Divide Cleaning Area Method for Multi-cleaning Robots Using Manufacturing Execution Systems Based on Contamination-level Estimation

Abstract

The Agricultural Products processing Center (APC) in Korea performs various tasks such as pre-cooling, sorting, packaging, and storage during the production of agricultural goods. To address the issue of dust occurring during the production processes at the APC, cleaning robots are employed. These robots are equipped to recognize the APC environment and perform cleaning tasks according to a specified pattern. However, variations in dust levels may lead to differences in cleaning effectiveness. Manufacturing Execution Systems (MES) can record the flow of agricultural products during production processes; this information can be utilized to predict dust areas and levels. The present study proposes a method based on MES to efficiently address the issue of dust at the APC. By setting the locations of the equipment within the APC and their respective discharge capacities for contaminants, and by considering the discrepancy between the input and output quantities of raw materials, the dust levels at various locations can be predicted. Based on the prediction, specific cleaning areas can be assigned to multiple cleaning robots. The utilization of MES can thus significantly increase the efficiency of multiple cleaning robots, ranging from 3% to as high as 22%.