As oil prices continue to rise internationally, shipping costs are also increasing rapidly. In order to reduce fuel costs, an economical shipping route must be determined by accurately predicting the estimated arrival time of ships. A common method in the evaluation of ship speed involves computing the total resistance of a ship using theoretical analysis; however, using theoretical equations cannot be applied for most ships under various operating conditions. In this study, a machine learning approach was proposed to predict ship speed over the ground using the automatic identification system (AIS) and noon-report maritime weather data. To train and validate the developed model, the AIS and marine weather data of the seventy-six vessels for a period one year were used. The model accuracy result shows that the proposed data-driven model has a satisfactory capability to predict the ship speed based on the chosen features.

With soaring oil prices worldwide, determining the most optimal routes for economical ship operation has become an important issue. Optimizing ship routes is economically important for ship operation, but it is also essential to meet the standards of environmental regulations recently imposed by the International Maritime Organization. For this purpose, various algorithms for determining ship routes have been developed to ensure the economical operation of ships via utilization of marine climate data and Automatic Identification System (AIS) data. However, such algorithms require a large amount of computational time and do not provide optimal routes because they do not consider practical operating conditions, such as weather and ocean conditions. In this study, an improved A* algorithm using AIS and weather data is proposed to overcome the limitation of the original A* algorithm, one of the most widely used path-finding algorithms. The improved A* algorithm uses an adaptive grid system that efficiently explores nodes according to map grid deformation by latitude. It finds economical routes by minimizing the estimated time of arrival generated by machine learning through 16-way node exploration. For verification of the proposed method, the original A* algorithm and improved A* algorithm were compared through a case study.

The human foot is easily deformed owing to the innate form of the foot or an incorrect walking posture. Foot deformations not only pose a threat to foot health but also cause fatigue and pain when walking; therefore, accurate diagnoses of foot deformations are required. However, the measurement of foot deformities requires specialized personnel, and the objectivity of the diagnosis may be insufficient for professional medical personnel to assess foot deformations. Thus, it is necessary to develop an objective foot deformation classification model. In this study, a model for classifying foot types is developed using image and numerical foot pressure data. Such heterogeneous data are used to generate a fine-tuned visual geometry group-16 (VGG16) and K−nearest neighbor (k-NN) models, respectively, and a stacking ensemble model is finally generated to improve accuracy and robustness by combining the two models. Through k-fold cross-validation, the accuracy and robustness of the proposed method have been verified by the mean and standard deviation of the f1 scores (0.9255 and 0.0042), which has superior performance compared to single models generated using only numerical or image data. Thus, the proposed model provides the objectivity of diagnosis for foot deformation, and can be used for analysis and design of foot healthcare products.