The application of low cost intelligent automation to assembly operations using a discrete event simulation

Abstract

This report proposes using a discrete event simulation approach to investigate the application and integration of Low Cost Intelligent Automation (LCIA) at a local automotive manufacturer located in Rosslyn, Pretoria. Low cost automation solutions are implemented in the form of Automated Guided Vehicles (AGVs) which are used as material handling mechanisms. Critical factors pertaining to the design and operation of these systems are the guide-path layout, the number of AGVs required and the vehicle dispatch rules. Literature on the subject matter showed that these design problems can be solved by using analytical models or simulation models. AGV systems are large and complex with various interrelating components that make up the complete system. Therefore, simulation modelling is chosen as the tool for this application as it is capable of handling these complex systems. The three main issues pertaining to the AGV system design are presented and analysed in this report. The conventional and tandem guide paths are analysed together with the workload-based Maximum Remaining Outgoing Queue Size (MROQS) dispatch rule and time-base Modified First Come First Serve (MFCFS) dispatch rule. These guide paths and dispatch rules are combined in four scenario models, each focussing on a specific combination of a guide path and dispatch rule. For each of the scenario models, the number of AGVs are varied and the changes in system performance are observed and documented. These scenario models are developed with discrete event simulation models, and are evaluated based on key performance indicators. The specific performance indicators comprised of the total throughput of the AGV system, the utilisation of AGVs and the buffer sizes at three stations. From the scenarios it is found that both the conventional and tandem guide path have similar results for both dispatch rules. Of the two guide paths, the conventional is more sensitive to the changes in dispatch rules. It is concluded that the best performing system is the tandem guide path with a MFCFS dispatch rule and a total of three AGVs in the system. It provided, on average a total output of 135 front and rear end bumpers. The AGVs in the loop are also fully utilised with a high utilisation of 85%. The proposes system also reduced the total walking distance by 37% and streamlines the material supply process to the station by eliminating unnecessary motions.