Big Data is the New Oil
Lorne and Vicki Tannas
Lorne and Vick Tannas are swine specialists and nucleus support for Genesus farms in China.
“You don’t know what you don’t know”, many of us continue to do things the same way we have because of the fear of change, even if it means spiraling down into disaster. Fear is a far more dominant force in human behavior than euphoria. Errors using inadequate data are far less than those using no data at all. Some of the best theorizing comes after collecting data because then you become aware of another reality. “You don’t know what you don’t know”.
Collecting data without a purpose is just a make work project. Start with comparative data and observe best practices, your farm versus another farm. This will get you thinking that there is “something going on”. Then progress to data mining of one’s own data to become informed. (Remember it is not about the data but about becoming informed). Why do I get these results when others are getting something else?
A worker of a competitor genetic company once said, “No wonder Genesus does better than the rest, you have all the best farmers buying your genetics”. At first glance, this seems funny, and then I think, I agree with him. When you look at the information based on the data collected from these “best farmers”, it does support his argument. Another farm is doing better than my farm because they are getting better results. My farm needs to do something that the “Best farmers” are doing and see if it helps. We then collect data and try out our hypothesis. In some cases, it is simple, like using better genetics, but in most cases, it is multifactorial and can be interdependent.
Let us look at a couple of Genesus farms: Farm A and Farm M. Farm M does many very good things but does not wean as many as he would like and has a lot of Pre-wean mortality.
Farm M then does a comparison to Farm A. Pre-wean mortality is one of those multifactorial things; health, environment, stockmanship and nutrition all play key roles. Looking at this data, we can see birth weights are low in Farm M. Total born and born alive are greater than in farm A but Farm A weans more and has less pre-wean mortality. We do not have to prove low birth weights will cause higher death loss in farrowing, this has been proven and we can find lots of data on the internet support that. Low birth weight pigs require higher temperatures more attention to detail (colostrum management, split suckling). Low birth weights pigs can contribute to chronic illness and spread it to larger healthy animals. If the average is 1.3 Kgs, the pigs born below 800 grams will have a low chance to survive. What is the lowest weight and how many? How broad is the bell curve? Meaning, how much variation is there in a litter or between litters? If we can shift the bell curve to the right by increasing the birth weight we would assume that the bottom enders will move right as well. If we can shift the bell curve to the right, we will have less pre-wean mortality and more piglets weaned.
Farm M hypothesis is; “If I have heavier birth weights my pre-wean mortality will go down and I will wean more pigs”.
Increasing feed pre-parturition is something easy to try but we need to become more informed. By digging deeper into the two farms, we find that Farm A starts to increase or up feed in gestation on about day 90. Farm M waits until day 100 to up feed the gestation sows. What happens in the uterus from day 85 is the embryo goes from 20 grams of protein a day requirement needs to 200 grams per day. Farm M by waiting until day 100 has his sows falling into a deficiency in protein uptake. The sow then starts to pull protein and essential nutrients from its own body. When I walked through the herd there were many sows starting to dish out at the tail head. This is a good way to see if sows are losing body mass.
A good hypothesis will open up more questions than it answers. This is what collecting data does. This is that euphoric moment when you become informed and start asking other questions and move to a new reality. “I now know what I didn’t know”.
Some other questions: Is age a factor? Gilts absorb vitamins and minerals at about 39%, where five parity sows absorption rate drops to 16%. Is this absorption rate a factor? Is my nutrient level pre-farrowing sufficient for litters of 16.12 total born? If we are formulating based on 12 and are getting 16 we are 33% deficient in piglet nutrient requirements. Am I split suckling litters over 12? How well is my fostering and balancing litters?
What is my environment like? And so on.
Data has become a resource, like oil, it can be mined to provide innovations and insight. You can have data without information but you cannot have information without data. Do something with the data you have, collect more and become informed. Do not just accept your current reality.
Categorised in: Global Tech
This post was written by Genesus