Da Rosa Righi R., Goldschmidt G., Kunst R., Deon C., André da Costa C. (2020). Towards combining data prediction and internet of things to manage milk production on dairy cows. Computers and Electronics in Agriculture, 01/02/2020, vol. 169, p. 1-13.
https://doi.org/10.1016/j.compag.2019.105156
https://doi.org/10.1016/j.compag.2019.105156
Titre : | Towards combining data prediction and internet of things to manage milk production on dairy cows (2020) |
Auteurs : | R. Da Rosa Righi ; G. Goldschmidt ; R. Kunst ; C. Deon ; C. André da Costa |
Type de document : | Article |
Dans : | Computers and Electronics in Agriculture (vol. 169, February 2020) |
Article en page(s) : | p. 1-13 |
Langues : | Anglais |
Langues du résumé : | Anglais |
Catégories : |
Catégories principales 06 - AGRICULTURE. FORÊTS. PÊCHES ; 6.4 - Production Agricole. Système de ProductionThésaurus IAMM PRODUCTION LAITIERE ; VACHE LAITIERE ; EXPLOITATION LAITIERE ; AGRICULTURE NUMERIQUE ; PREVISION ; PRODUCTIVITE ; ALIMENTATION DES ANIMAUX |
Résumé : | Dairy farms play a fundamental role in both social and economic areas, due to the high quantity of milk production they have. In this area, the nutritional management and the dairy control practices must be carefully analyzed, since they are directly related to the cows productivity in terms of quantity and quality of the milk produced by the animals. To the best of our knowledge, there is a research gap on proposing and evaluating computational models to automate and individualize the feeding of the animals and its impact on milk production. In this way, this work proposes the MooCare model, designed to assist producers in the management of dairy cattle, in order to obtain better productivity indexes. Using the Internet of Things devices, MooCare performs the automation and individualization of the animals feeding. Data gathered using such devices enable the MooCare scientific contribution, which consists of providing the milk production forecasting of each cow, using the ARIMA prediction engine. Thus, farmers can be notified in advance, so enabling them to react in terms of designing a better nutritional plan, which can be individualized for each cow. The evaluations comprise a simulation based on data of cow lactation obtained from a real farm. The experiments presented encouraging results since the assertiveness of the prediction indicated small differences if compared to the actual milk production. Besides, the notification system was able to forecast nutritional issues, allowing the farmers to design individualized plans for specific cows. |
Cote : | Réservé lecteur CIHEAM |
URL / DOI : | https://doi.org/10.1016/j.compag.2019.105156 |