Offshore operation causes the dynamic motion of offshore cranes and payload by the ocean environment. The motion of the payload lowers the safety and efficiency of the work, which may increase the working time or cause accidents. Therefore, we design a control method for the crane using artificial intelligence (AI) to minimize the heave motion of the payload. Herein, reinforcement learning (RL), which calculates actions according to states, is applied. Furthermore, the deep deterministic policy gradient (DDPG) algorithm is used because the actions need to be determined in a continuous state. In the DDPG algorithm, the state is defined as the motion of the crane and speed of the wire rope, and the action is defined as the speed of the wire rope. In addition, the reward is calculated using the motion of the payload. In this study, the heave motion of the payload was reduced by developing an agent suitable for adjusting the length of the wire rope. The heave motion of the payload was compared in the non-learning condition of the RL-based and proportional integral differential (PID) controls, and an average reduction rate of 30% compared to PID control was confirmed. The RL-based control performed better than the PID control under learned conditions.

No abstract

In this study, the coupled motion of a mooring system and multifloating cranes were analyzed. For the motion analysis, the combined equations of motions of the mooring line and multifloating cranes were introduced. The multibody equations for floating cranes were derived from the equations of motion. The finite element method (FEM) was used to derive equations to solve the stretchable catenary problem of the mooring line. To verify the function of mooring simulator, calculation results were compared with commercial mooring software. To validate the analysis results, we conducted an experimental test for offshore operation using two floating crane models scaled to 1:40. Two floating crane models and a pile model were established for operation of uprighting flare towers. During the model test, the motion of the floating cranes and tensions of the mooring lines were measured. Through the model test, the accuracy of the mooring analysis program developed in this study was verified. Therefore, if this mooring analysis program is used, it will be possible to perform a mooring analysis simulation at the same time as a maritime work simulation.

The financial business process worldwide suffers from huge dependencies upon labor and written documents, thus making it tedious and time-consuming. In order to solve this problem, traditional robotic process automation (RPA) has recently been developed into a hyper-automation solution by combining computer vision (CV) and natural language processing (NLP) methods. These solutions are capable of image analysis, such as key information extraction and document classification. However, they could improve on text-rich document images and require much training data for processing multilingual documents. This study proposes a multimodal approach-based intelligent document processing framework that combines a pre-trained deep learning model with traditional RPA used in banks to automate business processes from real-world financial document images. The proposed framework can perform classification and key information extraction on a small amount of training data and analyze multilingual documents. In order to evaluate the effectiveness of the proposed framework, extensive experiments were conducted using Korean financial document images. The experimental results show the superiority of the multimodal approach for understanding financial documents and demonstrate that adequate labeling can improve performance by up to about 15%.

No abstract