The strong growth of the world population will cause a major increase in demand for bovine milk, making it necessary to use various technologies to increase milk production efficiently. Some technologies that can contribute to solving part of this problem are those related to data analytics tools, big data, and sensor development. It is timely to review modern technologies and data analytics methods for milk predictions in view of supporting decision-making in dairy farms. To this end, a scoping review was carried out, which resulted in 151 articles of interest. Among the most important results, we found that (i) the identified studies are relatively recent with an average publication time of 5.95 years; (ii) the scope of the selected studies is mostly concentrated on milk and prediction (29%), early detection of lameness (26%), and timely detection of mastitis (13%); (iii) the type of analysis is mostly predictive (87%), and prescriptive is barely present (3%); (iv) the types of input data used in the studies are preferably historical (70%), and real-time data (25%) are used less frequently; (v) we found that the method of artificial neural networks (47%) and the convolutional neural networks (24%) are the most used for the studies regarding bovine milk output predictions. In the selected studies, the artificial neural network methods have considerable accuracy, recall, precision, and F1 Scores on average but with high ranges and standard deviations. (vi) Simulation tools are scarcely used, being present in 4% of cases. In the treatment of variability, the models reviewed are mostly deterministic (77%), and the stochastic models (5%) are considered in a small number of cases. Based on our analysis, we conclude that future research on decision-making tools will benefit from the advantages of artificial neural networks in data analytics combined with optimization–simulation methods.
This study provides a comprehensive scoping review with the aim of determining the mathematical methods applied to dairy cow replacements that will serve as a basis for future research in this field. In the WOS and Scopus databases, a search was carried out for peer-reviewed, English articles, where a process of discarding those that did not address the topic related to our objective was carried out, and where the titles, keywords, and full text were reviewed sequentially. We obtained a total of 40 selected articles. Dynamic programming is the most commonly used optimization technique, present in 58% of the studies, followed by stochastic simulation in 40%, and deterministic simulation in 8%. Machine learning methods or hybrid approaches are applied in only 5% of the cases. The review identifies milk production as the most frequently used response variable, appearing in at least 58% of the studies, and profit as the primary economic indicator, utilized in 78% of the cases. This research underscores the importance of these methods in improving the efficiency, profitability, and sustainability of dairy farming operations. Future research could address the inclusion in models of diseases and animal characteristics that have not yet been considered, as well as expand the scarce use of machine learning tools and the hybridization of such models with statistical ones.
This study aimed to explore Markov decision methods in order to solve the problem of dairy cow replacement, adding the special characteristics of two types of abortions due to different sanitary reasons that influence the economic, production, and reproduction performance of these animals. The model was successfully validated against other models published in the literature. Python code v.3.13 was used to solve the problem and to ease future extensions with the inclusion of new variables. The results constitute tools that allow the veterinarian to explore more realistic scenarios by running a Markov simulation model that avoids the complexities leading to the problem of dimensionality in dynamic optimization models. In our study, the economic value of the herd considering RA and NLA abortions shows that the maximum net benefit is USD 178.77 per cow, and non-pregnant cows are slaughtered upon reaching six months of lactation, a value that is within the range of values reported by the literature that we have identified. At the optimum, the replacement model extended with abortion generates a difference of USD 0.69 per cow per month compared to the model that does not include the special abortion features. The changes in the net present value of each cow according to the month of culling depend on the variability of milk income and slaughter value and heifers’ replacement values, suggesting that any measure that seeks to improve the economic benefit of dairy cows should take greater account of these variables.