Algorithms and Automated Material Handling Systems Design for Stacking 3D Irregular Stone Pieces

The motive of this research is to develop a good stacking method with an automatic material handling system and the procedures that can increase productivity, reduce production costs, and prevent labor injury. A diversity of products leads to a number of different kinds of stacking problems. Much research has been done focusing on two-dimensional arrangement for rectangles, circles or irregular shapes, and threedimensional regular-shaped objects such as rectangular boxes. To solve stacking problems, many algorithms such as the genetic algorithm, simulated annealing and other heuristic algorithms have been proposed. The three-dimensional stacking problem has a practical application in the transportation, manufacturing, and construction industries. There has been relatively little emphasis on three-dimensional irregular objects; however, stacking three-dimensional irregular objects has become more common in industry. In this thesis research, three heuristic algorithms are proposed to stack irregular stone pieces nested in a container with multiple layers. Primary functions of the heuristic algorithms include three major parts. First, it approximates irregular shapes to a cluster of straight lines. Secondly, it arranges the approximated angles one-by-one with the proposed step-by-step rule. Finally, it considers the weight of the stone pieces from the pixel calculation for reasons of stability. The first and second algorithms are based on the area and angle of the stone piece and the third one is based on the approximated weight of the stone. An automatic real-time stacking system including pneumatic devices, sensors, relays, a conveyor, a programmable logic controller, a robotic arm, and a vision system was developed for this study. The algorithms developed were tested by this automatic stacking system for better utilization. Three performance measures were presented in the experimental result. Comparisons between the results from three proposed algorithms and that from the bottom-back-left algorithm are made. Experimental data demonstrate that the utilizations and the stabilities of the three proposed algorithms are statistically better than that of the bottom-back-left algorithm. However, the cycle times of the three proposed algorithms have no statistical difference from that of the bottom-back-left algorithm. In addition, a statistical test between each proposed algorithm is also conducted. Both the utilizations and stabilities have statistical differences between each proposed algorithm while the cycle times do not. The results of this study show that the algorithm developed works effectively for solving the stone-pieces stacking problem.