Autonomy is the core capability of future systems, and architecture design is one of the critical issues in system development and implementation. To discuss the architecture of autonomous systems in the future, this paper reviews the developing progress of architectures from automation systems to autonomous systems. Firstly, the autonomy and autonomous systems in different fields are summarized. The article classifies and summarizes the architecture of typical automated systems and infer three suggestions for building an autonomous system architecture: extensibility, evolvability, and collaborability. Accordingly, this paper builds an autonomous waterborne transportation system, and the architecture is composed of the object layer, cyberspace layer, cognition layer, and application layer, the proposed suggestions made in the construction of the architecture are reflected in the inter-relationships at all layers. Through the cooperation of four layers, the autonomous waterborne transportation system can autonomously complete the system functions, such as system control and transportation service. In the end, the characteristics of autonomous systems are concluded, from which the future primary research directions and the challenges of autonomous systems are provided.

Robust detection and recognition of water surfaces are critical for autonomous navigation of unmanned surface vehicles (USVs), since any none-water region is likely an obstacle posing a potential danger to the sailing vehicle. A novel water region visual detection method is proposed in this paper. First, the input image pixels are clustered into different regions and each pixel is assigned a label tag and a confidence value by adaptive multistage segmentation algorithm. Then the resulting label map and associated confidence map are fed into a convolutional neural network (CNN) as training samples to train the network online. Finally, the online trained CNN is used to segment the input image again but with greater precision and stronger robustness. Compared with other deep-learning image segmentation algorithms, the proposed method has two advantages. Firstly, it dispenses with the need of manual labeling training samples which is a costly and painful task. Secondly, it allows real-time online training for CNN, making the network adaptive to the navigational environment. Another contribution of this work relates to the training process of neuro network. An effective network training method is designed to learn from the imperfect training data. We present the experiments in the lake with a various scene and demonstrate that our proposed method could be applied to recognize the water region in the unknown navigation environment automatically.

Precipitation serves as a crucial factor in the study of hydrometeorology, ecology, and the atmosphere. Gridded precipitation data are available from a multitude of sources including precipitation retrieved by satellites, radar, the output of numerical weather prediction models, and extrapolation by ground rain gauge data. Evaluating different types of products in ungauged regions with complex terrain will not only help researchers in applying scientific data, but also provide useful information that can be used to improve gridded precipitation products. The present study aims to evaluate comprehensively 12 precipitation datasets made by raw retrieved products, blended with rain gauge data, and blended multiple source datasets in multi-temporal scales in order to develop a suitable method for creating gridded precipitation data in regions with snow-dominated regions with complex terrain. The results show that the Multi-Source Weighted-Ensemble Precipitation (MSWEP), Global Satellite Mapping of Precipitation with Gauge Adjusted (GSMaP_GAUGE), Tropical Rainfall Measuring Mission (TRMM_3B42), Climate Prediction Center Morphing Technique blended with Chinese observations (CMORPH_SUN), and Climate Hazards Group Infrared Precipitation with Stations (CHIRPS) can represent the spatial pattern of precipitation in arid/semi-arid and humid/semi-humid areas of the Qinghai-Tibet Plateau on a climatological spatial pattern. On interannual, seasonal, and monthly scales, the TRMM_3B42, GSMaP_GAUGE, CMORPH_SUN, and MSWEP outperformed the other products. In general, the Precipitation Estimation from Remotely Sensed Information using Artificial Neural Networks-Cloud Classification System (PERSIANN_CCS) has poor performance in basins of the Qinghai-Tibet Plateau. Most products overestimated the extreme indices of the 99^th percentile of precipitation (R₉₉), the maximal of daily precipitation in a year (Rmax), and the maximal of pentad accumulation of precipitation in a year (R_5dmax). They were underestimated by the extreme index of the total number of days with daily precipitation less than 1 mm (dry day, DD). Compared to products blended with rain gauge data only, MSWEP blended with more data sources, and outperformed the other products. Therefore, multi-sources of blended precipitation should be the hotspot of regional and global precipitation research in the future.

In order to monitor and manage vessels in channels effectively, identification and tracking are very necessary. This work developed a maritime unmanned aerial vehicle (Mar-UAV) system equipped with a high-resolution camera and an Automatic Identification System (AIS). A multi-feature and multi-level matching algorithm using the spatiotemporal characteristics of aerial images and AIS information was proposed to detect and identify field vessels. Specifically, multi-feature information, including position, scale, heading, speed, etc., are used to match between real-time image and AIS message. Additionally, the matching algorithm is divided into two levels, point matching and trajectory matching, for the accurate identification of surface vessels. Through such a matching algorithm, the Mar-UAV system is able to automatically identify the vessel’s vision, which improves the autonomy of the UAV in maritime tasks. The multi-feature and multi-level matching algorithm has been employed for the developed Mar-UAV system, and some field experiments have been implemented in the Yangzi River. The results indicated that the proposed matching algorithm and the Mar-UAV system are very significant for achieving autonomous maritime supervision.

The prediction of ship location has become an increasingly popular research hotspot in the field of maritime transportation engineering, which benefits maritime safety supervision and security. Existing methods of ship location prediction based on motion characteristics have a large uncertainty and cannot guarantee trajectory prediction accuracy of the target ship. An improved method of location prediction using k-nearest neighbor (KNN) is proposed in this paper. An expanded circle area of the latest point of the target ship is first generated to find the reference points with similar movement characteristics in the constraints of distance and time intervals. Then, the top k-nearest neighbors are determined based on the degree of similarity. Relationships between the reference point of each neighbor and the latest points of the target ship are calculated. The predicted location of the target ship can then be determined by a weighted calculation of the locations of all neighbors at the predicted time and their relationships with the target ship. Experiments of ship location prediction in 10 min, 20 min, and 30 min were conducted. The correlation coefficient of the location prediction error for the three experiments was 0.992, 0.99, and 0.9875, respectively. The results show that ship location prediction with reference to multiple nearest neighbors with similar movements can provide better accuracy.

The intelligentization of unmanned surface vehicles (USVs) has recently attracted intensive interest. Visual perception of the water scenes is critical for the autonomous navigation of USVs. In this paper, an adaptive semantic segmentation method is proposed to recognize the water scenes. A semantic segmentation network model is designed to classify each pixel of an image into water, land or sky. The segmentation result is refined by the conditional random field (CRF) method. It is further improved accordingly by referring to the superpixel map. A weight map is generated based on the prediction confidence. The network trains itself with the refined pseudo label and the weight map. A set of experiments were designed to evaluate the proposed method. The experimental results show that the proposed method exhibits excellent performance with few-shot learning and is quite adaptable to a new environment, very efficient for limited manual labeled data utilization.