Volatile organic compounds (VOCs) as the components of atmospheric pollutants have been potentially harmful to the environment and human health. Swine production, one of the pillar livestock industries, is severely hampered by some air pollution issues, such as the emissions of odor and VOCs. The odor complaints were reported the most, accounting for 12.7% of the whole Chinese livestock industry in 2020, according to the Study "Analysis of National Odor Pollution Complaints in 2018-2020" from the National Ministry of Ecology and Environment. Odor pollution is considered as one of the major environment issues in the large-scale swine production, hindering the sustainable development of the industry. Therefore, it is critical to implement the odor pollution prevention and control measures in the large-scale swine production industry. However, the VOCs, components of the odorous pollutants, vary greatly in the organic molecules in the frequent present of the air, leading to difficulty in mitigation of odor. Generally, the VOCs are the significant precursors of ozone (O₃) and secondary aerosols (SOA). Once the concentration of VOCs exceeds a certain threshold value, a significant negative impact can be imposed on the environment, as well as human and animal health. Therefore, mitigation measures are essential to reduce the VOC emissions, and their negative impact on the environment and public health. In this review, VOCs emitted from the swine production was introduced from the aspects of characteristics and reduction potential. The types, characteristics, and sources of VOCs from pig farms were elucidated, together with the properties of odor-causing VOCs and detection. The factors that could impact the VOCs concentration were also illustrated, such as ventilation in barn, pig breed and manure cleaning. In addition, different testing methods and standards were selected for the variations in the composition of VOCs. The most common VOCs in swine production were found as the volatile fatty acids (VFA), olefins, alkanes, aromatic hydrocarbons, halogenated hydrocarbons, esters, ethers, phenols, indoles, organic sulfur, aldehydes, and ketones. Among them, the VFA, phenol, indole, and organosulfur were the most common odor-causing substances. The sources of VOCs were then analyzed for the efficient odor management. Manure, indoor air, and dust were the three main sources of VOCs in pig barns, which are considered as the primary risk of the environment and human health. The odor annoyance was found to be directly caused in the neighborhood, leading to the formation of ground-level ozone and secondary aerosols when exposed to the air. The detection methods of VOCs generally consisted of the discrete and continuous approaches. The typical VOCs measurement steps of discrete approaches were the sample collection, pre-treatment, and analysis. The commonly used continuous detection instrument was the proton-transfer-reaction time-of-flight mass spectrometry (PTR-TOF-MS). Finally, the primary VOCs treatment technologies were evaluated from three aspects of VOCs reduction in pig farms: source reduction, process control, and end treatment. Source reduction of VOCs, i.e., adjustments in the feed rations and the additives could improve the physiology and production performance of pigs. Process management of the VOCs reduction was achieved after adjustments in feeding and ventilation systems. The better manure cleaning practices were applied to limit the off-gassing from swine production. End treatment mainly included the VOCs treatment technologies of manure and ambient air treatment. Traditional VOCs treatment technologies were divided into two categories: recycling and degradation treatments. Photocatalytic oxidation, active material adsorption, low temperature plasma technology and degradation methods with the bio filters can be expected to gradually replace the traditional treatments and become the key technologies for the VOCs treatment. The finding can contribute to the more sustainable development in animal husbandry, in order to mitigate the negative environmental and health impacts caused by the VOC emissions.

Inspection robots have been widely used in modern agriculture. The high navigation accuracy is often required to rapidly position in complex dynamic environments in sheep farms. In this study, the 3D laser positioning and autonomous navigation were developed to tailor to the sheep farm. Firstly, a fusion approach was employed to combine the 3D LiDAR and an inertial measurement unit (IMU), in order to perceive the sheep farm environment. A tightly coupled radar-inertial navigation and mapping (RNAM) algorithm was used to construct the navigation map. Secondly, a viewpoint visibility approach was employed to preliminarily filter the dynamic point clouds. Egocentric ratio of pSeudo occupancy-based dynamic object removal (ERASOR) was integrated to enhance the dynamic point detection. The height and distance features were combined to further filter out the interference. Thirdly, the precise local positioning was achieved in an error state extended Kalman filter (ESEKF) using laser odometry and IMU. Stable global positioning was realized to enhance the adaptive monte carlo algorithm with the normal distribution transform-iterative closest point (NDT-ICP). Finally, a path planning was constructed to combine the A* algorithm with the timed-elastic-band (TEB) algorithm. Experimental results demonstrate that the filtering algorithm significantly improved the robotic positioning accuracy, compared with the conventional simultaneous localization and mapping (SLAM) algorithms without dynamic point cloud filtering. The average lateral and longitudinal deviations reached 35.2% and 28.7%, respectively. The overall positioning accuracy increased by 31.8%. The robot exhibited an average heading deviation below 2.4° with a standard deviation under 3.2° at the speeds of 0.3-0.5 m/s. While both lateral and longitudinal deviations were maintained, the average values were below 3.5 cm and standard deviations below 2.9 cm. Among the three motion modes—forward, backward, and line-changing- the forward mode demonstrated the highest accuracy, while the backward and line-changing modes shared slightly reduced precision. Nevertheless, all modes fully met the requirements of autonomous navigation in the agricultural robots. The high-precision map construction, positioning, and navigation were realized under complex dynamic environments in sheep farms. The 3D laser positioning and navigation can lay a foundation for the autonomous mobile platforms of the agricultural robots within sheep farm settings in complex environments.

Egg-laying chickens can inevitably perish daily in large-scale farms. However, manual inspection of deceased chickens cannot fully meet the large-scale production under harsh conditions. Consequently, inspection robots can be utilized as a highly efficient trend in modern agriculture. In this study, the inspection robot was performed on the egg-laying chickens house to swiftly and precisely detect the deceased chickens under complex environments. The detection of dead chickens were also introduced for the caged egg-laying chickens using an improved YOLOv8n model. The research objects were four-tier caged laying chickens on a large-scale laying chickens farm in Zhejiang Province, China. In order to improve the recognition effect and avoid misjudging the lying chickens as dead chickens. Thus, a "lie" label was added during annotation. The dataset was defined with a total of three labels: normal, lying, and dead, where the label was named "normal", "lie", and "dead", respectively. The dataset of 4000 images was constructed after labeling. The dead chickens were presented in 2000 images, while the lying chickens were presented in 1000 images. The dataset was randomly divided into a training set, a validation set, and a test set at a ratio of 6:2:2. Initially, cross stage partial hetconv (CSPHet) was employed to replace the C2f structure within the network using YOLOv8n. The gradient changes of feature maps were then focused across various levels. Additionally, the smaller 1×1 convolution kernels were utilized to substitute some 3×3 convolution kernels, effectively minimizing the model's parameters and computational burden while maintaining detection precision. Secondly, due to the acute occlusion issue in high-density caged chickens farming, occlusion among egg-laying chickens resulted in data loss. Therefore, the spatially improved attention module (SEAM) was integrated into the neck layer to compensate for the response loss in occluded regions and enhance the detection precision of occluded targets. Lastly, dynamics upsampling (DySample) was introduced to bolster the resolution of upsampled feature maps, and then emphasize the effective features of layer regions. A series of experiments were carried out to compare the effects of different input image sizes on the training.The results showed that when the input image size is 800 × 800 pixels, the precision, recall and mAP were the highest, which were 96.5%, 98.0% and 94.4% respectively. Furthermore, the highest recall was 98%, when the size of the input image was 640 × 640 pixels. The precision and mAP were 1.3 and 0.1 percentage points lower than the size of 800 × 800 pixels, respectively, but the required GPU was reduced by 24.4%. The more appropriate size of input image was selected as the 640 × 640 pixels, considering that the edge computing device of the robot was required to process the data of four cameras at the same time. After introducing SEAM mechanism, the box precision (box (P)) was improved by 0.6 percentage point increase, which confirmed the ameliorative effect of the SEAM mechanism on occlusion. The mAP of the improved YOLOv8n algorithm was 95.8% to detectthe dead chickens, with a precision of 96.6%, 2.46 MB of parameters, and an inference time of 25.8 ms. In addition, the improved YOLOv8n model was compared with the YOLO series models, such as YOLOv5n, YOLOv6n, and the original YOLOv8n. The results showed that the improved mAP increased by 2.5, 2.8, and 1.5 percentage points; the precision increased 1.7, 1.6 and 1.4 percentage points; the parameters are reduced by1.6%, 41.8% and 18.3%; the inference time decreased by 11.9%, 36.8% and 15.7% respectively, which showed that the improved YOLOv8n model has strong advantages in thecomprehensive performance. The real-time streaming protocol (RTSP) was configured to acquire real-time video streams for the real-time detection and recognition of deceased chickens among caged egg-laying chickens. The improved YOLOv8n model was successfully deployed on the edge device of the inspection robot, in order to validate the stability, applicability, and real-time performance. This finding can provide the technical support and reference to detect the dead chickens in the facility environment using edgedevices of robots.