dr.ing. IJMM (Iris) Boumans MSc

The focus of my research is how today's knowledge and technology can contribute to animal welfare. In particular, to find win-win situations between animal welfare and other sustainability issues (i.e. social, economic and environmental issues), while also identifying possible trade-offs between them. I concentrate on pigs but my research can be applied to all kinds of animals.

During my PhD, in which I specialised in modelling pig behaviour and sustainability of pig systems, I concluded that we do not optimally use feeding behaviour of pigs as an indicator for pig health and welfare. Just like humans, pigs have a circadian feeding pattern, with regular feeding times. Deviations in these patterns, such as a result of sickness or stress, can be difficult to detect by humans, however, may be detected more easily by sensors technology & AI. With this technology we are able to closely monitor individual animal behaviour and gain more insight in (ab)normal behavioural patterns and relations with productivity and animal welfare. This knowledge is also valuable for applied research, in which behavioural patterns can serve as indicators in an early warning system for animal health and welfare in farms.

Livestock experimental data has been collected and analysed for long time, but was mainly sampled at particular time points and analysed at group level. Sensor data, however, are often designed to collect data continuously and at individual level, which require different data analytics than current common (statistical) methods. This can include approaches such as feature engineering, time series analysis and machine learning algorithms. Furthermore, sensor data can be noisy, incomplete or erroneous, and pre-processing and cleaning methods to check and improve data quality is important before analysis. Development of methods to deal with data quality as well as methods for further analysis are therefore required and something I like to contribute to.

Because sensor technology for pig systems is relatively expensive it does not pay off for commercial growing-finishing farmers to invest in this yet, although I do expect this will change in the future. In gestating sows, however, sensor technology is already more common, and mainly used for management, but not for animal welfare. One of the problems is that we lack the right indicators for animal welfare from sensor data. Understanding within and between variation in animals is of major importance in this and here I foresee big opportunities for the future: to further unravel individual variation and long term behaviours of animals. This will allow us to find new and better indicators for animal welfare, as well as for optimal production results.

At the same time, applying sensor technology and measuring behaviour won’t be enough to improve animal welfare. It will also depend on what is done with this information. Furthermore, sensor technology is usually applied in large, intensive systems, which might not provide the best environment for a pig. Pigs have strong innate behaviours (e.g. exploring and foraging) that cannot be sufficiently performed in current intensive housing systems. Therefore, I think it is also important to develop farming systems that take the pig specific needs into account and allow for related behaviours. It is interesting to study how sensor technology can be best applied in these systems.