Search

In the field of precision livestock farming, many systems have been developed to identify the position of each cow of the herd individually in a specific environment. Challenges still exist in assessing the adequacy of the available systems to monitor individual animals in specific environments, and in the design of new systems. The main purpose of this research was to evaluate the performance of the SEWIO ultrawide-band (UWB) real time location system for the identification and localisation of cows during their activity in the barn through preliminary analyses in laboratory conditions. The objectives included the quantification of the errors performed by the system in laboratory conditions, and the assessment of the suitability of the system for real time monitoring of cows in dairy barns. The position of static and dynamic points was monitored in different experimental set-ups in the laboratory by the use of six anchors. Then, the errors related to a specific movement of the points were computed and statistical analyses were carried out. In detail, the one-way analysis of variance (ANOVA) was applied in order to assess the equality of the errors for each group of points in relation to their positions or typology, i.e., static or dynamic. In the post-hoc analysis, the errors were separated by Tukey's honestly significant difference at p > 0.05. The results of the research quantify the errors related to a specific movement (i.e., static and dynamic points) and the position of the points (i.e., central area, perimeter of the investigated area). Based on the results, specific information is provided for the installation of the SEWIO in dairy barns as well as the monitoring of the animal behaviour in the resting area and the feeding area of the breeding environment. The SEWIO system could be a valuable support for farmers in herd management and for researchers in the analysis of animal behavioural activities. [ABSTRACT FROM AUTHOR]

Due to technological developments, wearable sensors for monitoring the behavior of farm animals have become cheaper, have a longer lifespan and are more accessible for small farms and researchers. In addition, advancements in deep machine learning methods provide new opportunities for behavior recognition. However, the combination of the new electronics and algorithms are rarely used in PLF, and their possibilities and limitations are not well-studied. In this study, a CNN-based model for the feeding behavior classification of dairy cows was trained, and the training process was analyzed considering a training dataset and the use of transfer learning. Commercial acceleration measuring tags, which were connected by BLE, were fitted to cow collars in a research barn. Based on a dataset including 33.7 cow × days (21 cows recorded during 1–3 days) of labeled data and an additional free-access dataset with similar acceleration data, a classifier with F1 = 93.9% was developed. The optimal classification window size was 90 s. In addition, the influence of the training dataset size on the classifier accuracy was analyzed for different neural networks using the transfer learning technique. While the size of the training dataset was being increased, the rate of the accuracy improvement decreased. Beginning from a specific point, the use of additional training data can be impractical. A relatively high accuracy was achieved with few training data when the classifier was trained using randomly initialized model weights, and a higher accuracy was achieved when transfer learning was used. These findings can be used for the estimation of the necessary dataset size for training neural network classifiers intended for other environments and conditions. [ABSTRACT FROM AUTHOR]

Innovative systems and automated computational procedures, such as those based on computer vision or inertial wearable sensors, have recently been adopted to provide effective and accurate monitoring and analysis of cow behaviour and respond to different issues related to cow health and welfare. In this study, a new and open source algorithm, characterised by a linear computational time, was defined and implemented with the aim to improve real-time monitoring and analysis of walking behaviour of dairy cows. It was applied to a novel inertial sensor-based system composed of low-cost devices, including wearable sensors, open source software, operating with a 4-Hz sampling frequency. The algorithm computed the number of steps of each cow from accelerometer data by making use of statistically defined thresholds. Two vector variables were considered to study the accelerometer signals, i.e., Signal Vector Magnitude and Signal Magnitude Area. Algorithm accuracy was carried out by comparing total error ( E ) and Relative Measurement Error ( RME ), and a sensitivity analysis on the parameters of the computed thresholds was carried out to analyse the variation of the error made by the algorithm. The results showed that the algorithm produced an E equal to 9.5%, and a RME between 2.4% and 4.8%. The sensitivity analysis confirmed that the proposed thresholds provided the minimum errors and that RME is less suitable than E for measuring the accuracy of the step counter. In fact, the underestimated and overestimated numbers of steps counted by the algorithm tended to compensate each other in RME computation. [ABSTRACT FROM AUTHOR]

Changes in cow behaviour may occur in relation to health disorders. In literature the suitability of using behavioural changes to provide an early indication of disease is studied. The possibility of achieving a real-time analysis of a number of specific changes in behaviours, such as lying, feeding, and standing, is crucial for disease prevention. Cow feeding and standing behaviour detectors were modelled and validated by defining a methodology based on the Viola–Jones algorithm and using a multi-camera video-recording system to obtain panoramic top-view images of an area of the barn. Assessment of the detection results was carried out by comparison with the results generated by visual recognition. The ability of the system to detect cow behaviours was shown by the high values of its sensitivity achieved for the behaviours of feeding and standing which were about 87% and 86%, respectively. Branching factor values for the two behaviours showed that one false positive was detected for every 13 and 6 well-detected cows, respectively. On the basis of these research outcomes, the proposed system is suitable for computing cow behavioural indices and the real-time detection of behavioural changes. [ABSTRACT FROM AUTHOR]

Simple Summary: Gas concentration is a relevant parameter for the estimation of emissions in dairy farms, but few studies have investigated the influence of cow behavior and barn management on gas concentrations in open buildings. In this study, concentrations of ammonia, methane, and carbon dioxide were investigated in an open dairy barn in a hot Mediterranean climate. Since hot climate conditions cause heat stress to the cows, gas concentrations were statistically analyzed to assess whether variation of environmental and animal-related parameters produced significant effects on the level of gas concentrations in the barn environment. In this study, it was statistically proved that daily gas concentrations were influenced by both the effect of micro-climate conditions, connected with the barn typology, and of barn management on the animals. Therefore, the mitigation strategies for the reduction of these gases could be pursued through the improvement of the barn management aimed at modifying cow behavior and through the control of climatic conditions in relation to the building features. Measurement of gas concentrations constitutes basic knowledge for the computation of emissions from livestock buildings. Although it is well known that hot climate conditions increase gas emissions, in the literature the relation between gas concentrations from open barns and animal-related parameters has not been investigated yet. This study aimed at filling this gap by evaluating daily gas concentrations within an open-sided barn in hot Mediterranean climate. The influence of microclimatic parameters (MC) and cow behavior and barn management (CBBM) were evaluated for ammonia (NH3), methane (CH4), and carbon dioxide (CO2) concentrations. Results showed that both MC and CBBM affected concentrations of NH3 (p < 0.02), CH4 (p < 0.001), and CO2 (p < 0.001). Higher values of NH3 concentration were detected during the cleaning of the floor by a tractor with scraper, whereas the lowest NH3 concentrations were recorded during animal lying behavior. Measured values of CO2 and CH4 were highly correlated (C = 0.87–0.89) due to the same sources of production (i.e., digestion and respiration). The different management of the cooling systems during the two observation periods reduced significantly CH4 concentrations in the barn when the cooling system in the feeding area was switched off. Based on methodological choices due to the specific barn typology, parameters related to animals can provide information on the variation of gas concentrations in the barn environment in hot climate conditions. [ABSTRACT FROM AUTHOR]