Using statistics to save newborn piglets
![[Photo: Henk Riswick]](https://www.pigprogress.net/app/uploads/2021/04/001_780_rb-image-2711882-848x565.jpeg)
By including statistical models, scientists have developed a system that can help the farmer to monitor farrowing and to automatically and intelligently adjust the temperature.
Monitoring of farrowing can help reduce mortality in newborn piglets. The challenge is to divine exactly when farrowing begins so you do not waste time sitting around waiting.
This is a problem that scientists from Aarhus University have been attempting to solve. They have developed an automatic system that can predict the onset of farrowing more precisely. The system is based on online sensor information and a statistical model about sow behaviour around the time of farrowing, including nest-building.
You may also find interesting:
Small farrowing nest beneficial for piglet survival
When choosing to keep sows loose during farrowing and lactating, a smaller-sized nest appears to be more protective for piglets than a larger-sized nest, English researchers have found.
Does floor heating in individual pens need regulating?
Scientists have also developed an affiliated tool that can help the farmer to decide whether the floor heating in the individual pen needs to be regulated. The decision tool evaluates the costs associated with the heating in relation to the expected higher income from lower piglet mortality and then attempts to set a strategy that will give the best outcome for the farmer.
Research assistant Aparna Udupi, having done her PhD study on this subject at the university’s Department of Animal Science, said, “Research has shown that piglet mortality can be reduced by monitoring the farrowing or by regulating the temperature in the farrowing pens. These actions are difficult, however, if you do not know when exactly farrowing will take place and so have enough time to make and implement the right decisions.”
Sensors watch the sow
In recent years sensor technologies have been developed that, inter alia, can be used to monitor the physiological and behavioural changes in sows. For example, they monitor and record the individual sow’s water consumption and activity. The technology can be used to build a system that uses statistical models to predict farrowing with a greater degree of accuracy.
Previous studies have shown that monitoring the farrowing combined with farrowing assistance and good care of the young and weak piglets, such as warming them up and give them colostrum, can reduce postnatal mortality by up to 50% – but this is not always what happens in practice.
You may also find interesting:
Danish finding pig improvement everywhere
Just when you think that improvement in pig growth efficiency might have reached a zenith, Denmark finds ways to further increase output.
High operating costs in Denmark
Udupi: “High operating costs mean that you rarely do these things in Danish herds. A more accurate prediction of farrowing will be able to remedy this,” says Aparna Udupi.
Judicious use of floor heating
Scientists from Aarhus University have previously shown that timely warming of the floor to a temperature that is optimal for the newborn piglets can reduce piglet mortality. The timing of the start and finish of floor heating needs to be balanced against the temperature needs of both the sow and the piglets, and how long it takes to heat up the floor plus the costs of heating. Here, too, a more accurate assessment of farrowing would be valuable both economically and for the sake of the animals’ survival and welfare.
Also read: Extra heating reduces mortality of newborn pigs
Providing extra heat in the farrowing house can save the lives of many newborn pigs. Research at Aarhus University shows how underfloor heating can combine with room temperature to prevent hypothermia and deaths.
Udipi suggested, “With a few adjustments, our system can also be designed to predict when the sow starts nesting. This would give farmers the chance to allocate extra straw to the sow at the right time.”
Prediction of sensor-based algorithms
She predicts that the same sensor-based algorithm that she has been helping to develop can be used in other contexts, for example to predict disease outbreaks or outbreaks of behavioural problems such as tail-biting.
You can read the scientific article Hidden phase-type Markov model for the prediction of onset of farrowing for loose-housed sows in the journal Computers and Electronics in Agriculture.
Join 18,000+ subscribers
Subscribe to our newsletter to stay updated about all the need-to-know content in the pigsector, three times a week.
