pressure level of the machine under specific drilling regimes.