Wear detection in mechanical equipment
Aspects of the technology described herein describe a system for detecting and reducing wear in industrial equipment. Aspects of the technology use 3D image data from a field inspection of industrial equipment to identify and quantify wear. The wear can be detected by providing the images from the field inspection to a computer classifier for recognition. Aspects of the technology also use machine learning to recommend a change to the operation of the equipment to minimize wear. Such a change could include load shedding, lube oil feed rate changes, prompts for maintenance, etc. Through incorporation of the wear data into the control system, the equipment can automatically change operation to improve wear performance and increase the durability and lifetime of the equipment.
1. A method for detecting equipment wear, the method comprising:
receiving a three-dimensional image from a three-dimensional scanning rig depicting a surface of a portion of a piece of equipment, wherein the three-dimensional scanning rig comprises at least a left scanner and at least a right scanner, the at least left scanner and the at least right scanner attached to a handle;
using a machine classifier to detect a specific wear pattern in the image, wherein the machine classifier is trained using labeled images of the specific wear pattern; and
generating a recommendation to change the piece of equipment's operation, wherein the change is predicted to decrease a rate of growth for the specific wear pattern.
2. The method of claim 1 , wherein the recommendation is automatically generated by a machine learning process that takes the wear pattern as input along with monitoring data for the piece of equipment.
3. The method of claim 2 , wherein the monitoring data comprises an analysis of lubrication used with the equipment.
4. The method of claim 2 , wherein the machine learning process is an association rule engine.
5. The method of claim 1 , wherein the machine classifier is a convolutional neural network.
6. The method of claim 1 , wherein the wear pattern indicates wear selected from the group consisting of pitting, spalling, abrasive wear, corrosive wear, machine marks, honing marks, and thermal stress.
7. The method of claim 1 , wherein the portion is a gear tooth.