Students' Research Circle    
» 2021
Call for papers
The conference
Veterinary Session
Veterinary Jury
Biology Session
Home » Archive » 2021

TDK conference 2021

Developing and applying automatic walk-over-weighing system in extensive cattle herds
Fodor Bálint - year 5
University of Veterinary Medicine Budapest, Department Animal Breeding, Nutrition and Laboratory Animal Science
Supervisor: Dr. Ákos Maróti-Agóts


Already precision livestock breeding devices are not only available in intensive animal husbandry, where electronic systems are easily can be built, but also it may become possible in extensive stocks such as hungarian grey cattle herds.

Weighing usually happens in stock with using mechanic or digital scale after waiting until that the animal calmly stands, only after this the weighing can be done. Devices which can detect the animals’ body weight without they would been stopped called dynamic or automatic walk-over-weighing scales.

In extensive herds the individuals rarely get contact with humans and this is the reason why weighing procedure is usually problematic. Because of animal protection and animal welfare reasons there were several attempts to develope large amount of data provider (Big Data) devices which can be able to serve precision animal husbandry’s needs. Our institute started development of a walk-over-weighing automatic, dinamic scale within the framework of a tender.

The dynamic scale was designed and manufactured by Emalog Kft based on experiences and literature data (Cveticanin, 2004).

The prototype of the scale was tested at two locations in Apaj (n = 52) and Hortobágy-Kecskés (n = 103). The traditional measurement was performed in Apaj with a digital scale (Tru-Test, ID-500), in Kecskés with a mechanical scale (Metripond, 1997).

The raw measurement results of the dynamic automatic scale were processed with three data cleaning and estimation algorithms, and then statistically compared with the results of the traditional method. The results obtained with the dynamic scale and the values measured with the conventional scale were within +/- 5%, which is in agreement with the data in the literature (Dickinson, 2013; González, 2014).

The animal welfare, animal protection and ethological experience of the application of the scale was characterized by the ratio of the time spent on the measurement and the measurement incidents - carving, turning, eruption, etc.

We conclude that the automatic scale significantly speeded up the animal measurement process because the well-placed scale (Apaj) reduced the herding time. On the other hand, unaccustomed animals, if they had time to notice the scales (Hortobágy) for the first time, then the „meeting” significantly prolonged the herding time.

As a continuation of our work, connected to the VATEM3 system from the body surface of the animals obtained from the standard side and top view image (planimetry), we will examine the weighing measurement results estimated with the artificial intelligence deep learning algorithm using a similar methodology.

List of lectures