The expansion of online retail markets has driven the development of personalized product recommendation services leveraging platform-based product and customer data. Large retailers have implemented seller-oriented recommendation systems, where AI analyzes POS sales data to identify similar stores and recommend products not yet introduced but successful elsewhere. However, small and medium-sized retailers face challenges in adapting to rapidly evolving online market trends due to limited resources. This study proposes a recommendation algorithm tailored for small-scale retailers using sales data from an online shopping mall. We analyzed 600,000 transaction records from 13,607 sellers and 95,938 products, focusing on Beauty Supplies, Kitchenware, and Cleaning Supplies categories. Three algorithms—Attentional Factorization Machines (AFM), Deep Factorization Machines (DeepFM), and Neural Collaborative Filtering (NCF)—were applied to recommend top 10% weekly sales items, with an ensemble model integrating their strengths. To address class imbalance, the Synthetic Minority Oversampling Technique (SMOTE) was employed, and performance was evaluated using AUC, Accuracy, Precision, and Recall metrics on separate training and test datasets. The ensemble model outperformed individual models across all metrics, while DeepFM excelled in Precision. These findings demonstrate that ensemble-based recommendation algorithms enhance recommendation accuracy for suppliers in large-scale online retail environments, offering practical implications for small-scale retailers.

Vertical takeoff and landing (VTOL) drones are increasingly recognized as an important solution for last-mile delivery in the food and beverage sector, owing to their rapid deployment capabilities and high operational flexibility. In particular, growing interest in drone delivery services has been observed among fast food and coffee franchises, where rapid delivery is essential due to the time-sensitive nature of food and beverage items intended for immediate consumption. Despite this trend, there remains a lack of research on the structural modeling of flight routes for VTOL drones operating under automatic flight conditions, and on the implementation of first-come-first-served (FCFS) delivery services utilizing predefined flight routes. Accordingly, this study comprehensively describes the operations for food and beverage delivery services using VTOL drones. In particular, it addressed the use of multiple drones to conduct FCFS-type multi-point delivery services along fixed routes suitable for automatic flight.

Anomaly detection technique for the Unmanned Aerial Vehicles (UAVs) is one of the important techniques for ensuring airframe stability. There have been many researches on anomaly detection techniques using deep learning. However, most of research on the anomaly detection techniques are not consider the limited computational processing power and available energy of UAVs. Deep learning model convert to the model compression has significant advantages in terms of computational and energy efficiency for machine learning and deep learning. Therefore, this paper suggests a real-time anomaly detection model for the UAVs, achieved through model compression. The suggested anomaly detection model has three main layers which are a convolutional neural network (CNN) layer, a long short-term memory model (LSTM) layer, and an autoencoder (AE) layer. The suggested anomaly detection model undergoes model compression to increase computational efficiency. The model compression has same level of accuracy to that of the original model while reducing computational processing time of the UAVs. The proposed model can increase the stability of UAVs from a software perspective and is expected to contribute to improving UAVs efficiency through increased available computational capacity from a hardware perspective.

Vertical takeoff and landing (VTOL) is a core feature of unmanned aerial vehicles (UAVs), which are commonly referred to as drones. In emerging smart logistics, drones are expected to play an increasingly important role as mobile platforms. Therefore, research on last-mile delivery using drones is on the rise. There is a growing trend toward providing drone delivery services, particularly among retailers that handle small and lightweight items. However, there is still a lack of research on a structural definition of the VTOL drone flight model for multi-point delivery service. This paper describes a VTOL drone flight route structure for a multi-drone delivery service using rotary-wing type VTOL drones. First, we briefly explore the factors to be considered when providing drone delivery services. Second, a VTOL drone flight route model is introduced using the idea of the nested graph. Based on the proposed model, we describe various time-related attributes for delivery services using drones and present corresponding calculation methods. Additionally, as an application of the drone route model and the time attributes, we comprehensively describe a simple example of the multi-drone delivery for first-come-first-served (FCFS) services.

Distribution and logistics industries contribute some of the biggest GDP(gross domestic product) in South Korea and the number of related companies are quarter of the total number of industries in the country. The number of retail tech companies are quickly increased due to the acceleration of the online and untact shopping trend. Furthermore, major distribution and logistics companies try to achieve integrated data management with the fulfillment process. In contrast, small and medium distribution companies still lack of the capacity and ability to develop digital innovation and smartization. Therefore, in this paper, a deep learning-based demand forecasting & recommendation model is proposed to improve business competitiveness. The proposed model is developed based on real sales transaction data to predict future demand for each product. The proposed model consists of six deep learning models, which are MLP(multi-layers perception), CNN(convolution neural network), RNN(recurrent neural network), LSTM(long short term memory), Conv1D-BiLSTM(convolution-long short term memory) for demand forecasting and collaborative filtering for the recommendation. Each model provides the best prediction result for each product and recommendation model can recommend best sales product among companies own sales list as well as competitor’s item list. The proposed demand forecasting model is expected to improve the competitiveness of the small and medium-sized distribution and logistics industry.