• Journal of Navigation and Port Research
• /
• v.48 no.4
• /
• pp.327-334
• /
• 2024

Amidst the burgeoning research on autonomous ships, there has been a marked increase in the development of new systems and equipment for these vessels. This study identifies specific assets and functional requirements essential for managing autonomous ships, and presents a system architecture applicable to conventional and autonomous vessels alike. The recommended fleet asset management system encompasses ship information, maintenance, crew, and operational management. Assets specific to autonomous ships include the MASS rating within ship information, advanced software and hardware for maintenance, and remote operators in the crew management. These asset functions are managed through separate, modular systems that permit easy attachment and detachment, hence ensuring compatibility with both traditional and autonomous ships. Moreover, the system is web-based which facilitates access through an internet environment without necessitating the installation of any program.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.12
• /
• pp.5669-5684
• /
• 2018

The new mobile edge network architecture has been required for an increasing amount of traffic, quality requirements, advanced driver assistance system for autonomous driving and new cloud computing demands on highway. This article proposes a hierarchical cloud computing architecture to enhance performance by using adaptive data load distribution for buses that play the role of edge computing server. A vehicular dynamic cloud is based on wireless architecture including Wireless Local Area Network and Long Term Evolution Advanced communication is used for data transmission between moving buses and cars. The main advantages of the proposed architecture include both a reduction of data loading for top layer cloud server and effective data distribution on traffic jam highway where moving vehicles require video on demand (VOD) services from server. Through the description of real environment based on NS-2 network simulation, we conducted experiments to validate the proposed new architecture. Moreover, we show the feasibility and effectiveness for the connected car media service on highway.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.51 no.5
• /
• pp.182-189
• /
• 2002

This paper presents a framework of hybrid dynamic control systems for the motion control of wheeled mobile robot systems with nonholonomic constraints. The hybrid control system has the 3-layered hierarchical structure: digital automata for the higher process, mobile robot system for the lower process, and the interface as the interaction process between the continuous dynamics and the discrete dynamics. In the hybrid control architecture of mobile robot, the continuous dynamics of mobile robots are modeled by the switched systems. The abstract model and digital automata for the motion control are developed. In high level, the discrete states are defined by using the sensor-based search windows and the reference motions of a mobile robot in low level are specified in the abstracted motions. The mobile robots can perform both the motion planning and autonomous maneuvering with obstacle avoidance in indoor navigation problem. Simulation and experimental results show that hybrid system approach is an effective method for the autonomous maneuvering in indoor environments

• Journal of the Institute of Convergence Signal Processing
• /
• v.9 no.3
• /
• pp.230-237
• /
• 2008

Recently, lots of interest and competition in developments related to the autonomous-vehicle robot are being further increased. However, the absence of the standard architectures for effectively controlling the autonomous-vehicle robot led to many difficulties such as the long duration of development and the uncompatibility with other autonomous-vehicle robots. Accordingly, we implemented a mobile autonomous-vehicle robot system based on JAUS standard architecture. The mobile robot communicates with the remote-control system by using wireless LAN UDP/IP JAUS command massages. Its effectiveness is showed through the experimental results related to the navigation of implemented robot.

• ETRI Journal
• /
• v.44 no.6
• /
• pp.1004-1019
• /
• 2022

For autonomous driving without high-definition maps, we present a model capable of generating multiple plausible paths from egocentric images for autonomous vehicles. Our generative model comprises two neural networks: feature extraction network (FEN) and path generation network (PGN). The FEN extracts meaningful features from an egocentric image, whereas the PGN generates multiple paths from the features, given a driving intention and speed. To ensure that the paths generated are plausible and consistent with the intention, we introduce an attentive discriminator and train it with the PGN under a generative adversarial network framework. Furthermore, we devise an interaction model between the positions in the paths and the intentions hidden in the positions and design a novel PGN architecture that reflects the interaction model for improving the accuracy and diversity of the generated paths. Finally, we introduce ETRIDriving, a dataset for autonomous driving, in which the recorded sensor data are labeled with discrete high-level driving actions, and demonstrate the state-of-the-art performance of the proposed model on ETRIDriving in terms of accuracy and diversity.

• Journal of information and communication convergence engineering
• /
• v.21 no.4
• /
• pp.346-350
• /
• 2023

Trajectory prediction is an essential element for driving autonomous vehicles, and various trajectory prediction models have emerged with the development of deep learning technology. Convolutional neural network (CNN) is the most commonly used neural network architecture for extracting the features of visual images, and the latest models exhibit high performances. This study was conducted to identify an efficient CNN backbone model among the components of deep learning models for trajectory prediction. We changed the existing CNN backbone network of multiple-trajectory prediction models used as feature extractors to various state-of-the-art CNN models. The experiment was conducted using nuScenes, which is a dataset used for the development of autonomous vehicles. The results of each model were compared using frequently used evaluation metrics for trajectory prediction. Analyzing the impact of the backbone can improve the performance of the trajectory prediction task. Investigating the influence of the backbone on multiple deep learning models can be a future challenge.

• Journal of the Society of Naval Architects of Korea
• /
• v.61 no.2
• /
• pp.98-105
• /
• 2024

Due to the low accuracy of measured data obtained from low-cost GNSS and IMU devices, it was hard to secure the required accuracy of the measured position and heading angle for autonomous navigation which was conducted by a model-scale marine mobility. In this paper, a localization technique using the Extended Kalman Filter (EKF) is proposed for coping with the issue. First of all, a position and heading angle estimator is developed using EKF with the assumption of a point mass model. Second, the measured data from GNSS and IMU, including position, heading angle, and velocity are used for the estimator. In addition, the heading angle is additionally obtained by comparing the LiDAR point cloud with map information for a temporal water tank. The newly acquired heading angle is integrated into the estimator as an additional measurement to correct the inaccuracy in the heading angle measured from the IMU. The effectiveness of the proposed approach is investigated using data acquired from preliminary tests of the model-scale autonomous marine mobility.

• 제어로봇시스템학회:학술대회논문집
• /
• 1991.10b
• /
• pp.1585-1590
• /
• 1991

This paper presents the principles for design of autonomous systems whose behavior is based on models that support the various tasks that must be performed. We propose a model-based architecture aimed at reducing the computational demands required to integrate high level symbolic models with low level dynamic models. Model construction methods are illustrated to outfit such an architecture with the models needed to meet given objectives.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.30 no.11 s.254
• /
• pp.1348-1357
• /
• 2006

In a hybrid three-layer control architecture(deliberative, sequencing, and reflexive), the lowest reflexive layer consists of resources, actions, an action coordinator, and motion controllers. Because the execution of individual components in the reflexive layer should be done in real-time, each component has to be simple and, due to this reason, the Linux-RTAI(Real-Time Application Interface for Linux) has been used as an operating system. In this paper, a navigation control architecture, which combines the components in the reflexive layer and the navigation-related modules in the sequencing layer, is proposed. And then, as basic components, four actions(Goto, Avoid, Move, and EmergencyStop) are designed. Experimental results confirm the effectiveness of the proposed architecture and the performance of individual associated actions.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2001.04a
• /
• pp.393-397
• /
• 2001

This paper addresses an Intelligent Agent-based Open Architecture Cell Controller for Intelligent Manufacturing System(IMS). With an Intelligent Agent approach, the IMS will be a independent, autonomous, distributed system and achieve a adaptability to change of manufacturing environment. As the development methodology of Open Architecture Cell Controller, an object-oriented modeling technique is employed for building models associated with IMS operation, such as resource model, product model, and control model. Intelligent Agent-based Open Architecture Cell Controller consists of two kinds of dependant agents, that are the active agent and the coordinator agent. The Active agent is contributed to control components of IMS in real-time. The coordinator agent has great role in scheduling and planning of IMS. It communicates with other active agents to get information about status on system and generates the next optimal task through the making-decision logic and dispatch it to other active agent.