PERANCANGAN SISTEM WAREHOUSE BERBASIS TEKNOLOGI OCR UNTUK MENINGKATKAN EFEKTIVITAS DAN EFISIENSI

Abstract

The rapid development of information technology is massive worldwide, racing to create solutions as a form of digital supply chain movement, massive use of information technology is also being carried out in the logistics business world which is part of supply chain activities, requiring accurate and fast information is a very important area to guarantee delivery of goods on time and on target. One important aspect of the logistics process is effective warehouse management. One of the main challenges in effective warehouse management is the process of receiving goods in the warehouse which still causes errors such as writing errors, inaccurate recording of the quantity, type and name of goods received and removed from the warehouse which takes quite a long time to process. There has been a lot of research carried out on data capturing activities with the application of OCR but not much has been found that focuses on managing printed and handwritten physical documents in the warehouse area, and preparation of data bases on websites connected to systematic OCR technology, and there is no analysis of technology investment. OCR regarding cost efficiency. Therefore, to help overcome this problem, this research carried out the design and implementation of a warehouse system based on OCR (Optical Character Recognition) technology with the name WareOCR and analyzed it from an investment perspective using the ROI method. From the research results obtained using the OCR creation model using the CNN method, the accuracy of the WareOCR system on test data is 93%, Return on Investment is calculated at 60.38%, in the NPV calculation in year 3 the figure is IDR 112,911,033 where NPV > 0 for the project implementing WareOCR it is acceptable that this investment provides added value of 493% of the total project cost. This research  also  has   research   limitations  in  the  form of  invoice  formats  which  need  to  be standardized so they can be read with the WareOCR tool and the advantages of this research are that they can shorten the working time of the data entry warehouse in inputting information into the database.