Intelligent production requires improved data analytics and better technological possibilities to improve system performance and decision making. With the widespread use of the machine learning process, a growing need has arisen for processing extensive production data, equipped with high volumes, high speed, and high diversity. At this point, deep learning provides advanced analysis tools for processing and analyzing extensive production data. The deep convolutional neural network (DCNN) displays state-of-the-art performance on many grounds, including metal manufacturing surface defect detection. However, there is still space for improving the defect detection performance over generic DCNN models. The proposed approach performed better than the associated methods in the particular area of surface crack detection. The defect zones of disjointed results are classified into their unique classes by a DCNN. The experimental outcomes prove that this method meets the durability and efficiency requirements for metallic object defect detection. In time, it can also be extended to other detection methods. At the same time, the study will increase the accuracy quality of the features that can make a difference in the deep learning method for the detection of surface defects.

SummaryComplete coverage planning (CCP) is a task to cover the entire area on the map, according to the job description of the autonomous mobile robot. The most widely used method for CCP in the literature is the grid‐based coverage method. In this method, the problem is processing the partially filled cell as completely filled, which reduces the coverage performance. The ability to use the clustering method, which will be created by considering the characteristics of the environment, was determined as a research question to solve this problem. In this direction, it is aimed to use K‐means++ algorithm, which is a widely used clustering algorithm and segmentation technique. In this context, an offline K‐means++ complete coverage planning (Km++CCP) method, in which the navigable area on the map of the indoor where a mobile robot will navigate is clustered using the K‐means++ algorithm and the centroids can be used as waypoints, is proposed. To test the proposed method, 2 simulations and 36 real‐world experiments were conducted. The indoor coverage ratio of Km++CCP was calculated higher than the grid‐based method in all experiments.

An autonomous mobile robot needs a map of the environment and location information relative to the map. Simultaneous Localization and Mapping (SLAM) is a prediction process in which the autonomous mobile robot can use this map to determine its position while building a consistent map. The purpose of this study is to examine the effect of geometric objects on SLAM performance. In this direction, three different experimental areas including equilateral triangular prisms, square prisms and cylinders are designed in Gazebo. The fourth experiment area includes all three geometric objects used in the study. When the mapping times of the four experimental areas were compared, it was seen that the fastest scenario is achieved within triangular-only objects (9 min 55 sec) and the slowest within square (10 min 43 sec). In terms of measures, the generated map including the triangular prisms is the closest to the actual measures of the simulated area. Accordingly, the mapping error was calculated as 0.171 m 2 per 1 m 2 in an interior made of triangular prisms, and 0.682 m 2 in an interior made of square prisms. The obtained results show that the shapes of the geometric objects directly affect the performance of SLAM.