Advanced lumber manufacturing model for increasing yield in sawmills using GPR-based defect detection system

摘要

This paper proposes an advanced lumber manufacturing model for real-time process control in saw mills in order to increase the yield of high value defect-free lumber. Detecting subsurface defects by scanning canted logs and generating a process plan to cut the logs can increase the yield of high-grade lumber in a saw mill industry. The defect detection process is performed using the ground-penetrating radar (GPR) system. More recently, a defect detection algorithm was developed to process GPR scanned data using the MATLAB? software. This research uses the distance and depth coordinates generated by the defect detection algorithm to develop the process plan that generates a cutting sequence for the resaw machine. The process plan is in the form of an algorithm written in MATLAB? with a simple user interface. The generated cutting sequence was validated by comparing to the conventional sawing sequence, where the operator of the resaw machine randomly performs the cutting of boards. An increase in the yield (in terms of dollar value) of about 27% was noticed using the GPR-based detection system which can map interior defects prior to sawing the log and enable an optimal sawing pattern.