Automated Shovel Tooth Wear Monitoring with Machine Vision

Tooth wear on mining shovels presents a challenge for productivity. A shovel with worn teeth suffers from increased digging forces, longer fill times, and an increased likelihood of lost or broken teeth and adapters. Efficient maintenance planning relies on knowledge of tooth decay rates and patterns. A fully automated wear monitoring system can record detailed logs of tooth wear, permitting G.E.T. personnel to optimize maintenance scheduling. Studies have demonstrated that planned tooth change-outs result in significantly reduced costs as compared to unplanned maintenance. A machine-vision-based tooth-wear monitoring method is introduced. A rugged camera is installed on the boom of cable shovels, or on the stick of hydraulic shovels. The images are processed using a photogrammetric algorithm to determine the individual tooth lengths. Tooth length data is collected at intervals during shovel operations, forming a profile of the current wear state, speed of decay, wear pattern, and expected lifespan of teeth. A case study in a South American mine on a Komatsu PC8000 hydraulic front shovel is presented. Tooth wear is tracked over a 6-day period and wear rate statistics are computed for 6 teeth. This validates the system and provides an example of how wear patterns can be recorded and analysed by an automatic system. Armed with this data, mine personnel can reduce the cost and lost time due to unplanned maintenance, and observe the effects of mine conditions and operator behaviour on tooth wear.