Subscribe
Share
Share
Embed
University of Minnesota Extension Swine team gives research updates through podcasts.
Sarah Schieck Boelke: Hello, and welcome to Minnesota's Swine and U podcast series, a University of Minnesota Extension swine program. Today's podcast is a research update on precision and accuracy of the doser in electronic sound feeders. My name is Sarah Schieck Boelke, your host, and I'm a swine Extension educator with the University of Minnesota. Joining me today to talk about his research is Samuel Berrio, who is a graduate student in the Department of Animal Science. So welcome, Samuel.
Samuel Berrio: Hello, Sarah. Thank you for the invitation and the introduction.
Sarah Schieck Boelke: Yes. So to get us started today, Samuel, will you tell us a little bit about yourself and including who you are working with on this research?
Samuel Berrio: Yeah. So, I am a veterinarian, and currently, I'm doing a master's in animal science, with a focus on swine nutrition and management. And I'm working under the guidance of doctor Lee Johnston and doctor Pedro Riola.
Sarah Schieck Boelke: So just another question to add in here. But so you said that you are a veterinarian. So what prompted you to do a master's in animal science, and and what led you to the University of Minnesota for that?
Samuel Berrio: Yeah. So after I completed the DVM I'm from Columbia, I did in Columbia, I got accepted into the MAST program. That is an exchange program from the university.
And, yeah, that's the way I arrived to to Minnesota and started working with doctor Pedro and doctor Lee after that.
Sarah Schieck Boelke: Alright. Very good. At least, myself, I always find it interesting, the paths that students take to come to the University of Minnesota for their graduate work. Okay. So now we can get to, I guess, the main reason, why you're talking to me today, and that is about the research that you've recently completed.
So would you share with us what is the research topic that you'll be sharing and talking about today?
Samuel Berrio: Yes. So today, I'll be sharing one of our research projects that is about the precision and accuracy of a supplement doser using with electronic sow feeders for gestating sows called Gestal Multi Unit. This equipment operates with the Gestal 3G feeding system and allows individual sow supplementation in group housing so they can receive targeted supplementation at the specific moments during gestation, which creates an opportunity for both management and research. And we wanted to know, like, how accurate and precise is this equipment. So that's that's the question behind this research.
Why we did this was because in a parallel study where we evaluated zinc supplementation during gestation, we used this equipment to supplement the sows.
Sarah Schieck Boelke: Very good. And I definitely would like to learn more. But before we get to some of those questions for you to explain, how is your research funded?
Samuel Berrio: Okay. So this research was funded by the USDA through a grant program focused on evaluating effects of high levels of zinc during cell gestation. And as a part of this project, as I mentioned before, we use these multiunits electronic dosers to deliver the zinc supplement to the sows that are warehousing in in group. And Jyga Technologies donated this equipment and collaborator collaborated with us throughout the project.
Sarah Schieck Boelke: Alright. Thank you for sharing that, and thank you to our funding source as well as the donators for the equipment. As a lot of people don't always realize that research that's done by our universities, including our University of Minnesota, can't be done without support from either a funding source for the project and/or also, sources for equipment that might be used. So I always like to make sure that we give credit to those funding sources. So now, Samuel, let's get into talking more about your research.
Can you give a brief introduction to your project explaining why it was a valuable project to do for the industry?
Samuel Berrio: Okay. Yeah. So before this technology is implemented more broadly, and this this is a a new product. So we, as I mentioned before, we collaborated with Jyga, and we the objective of this study was to evaluate the precision and the accuracy of this equipment under commercial conditions and identify potential sources of variation in the total amount of supplement delivered and also, like, about the precision and accuracy of the individual drops because all the the total amount delivered by the equipment was done in multiple drops, say, from four to six. Some important definitions to keep in mind is the accuracy that refers to the deviation from the programming target and precision that refers to how consistent the deliveries were in repeated measurements.
And okay. So this equipment is integrated with the Gestal 3G electronic self feeding system. And the way it work it was working during our study was as a proportion of the total feed allowance. So was 7% of the total feed allowance. So if a if a sow was programmed to receive two kilograms, 140 grams were going to be delivered by the multiunit and the rest of it by the feeding line.
And this was done in a commercial sow farm where they managed five different feeding allowances, and we used these ones as well for the study. So the target of the supplement was between 119 to 207 grams daily for an individual sow. And okay, so this equipment has a capacity of about 9 kilograms. And, yeah, it's you have to fill it manually unless you have to you have it connected to to a feeding line.
But in our case, was fill it manually, and you can use it for delivering any supplement or product that that you want.
Sarah Schieck Boelke: Interesting. So, basically, you were well, I guess, like you said, testing the precision and accuracy of this doser to make sure that not just once, but several times, that it gave the correct dosage of that supplement, and it continued to be precisely and accurate at at that dosage. Right? And and also it sounds like there was different dosages depending on what, like, the total amount of feed that that sow would have received.
Samuel Berrio: Right.
Sarah Schieck Boelke: Is that correct?
Samuel Berrio: Right. And since it worked as a proportion, for example, the feeding the feed allowances were based on parity and body condition. So for example, a gilt average gilt will receive a 1.7 kilogram.
An average sow will receive 2 kilograms, and so the 7% of this total allowance will change as well.
Sarah Schieck Boelke: Mhmm.
Samuel Berrio: And this was done in in a commercial sow farm in in Marshall, Minnesota, and we tested the equipment for about six months last year.
Sarah Schieck Boelke: So you had several repetitions. So do you wanna explain and you have a little bit, but a little bit more, how did you complete this study?
Samuel Berrio: Yes. So during this period, we tested 88 feeding stations, so 88 multi-unit electronic dosers. And at the end, we collected and we would have a total of 170 observations with the individual weights of each drop. Right? Since the equipment delivered the supplement in multiple drops.
We collected the individual drops, weight weighed them, and by adding them, we know how much was delivered in each feeding event. So the way we did this, that was before each sampling, we calibrated the equipment. And by doing this, we we tell the system how many grams are being delivered in a specific amount of time. So this was very important to to be done before each testing.
And we use an RFID ear tag. We add it into the pen where is located where where was located, the the equipment test. And we brought it next next to the sensor to trigger the feeding event and simulate, like a daily sow feeding test. Then we collect it with a an aluminum tray the the amount delivered by each drop and weigh it with a digital scale and start building the dataset in a spreadsheet using a tablet.
Sarah Schieck Boelke: So now that you've shared how you've completed this study, what were the results that you got?
Samuel Berrio: Okay. Yes. When we look like the total amount delivered by the multiunit in each feeding event, that is the sum of all of the drops, we observed that the precision was better at higher allowances. So the deliveries were more consistent when the target amount was larger. The distributions were less spread out.
And across allowances, the coefficient of variation ranged between 7.3 and 13.3. And in terms of accuracy, the system consistently delivery is slightly less than the programming target and on average was about 10 grams below than the target value. That that was about 7% of under-delivery. Yes. This was at the total amount delivered.
And we also did an analysis. We evaluate the performance analysis at the individual drop level. So during this study, the equipment was set to deliver a fixed proportion of the total feed allowance. In this case, it was 7%. But in many research and management situations, it might be more useful to deliver a fixed amount of supplement independently of the total feed allowance.
So let's say, I wanna give 30 grams to all of the sows or 50 grams to all of the sows. And right now, with recent software versions, it's possible to do it. But at the moment, when we we're doing the trial, it was not possible to make the equipment work that way, but it's possible right now. So since our dataset included the individual drop weights, we observed that many of these individual drops shared a common target value. Most of them were centered around 32.5, 32.5 grams.
So what we did was to pull many of these drops together that share discount distribution. So we pulled 546 individual drops to evaluate the precision and accuracy of at this smallest delivery unit. So in other words, we could answer, like, how how how precise and accurate will work the multiunit if we have a target value of 32 grams, 32.5 grams. And we observed a coefficient of variation of 17% and, I mean, absolute deviation of 4.2 grams. So on average, each individual drop was inaccurate by 4.2 grams.
Yeah. We also observed that 92% of these drops with a target value of 32.5 grams were between 26 and 45 grams with a yeah. With 8% of extreme values.
Sarah Schieck Boelke: But it seems like, um, it wasn't too far off.
Samuel Berrio: Right. Right.
Sarah Schieck Boelke: So you mentioned um, that during the study, there was some things that maybe the software or the system didn't allow you to do, but now a person could do. So I'm assuming the supplier of the equipment, did they take kinda some of that feedback of what you were seeing with your research project to make updates to their system?
Samuel Berrio: Right. Yeah. So we we were working together about this, and, yeah, we agreed that a configuration with a fixed amount will be, like, the most valuable or at least that both options are possible.
Sarah Schieck Boelke: Mhmm.
Samuel Berrio: And and, yeah, right now with the recent software versions, it's possible to do it.
Sarah Schieck Boelke: Good. That was that was, good that you were able to have that working relationship, and I'm sure the supplier of the equipment was happy about that as well.
Samuel Berrio: Yeah.
Sarah Schieck Boelke: So based off of your results, what conclusions can be made from your study?
Samuel Berrio: We concluded that under these conditions and how it was operating, the equipment consistently underdelivered, supplemented relative to the programming target with an average of 10.1 grams or 6.7% less than the expected value. And if the equipment is operating like this, this can be accounted in in the in the settings. So if you serve this consistently, if it's a you can account, like, program 7% more. Right? But yeah.
And, also, the precision and accuracy improved as a programmer. A value increased, meaning that the system was performing better when the total amount of supplement delivered was was larger and when it was being delivered in multiple drops. And since right now, it's possible to operate the multiunit as a fixed amount, we which might be more useful for research and management application. We expect that this configuration could improve the accuracy and precision of of the equipment.
Sarah Schieck Boelke: So these results that you shared, why are they important takeaways?
Samuel Berrio: They are important because it helps us to understand how the equipment work at different levels. Right? Like working operating in multiple drops or at individual drops since so before this technology is more broadly applied, it's important to know how it's performing and also to identify things that can be improved. And since the recent versions now allow it, then we evaluate it with the data that we had available with an underlying target value of 32 grams. But you can set up that amount based on on your necessities.
Sarah Schieck Boelke: And like you mentioned too, I'm assuming that some important takeaways were were also how you communicated with the company so that they were able to make those updates.
Samuel Berrio: Right. Right. Yeah. It was was we did we had good meetings and conversations about the equipment and the potential things that were affecting how it was operating or things to improve and was was very good experience working with them.
Sarah Schieck Boelke: Mhmm.
Samuel Berrio: But, yeah, definitely, this equipment has a lot of potential. And because it creates a lot of opportunities for precision feeding in both research and commercial sow management, gives us the ability to target individual sows at the specific stages of gestation in group housing. So you can have 50,100 sows housed together, and just with the computer, you can set up which of them you want to target and at which moment. So it's very it gives you a lot of flexibility in the management.
Sarah Schieck Boelke: So we're coming to the end of our podcast here, Samuel. To wrap up our discussion, do you have any closing remarks or anything that you wish you would have said earlier but maybe forgot?
Samuel Berrio: No. I think a special acknowledgment to to USDA and JygaTechnologies and Boerboom as well. That is the commercial cell farm where we did the this research and to my advisers. And, also, as I mentioned before, well, this equipment gives us many opportunities for precision feeding approaches, and it's opened the door, like, to more flexible feeding strategies and research questions that will be difficult to study with conventional feeding systems or by doing it manually.
Sarah Schieck Boelke: Well, thank you, Samuel, for sharing your research on precision and accuracy of the doser in electronic sow feeders.
Samuel Berrio: No. Thank you for the invitation.
Sarah Schieck Boelke: And thank you to those listening to the University of Minnesota Swine & U podcast. This has been Sarah Schieck Boelke, swine extension educator, along with Samuel Berrio, animal science graduate student. To further connect with the University of Minnesota Swine Extension, please visit the swine-specific webpages on the University of Minnesota Extension's website at www.extension.umn.edu/swine. And on those pages, you will find connections to our blog and our Facebook page. To further learn about research being done by our swine faculty in veterinary medicine, please visit their swine in Minnesota blog at www.umnswinenews.com.