All media

ILVO press release - Thursday, February 4, 2016

ILVO-KU Leuven doctoral defense "Precision animal husbandry: sensors keep an eye on welfare, health and production in pigs"

Jarissa Maselyne recently defended her doctoral thesis “Automatic monitoring of the eating and drinking patterns of pigs: towards a warning system for productivity, health and welfare problems in individual pigs.

Abnormal eating and drinking behavior in pigs can be discovered in early stages using a new sensor system developed by ILVO – KU Leuven researcher Jarissa Maselyne. Abnormal eating and drinking patterns can be a signal of underlying problems in terms of animal health and welfare, which can also signal production problems. With this system, problems such as fever or severe lameness can be discovered even after 1 day. “In the larger pig production systems, monitoring individual pigs just by looking at them is becoming more and more difficult. If you can add sensor monitoring then we think you get an important added value for the animal and for farm management. The growing societal concern for animal welfare also points to a need for intensive individual follow-up of each animal, and this system answers that need.”

Technical background: RFID sensor for eating and drinking behavior

The standard feed and drinking systems were equipped with a High Frequency Radio Frequency Identification (RFID) system. Each pig was outfitted with a passive RFID transponder with a unique number in the ear. In this way, each pig’s visit to the feed trough or drinking nipple was registered and the individual patterns could be observed. In her doctoral research, Jarissa Maselyne paid a great deal of attention to comparing the RFID data with the reality in the barn based on human observations. From the RFID transponders fixed to the feed trough or drinking nipple, the number, duration and time of eating and drinking episodes could be derived. Water usage could also be estimated based on the RFID on the drinking nipples.

Four warning systems were developed and compared as a way of noticing changes in eating and drinking behavior of individual animals. For this purpose, the number of RFID registrations and the time between meals were the most important parameters. For each pig, the “normal” values were calculated. This is important to determine the normal variation in the eating and drinking patterns of each pig (e.g. age effects) to differentiate from abnormal variation (as indication of a problem). Every pig was used as its own reference and the limits are animal-specific. The use of constant, non-individual limits were also tested in the system.

When does the alarm bell go off?

During the study, abnormal eating and drinking behavior was detected via the warning system. These were signaled each day as a warning when a pig exceeded a certain limit. The warnings were then compared with extensive human observations of the group of pigs. Per day, the correct and incorrect warnings produced by the system were determined. From this analysis, it appeared that the system based on individual limits and on the basis of the number of RFID registrations was the best. This system had a sensitivity level of 66% (percentage of pig-days with a problem that was correctly signaled), specificity of 98% (percentage of pig-days without problems that also yielded no warnings), and precision of 67% (percentage of correct warnings) for all health, welfare and production problems detected by the observers. In other words, 2 of 3 problems were detected and 2 of 3 warnings were correct. Serious problems like severe lameness and fever were detected on average after 1.1 days.

Ready to use?

According to the researcher, the precision and the sensitivity of the system can be improved in a follow-up study. “It is also a good idea to precisely calculate how much the system could save in a practice situation in terms of time spent, reduction in antibiotic use, costs and improvements in efficiency.” The first results are promising - for the farmers, the pigs and the consumers. Technically, the system is within reach of every farmer. Maselyne says, “The investment cost per pig will be higher if they are housed in small groups because each feed trough needs to have an RFID system.” But she is convinced that it will offer value to the sector.

The growing demand for pork and accompanying scaling up of pig farms and the growing societal concern for animal welfare all call for intensive monitoring of animals on an individual level. This RFID system answers those needs. This type of electronic ID on pigs can also help to integrate information on breed, farm management data, medicine use, and slaughter data.


Greet Riebbels, ILVO communications,, +32 486 26 00 14
Jarissa Maselyne, doctoral student,
Lieve Herman en Annelies Van Nuffel, ILVO promotor,
Wouter Saeys, KU Leuven promotor,

Additional info

This doctoral study was originally part of the PigWise project and is a collaboration between KU Leuven and ILVO, funded by IWT. The study is also related to ISense, a research platform for precision agriculture at ILVO. The aim of this initiative is to track the fast technological evolution in terms of sensors and the related data analysis, and also to quickly provide practical applications of these technologies in the agro-food chain.