A packaging company is designing reusable multi-folding plastic box with RFID function. They want to identify material flow and inventory status of multi-folding box by adapting RFID technology and reallocate the boxes to required distribution centers. This reusable box may contain various types of materials such like liquids, metals, farm products and so on. Recently, RFID technology has been widely applied in the industrial fields, especially in logistics. However, RFID technology has some problems. One of the problems is that it doesn't guarantee the RFID recognition with environment of metals and liquids. We cannot overlook this fact because the contents of box is not fixed and may contain various kinds of materials. The purpose of this research is to analyze the influence of unspecified contents on the RFID detectability of reusable multi-folding box and predict the reading rate on various conditions. At first, we selected a list of test materials from expert interview and literature survey and performed experimental design to investigate influence by materials. Then, we built a prediction model using support vector machine (SVM) to predict reading rate on various conditions. The proposed model is based on intelligent machine learning algorithm. It gave a high accuracy of prediction, and provided robust combinations of conditions as to detectability on various contents. Finally, we performed sensitivity analysis about marginal detectability of influencing factors and found out tolerable level of factors.