Deep reinforcement learning (DRL), a vital branch of artificial intelligence, has shown great promise in mobile robot navigation within dynamic environments. However, existing studies mainly focus on simplified dynamic scenarios or the modeling of static environments, which results in trained models lacking sufficient generalization and adaptability when faced with real-world dynamic environments, particularly in handling complex task variations, dynamic obstacle interference, and multimodal data fusion. Addressing these gaps is essential for enhancing its real-time performance and versatility. Through a comparative analysis of classical DRL algorithms, this study highlights their advantages and limitations in handling real-time navigation tasks under dynamic environmental conditions. In particular, the paper systematically examines value-based, policy-based, and hybrid-based DRL methods, discussing their applicability to different navigation challenges. Additionally, by reviewing recent studies from 2021 to 2024, it identifies key trends in DRL-based navigation, revealing a strong focus on indoor environments while outdoor navigation and multi-robot collaboration remain underexplored. The analysis also highlights challenges in real-world deployment, particularly in sim-to-real transfer and sensor fusion. Based on these findings, this paper outlines future directions to enhance real-time adaptability, multimodal perception, and collaborative learning frameworks, providing theoretical and technical insights for advancing DRL in dynamic environments.

Traditional approaches to monitoring water quality in aquaculture tanks present numerous limitations, including the inability to provide real-time data, which can lead to improper feeding practices, reduced productivity, and potential environmental risks. To address these challenges, this study aimed to create an accurate water quality monitoring system for Asian seabass fish farming in aquaculture tanks. This was achieved by enhancing the accuracy of low-cost sensors using simple linear regression and validating the IoT system data with YSI Professional Pro. The system's development and validation were conducted over three months, employing professional devices for accuracy assessment. The accuracy of low-cost sensors was significantly improved through simple linear regression. The results demonstrated impressive accuracy levels ranging from 76% to 97%. The relative error values which range from 0.27% to 4% demonstrate a smaller range compared to the values obtained from the YSI probe during the validation process, signifying the enhanced accuracy and reliability of the IoT sensor by using simple linear regression. The system's enhanced accuracy facilitates convenient and reliable real-time water quality monitoring for aquafarmers. Real-time data visualization was achieved through a microcontroller, Thingspeak, Virtuino application, and ESP 8266 Wi-Fi module, providing comprehensive insights into water quality conditions. Overall, this adaptable tool holds promise for accurate water quality management in diverse aquatic farming practices, ultimately leading to improved yields and sustainability.

The oil palm industry is vital for the Malaysian economy. However, it is threatened by the Ganoderma boninense fungus, which causes basal stem rot (BSR) disease. Foliar symptoms of the disease include the appearance of several unopened spears, flat crowns, and small crown size. The effect of this disease depends on the severity of the infection. Currently, the disease can be detected manually by analyzing the oil palm tree’s physical structure. Terrestrial laser scanning (TLS) is an active ranging method that uses laser light, which can directly represent the tree’s external structure. This study aimed to classify the healthiness levels of the BSR disease using a machine learning (ML) approach. A total of 80 oil palm trees with four different healthiness levels were pre-determined by the experts during data collection with 40 each for training and testing. The four healthiness levels are T0 (healthy), T1 (mildly infected), T2 (moderately infected), and T3 (severely infected), with 10 trees in each level. A terrestrial scanner was mounted at a height of 1 m, and each oil palm was scanned at four positions at a distance of 1.5 m around the tree. Five tree features were extracted from the TLS data: C200 (crown slice at 200 cm from the top), C850 (crown slice at 850 cm from the top), crown area (number of pixels inside the crown), frond angle, and frond number. C200 and C850 were obtained using the crown stratification method, while the other three features were obtained from the top-down image. The obtained features were then analyzed by principal component analysis (PCA) to reduce the dimensionality of the dataset and increase its interpretability while at the same time minimizing information loss. The results showed that the kernel naïve Bayes (KNB) model developed using the input parameters of the principal components (PCs) 1 and 2 had the best performance among 90 other models with a multiple level accuracy of 85% and a Kappa coefficient of 0.80. Furthermore, the combination of the two highest PC variance with the most weighted to frond number, frond angle, crown area, and C200 significantly contributed to the classification success. The model also could classify healthy and mildly infected trees with 100% accuracy. Therefore, it can be concluded that the ML approach using TLS data can be used to predict early BSR infection with high accuracy.

A combine harvester has been widely employed for harvesting paddy in Malaysia. However, it is one of the most challenging machines to operate when harvesting grain crops. Improper handling of a combine harvester can lead to a significant amount of grain loss. Any losses during the harvesting process would result in less income for the farmers. Grain loss sensing technology is automated, remote, and prospective. It can help reduce grain losses by increasing harvesting precision, reliability, and productivity. Monitoring and generating real-time sensor data can provide effective combine harvester performance and information that will aid in analyzing and optimizing the harvesting process. Thus, this paper presents an overview of the conventional methods of grain loss measurements, the factors that contribute to grain losses, and further reviews the development and operation of sensor components for monitoring grain loss during harvest. The potential and limitations of the present grain loss monitoring systems used in combine harvesting operations are also critically analyzed. Several strategies for the adoption of the technology in Malaysia are also highlighted. The use of this technology in future harvesting methods is promising as it could lead to an increase in production, yield, and self-sufficiency to meet the increasing demand for food globally.

Insect detection and counting constitute a considerable challenge in the field of agriculture. However, among various biotic issues of agricultural production, pest infestation is a major challenge, with the humid environment surrounding the crops encouraging the survival and proliferation of pests. In addition, electronic traps need protection, especially from rain. This study describes the design and development of a prototype for an automatic pest sampling and detection system for agricultural crops. To the best of our knowledge, the proposed system is the first motorized automatic trap developed to handle monitoring operations in two directions of precise movement (i.e., clockwise and counterclockwise) and cover four directions of insect sampling that provides additional details of insect infestation direction. A square-shaped sticky box was designed, and an optical sensor was attached to a scalable arm. The movements of the sticky box and the camera arm were generated by motors. Preprocessing was conducted by using morphological operations, whereas insect detection and counting were implemented by an algorithm of connected components labeling that applied by using MATLAB image processing toolbox. Different kernel functions, such as disk, diamond, square, and sphere, were used as matching functions for the insect detection and counting algorithm. The average accuracy of the highest sphere kernel was 85.2%. Test results of the hardware show the reliability, flexibility, and system protection of the automatic system to provide accurate movements in two degrees of freedom.