WEBVTT 00:00:10.220 --> 00:00:14.039 Creeping across the pasture land. Got dirt on 00:00:14.039 --> 00:00:18.719 the boots and a working hand. From the milk house 00:00:18.719 --> 00:00:24.000 home to the auction call. These are the folks 00:00:24.000 --> 00:01:00.710 standing strong and tall. Yeah, this is... Welcome 00:01:00.710 --> 00:01:03.869 to the Bullvine Podcast. Victor Cabrera came 00:01:03.869 --> 00:01:06.450 to the U .S. from Peru as a student who waited 00:01:06.450 --> 00:01:09.069 months just to get a single research paper in 00:01:09.069 --> 00:01:11.810 the mail. Sixteen years later, at the University 00:01:11.810 --> 00:01:14.510 of Wisconsin -Madison, he's built a system called 00:01:14.510 --> 00:01:17.250 Dairy Brain that predicts mastitis days before 00:01:17.250 --> 00:01:20.769 any clinical signs appear. Over 90 % accuracy, 00:01:21.069 --> 00:01:24.549 on real farms, with data your operation already 00:01:24.549 --> 00:01:28.900 generates. So why is a $40 billion industry still 00:01:28.900 --> 00:01:31.840 running five software systems that can't talk 00:01:31.840 --> 00:01:34.200 to each other? That's what this conversation 00:01:34.200 --> 00:01:38.560 is about. Now, coming not from a North American 00:01:38.560 --> 00:01:43.280 background into our production systems, can you 00:01:43.280 --> 00:01:45.980 explain maybe how that brought a different perspective 00:01:45.980 --> 00:01:47.840 and how that's maybe influenced your career? 00:01:49.260 --> 00:01:54.340 Oh, for sure. It makes a huge difference because 00:01:54.340 --> 00:02:02.079 my feeling, and I still kind of see this related 00:02:02.079 --> 00:02:05.239 on my students who are international in the big 00:02:05.239 --> 00:02:13.719 majority, come or we come with a lot of... passion 00:02:13.719 --> 00:02:17.659 and expectatives. And so we give it all, basically. 00:02:17.860 --> 00:02:21.240 I think that's a big difference. We are coming 00:02:21.240 --> 00:02:26.500 and we see here a huge opportunity and we try 00:02:26.500 --> 00:02:30.139 to take the best advantage of that. And we are 00:02:30.139 --> 00:02:33.280 ready to learn as much as we can and take advantage 00:02:33.280 --> 00:02:37.479 of the technology and try to, I mean, there are 00:02:37.479 --> 00:02:39.580 things like, for example, all these professors 00:02:39.580 --> 00:02:43.180 who We only knew because of the publications 00:02:43.180 --> 00:02:46.379 or the books that we were studying in our countries. 00:02:46.539 --> 00:02:51.740 They were here in the offices, and we could visit 00:02:51.740 --> 00:02:54.479 with them. They were very approachable. One thing 00:02:54.479 --> 00:02:57.419 we feel very different, at least that was my 00:02:57.419 --> 00:03:00.259 experience in Peru, for example, if I knew of 00:03:00.259 --> 00:03:04.659 a famous professor or researcher. I didn't have 00:03:04.659 --> 00:03:08.159 access to that person very easily. At the time 00:03:08.159 --> 00:03:11.759 I was studying as well, if I needed a new paper, 00:03:12.000 --> 00:03:14.979 I needed to request it. It took several weeks 00:03:14.979 --> 00:03:19.479 or months until I get it. The fact of coming 00:03:19.479 --> 00:03:23.039 here and have this all handy makes a huge difference. 00:03:23.300 --> 00:03:26.800 And we give it for granted nowadays, right? And 00:03:26.800 --> 00:03:29.840 even now with all this AI and the internet, we 00:03:29.840 --> 00:03:32.379 have access to everything. But I think that's 00:03:32.379 --> 00:03:36.039 something, Still, there is a huge advantage here 00:03:36.039 --> 00:03:39.960 in the developed world compared with the developing 00:03:39.960 --> 00:03:46.219 world, right? And so I think we feel that and 00:03:46.219 --> 00:03:49.460 we appreciate that. I think that's the main point. 00:03:49.520 --> 00:03:52.460 We appreciate that and we take the best advantage 00:03:52.460 --> 00:03:56.639 of that. Yeah, I was noticing, and I go through 00:03:56.639 --> 00:04:00.240 some of the research papers and stuff you've 00:04:00.240 --> 00:04:02.379 done, and one of the things I notice is, You're 00:04:02.379 --> 00:04:06.039 not always advocating for more technology, but 00:04:06.039 --> 00:04:08.680 you're talking about how it all works together 00:04:08.680 --> 00:04:12.580 and how it's all fragmented. Was that one of 00:04:12.580 --> 00:04:14.479 the first things that really stood out to you 00:04:14.479 --> 00:04:16.540 in your research, or how does that come through? 00:04:17.720 --> 00:04:21.459 No, I think it is somehow related. As you know, 00:04:21.560 --> 00:04:24.779 we have developed in my lab since beginning my 00:04:24.779 --> 00:04:28.399 career here. a large number of decision support 00:04:28.399 --> 00:04:32.079 tools that actually combine the latest scientific 00:04:32.079 --> 00:04:35.120 knowledge with the economics and with the decision 00:04:35.120 --> 00:04:37.959 making on farms, trying to do something that 00:04:37.959 --> 00:04:46.360 is very usable and easy to interpret tools for 00:04:46.360 --> 00:04:49.339 dairy producers or dairy decision makers, right? 00:04:49.480 --> 00:04:53.100 We have worked heavily on that, but as we were 00:04:53.100 --> 00:04:58.189 working on those, one of the limitations we encounter 00:04:58.189 --> 00:05:03.009 over and over even to now is the fact that this 00:05:03.009 --> 00:05:07.389 in order to work well in the production systems 00:05:07.389 --> 00:05:12.790 they need to be basically customized to the characteristics 00:05:12.790 --> 00:05:16.930 of each one of these systems and therefore they 00:05:16.930 --> 00:05:21.410 need data and the data in some cases could be 00:05:22.110 --> 00:05:26.029 more general uh if you have or want more specific 00:05:26.029 --> 00:05:29.829 decision making uh the more data you feed to 00:05:29.829 --> 00:05:32.089 these tools the better would be the decision 00:05:32.089 --> 00:05:37.189 making right so uh our constraint was and it 00:05:37.189 --> 00:05:39.529 still is the fact that we need to extract the 00:05:39.529 --> 00:05:43.389 data and many times the extraction is not only 00:05:43.389 --> 00:05:46.449 from one system we need several systems within 00:05:46.449 --> 00:05:49.449 the farm we need to combine the data maybe in 00:05:49.449 --> 00:05:53.329 our spreadsheet and then we feed the data to 00:05:53.329 --> 00:05:58.089 the tools to work better right and then we figure 00:05:58.089 --> 00:06:00.529 out we say i mean this data is already available 00:06:00.529 --> 00:06:03.589 on the farms but it's so fragmented that it's 00:06:03.589 --> 00:06:06.610 difficult to put it together right and nowadays 00:06:06.610 --> 00:06:11.209 we have so many systems that could help us on 00:06:11.209 --> 00:06:14.949 that right and so we can write scripts that combines 00:06:14.949 --> 00:06:19.620 the management software with the fit software 00:06:19.620 --> 00:06:22.300 together with the genetic software and with the 00:06:22.300 --> 00:06:25.079 data that comes directly from the dhi companies 00:06:25.079 --> 00:06:29.579 and so that was actually the division and the 00:06:29.579 --> 00:06:32.180 start of this daily brain project we are running 00:06:32.180 --> 00:06:36.579 that is try to integrate these data sources uh 00:06:36.579 --> 00:06:40.920 close to real time as the data being generated 00:06:40.920 --> 00:06:44.920 they should be uh integrated and that in our 00:06:44.920 --> 00:06:50.720 opinion will bring much and better insights to 00:06:50.720 --> 00:06:53.319 the decision -maker that are not available nowadays. 00:06:54.680 --> 00:06:57.759 Now, there is a lot of, and we get it especially 00:06:57.759 --> 00:07:00.839 online, we get a lot of things when Peters talk 00:07:00.839 --> 00:07:04.420 about intuition and what they've all felt and 00:07:04.420 --> 00:07:09.160 what they've learned growing up versus the data. 00:07:09.420 --> 00:07:11.339 And they kind of do, I know the saying is kind 00:07:11.339 --> 00:07:13.879 of the eye of the master fattened cow, meaning 00:07:13.879 --> 00:07:18.550 that it's the farmer. needs to be kind of it's 00:07:18.550 --> 00:07:21.209 a good farmer who does it and your kind of research 00:07:21.209 --> 00:07:23.829 from my findings and when we've leveraged a lot 00:07:23.829 --> 00:07:26.170 of your research you essentially kind of say 00:07:26.170 --> 00:07:28.550 that well no it's a combo of both but especially 00:07:28.550 --> 00:07:32.189 a lot of using the data have you had a lot of 00:07:32.189 --> 00:07:34.149 times or is there an instance where you found 00:07:34.149 --> 00:07:36.750 when you were talking with producers who maybe 00:07:36.750 --> 00:07:38.449 didn't believe in data when you've had to convince 00:07:38.449 --> 00:07:40.350 them and it's worked well and how do you talk 00:07:40.350 --> 00:07:44.470 to them about that i i think the best way uh 00:07:45.230 --> 00:07:49.149 In my experience, working with producers and 00:07:49.149 --> 00:07:54.149 decision makers in general is showing specific 00:07:54.149 --> 00:08:00.250 examples, right? And hopefully previous farmers, 00:08:00.389 --> 00:08:03.610 how they made decisions, how they changed what 00:08:03.610 --> 00:08:07.509 they were doing because they were able to insert 00:08:07.509 --> 00:08:10.750 data and make more data -driven decision making. 00:08:11.519 --> 00:08:16.759 And we do that, I mean, in the areas of reproduction 00:08:16.759 --> 00:08:23.980 or feed efficiency or maybe on the, I mean, one 00:08:23.980 --> 00:08:26.600 specific example that I think it's difficult 00:08:26.600 --> 00:08:34.399 to grasp for anyone is the replacement problem 00:08:34.399 --> 00:08:38.299 on dairy herds, for example, right? We tend to 00:08:38.299 --> 00:08:44.429 think normally. on the instant reward. If I'm 00:08:44.429 --> 00:08:47.889 going to call this cow, it's going to be a huge 00:08:47.889 --> 00:08:51.190 expense on the farm because even though I'm going 00:08:51.190 --> 00:08:53.690 to receive some salvage value by selling this 00:08:53.690 --> 00:08:57.110 animal, I will need to replace this animal, which 00:08:57.110 --> 00:08:59.870 is much more expensive than what I will receive 00:08:59.870 --> 00:09:03.269 as a salvage value. So it is a net negative, 00:09:03.529 --> 00:09:07.549 a huge net negative. However, if we look in the 00:09:07.549 --> 00:09:12.960 long term, if we project, with the best modeling 00:09:12.960 --> 00:09:16.220 that we can do nowadays. And actually, the farmer 00:09:16.220 --> 00:09:18.580 or the decision maker doesn't need to do any 00:09:18.580 --> 00:09:21.820 modeling. Just to trust what the models will 00:09:21.820 --> 00:09:24.879 do, it will show that many times it's worthwhile 00:09:24.879 --> 00:09:27.960 to do that transaction of replacement because 00:09:27.960 --> 00:09:32.500 in the long term, that means a better net return 00:09:32.500 --> 00:09:35.620 overall on the herd. You have a space on the 00:09:35.620 --> 00:09:39.379 farm, and you need to use that space, that lot. 00:09:39.980 --> 00:09:42.899 with the best animal you can in the long term 00:09:42.899 --> 00:09:47.580 not today right because and and that in order 00:09:47.580 --> 00:09:51.659 to to convey that idea it's it's difficult because 00:09:51.659 --> 00:09:55.720 it's difficult to convince a farmer that has 00:09:55.720 --> 00:09:59.659 a cow that is producing 50 pounds of milk uh 00:09:59.659 --> 00:10:06.899 to replace her right uh because uh i mean i'm 00:10:06.899 --> 00:10:10.940 gonna be uh and It's not only that I'm going 00:10:10.940 --> 00:10:12.899 to be losing that production, I'm going to be 00:10:12.899 --> 00:10:18.340 spending time on doing this. But if we look as 00:10:18.340 --> 00:10:22.240 a business proposition in the long term, maybe 00:10:22.240 --> 00:10:24.980 it is worth it. But the only way to convince 00:10:24.980 --> 00:10:29.200 that is with data. And when we do the simulation, 00:10:29.620 --> 00:10:34.139 they may still think, this is a simulation, this 00:10:34.139 --> 00:10:35.860 is a forecast, I don't know if that's going to 00:10:35.860 --> 00:10:41.039 happen. We can demonstrate that, and that's what 00:10:41.039 --> 00:10:43.940 we're trying to do now, actually going backwards, 00:10:44.080 --> 00:10:47.360 looking at the data and see what has happened 00:10:47.360 --> 00:10:50.879 and what could have happened, right? Or we have 00:10:50.879 --> 00:10:53.399 the opposite as well. In this situation, the 00:10:53.399 --> 00:10:57.600 decision was made, and this is what we have seen 00:10:57.600 --> 00:11:02.940 years afterward, right? One of my favorite quotes, 00:11:03.059 --> 00:11:05.820 and I don't know if it's verbatim, but that you 00:11:05.820 --> 00:11:08.769 have is kind of, Data is a liability until it 00:11:08.769 --> 00:11:11.750 becomes a decision. And I really like that because 00:11:11.750 --> 00:11:13.309 it's something we kind of try to do a lot here. 00:11:13.590 --> 00:11:15.690 It's one thing to have the data, but how do you 00:11:15.690 --> 00:11:17.629 make decisions and how do you work with that? 00:11:18.610 --> 00:11:22.330 In a world where everyone worships data, but 00:11:22.330 --> 00:11:24.769 it's only as good as how it helps you make decisions, 00:11:25.230 --> 00:11:28.570 how do you find that that does help decisions 00:11:28.570 --> 00:11:30.690 and how do you find that people should use the 00:11:30.690 --> 00:11:34.759 data to make those decisions? Sure, I completely 00:11:34.759 --> 00:11:37.000 agree with the quote that you mentioned. I mean, 00:11:37.019 --> 00:11:41.500 data by itself has no value, right? The data 00:11:41.500 --> 00:11:46.539 needs to become information first, and after 00:11:46.539 --> 00:11:50.340 information needs to be translated into actionable 00:11:50.340 --> 00:11:54.610 decisions in order to have a value. And I think 00:11:54.610 --> 00:11:57.830 one important point we made with these tools 00:11:57.830 --> 00:12:01.309 and the daily brain and other things we are working, 00:12:01.570 --> 00:12:04.230 another thing big we are working is this ruminant 00:12:04.230 --> 00:12:07.610 farm system model. It's the idea of actually 00:12:07.610 --> 00:12:14.070 to convert the data into specific actions on 00:12:14.070 --> 00:12:20.570 the farm that will have a very concrete outcome 00:12:20.570 --> 00:12:26.750 with normally, with an economic benefit. I mean, 00:12:26.789 --> 00:12:30.090 I think also thinking in the future, these benefits 00:12:30.090 --> 00:12:34.429 should also translate to other things like environmental 00:12:34.429 --> 00:12:39.070 stewardship or welfare of the animals. I mean, 00:12:39.090 --> 00:12:41.330 that's coming, and we need to be conscious about 00:12:41.330 --> 00:12:44.649 those as well. But in general, I'm talking mostly 00:12:44.649 --> 00:12:47.509 because I'm related with the North American dairy 00:12:47.509 --> 00:12:50.250 industry. I have been lucky enough to see other 00:12:50.250 --> 00:12:54.519 parts of the world. In North America, specifically 00:12:54.519 --> 00:12:57.860 U .S., the bottom line is king, right? I mean, 00:12:57.879 --> 00:13:02.019 what it moves the industry is the net return 00:13:02.019 --> 00:13:06.519 at the end. And based on that, actually, we can 00:13:06.519 --> 00:13:12.860 promote those things that are economical. They 00:13:12.860 --> 00:13:15.340 have strong economic benefits, but at the same 00:13:15.340 --> 00:13:21.340 time, will likely... will relate to other benefits 00:13:21.340 --> 00:13:25.519 as well. Now, you've mentioned a couple times 00:13:25.519 --> 00:13:27.460 about the dairy brain, and I'm really fascinated 00:13:27.460 --> 00:13:30.600 by this. Could you maybe just give us some background 00:13:30.600 --> 00:13:34.200 on how it came about and what it does and why 00:13:34.200 --> 00:13:37.740 producers should be wanting to learn more about 00:13:37.740 --> 00:13:43.820 it? Sure. I mean, in a nutshell, the dairy brain, 00:13:44.019 --> 00:13:52.350 luckily, the name has had quite a visibility 00:13:52.350 --> 00:13:56.129 because it means what it says. It is providing 00:13:56.129 --> 00:14:00.990 brain to the dairy based on data -driven decisions, 00:14:01.350 --> 00:14:05.950 right? The bottom line of the dairy brain is 00:14:05.950 --> 00:14:10.389 basically help producers to make much better 00:14:10.389 --> 00:14:13.009 informed decisions on their farms. That's the 00:14:13.009 --> 00:14:16.720 bottom line. Now, to reach that point, which 00:14:16.720 --> 00:14:21.960 has a strong value proposition we do need to 00:14:21.960 --> 00:14:28.080 move backwards to the data and in order to get 00:14:28.080 --> 00:14:31.120 the data in the right position we need to integrate 00:14:31.120 --> 00:14:35.240 the data from different sources and we know the 00:14:35.240 --> 00:14:40.740 industry has data very fragmented and maybe there 00:14:40.740 --> 00:14:43.519 are other industries as well, but particularly 00:14:43.519 --> 00:14:46.440 in the dairy industry, we have silos of data 00:14:46.440 --> 00:14:49.779 that they don't talk to each other. It is not 00:14:49.779 --> 00:14:52.500 uncommon to reach a farm that will have several 00:14:52.500 --> 00:14:54.779 computers, each one dedicated to a different 00:14:54.779 --> 00:14:58.559 software, for example, right? Which is, if you 00:14:58.559 --> 00:15:01.580 think out of the industry, it's kind of surprising 00:15:01.580 --> 00:15:06.100 because in order to to have a list of animals 00:15:06.100 --> 00:15:09.720 for a purpose you need to print or extract data 00:15:09.720 --> 00:15:12.740 from one software from another and manually or 00:15:12.740 --> 00:15:15.620 or yeah manually most of the time you need to 00:15:15.620 --> 00:15:18.840 connect the data to do the decisions and it is 00:15:18.840 --> 00:15:21.919 uh in these times it is actually inefficient 00:15:21.919 --> 00:15:27.360 and prone to errors and actually it it precludes 00:15:27.360 --> 00:15:32.139 to do more sophisticated advanced decision making 00:15:32.139 --> 00:15:36.320 now we can use all these analytics that are artificial 00:15:36.320 --> 00:15:39.360 intelligence and machine learning and I mean 00:15:39.360 --> 00:15:42.600 it's not the matter of just using them but in 00:15:42.600 --> 00:15:45.100 order to use these sophisticated analytics we 00:15:45.100 --> 00:15:48.200 need the data right and in order to provide that 00:15:48.200 --> 00:15:52.480 additional value to the information we need to 00:15:52.480 --> 00:15:56.799 have all the data integrated so basically a big 00:15:56.799 --> 00:16:01.279 part of the daily brain consists on extracting 00:16:01.279 --> 00:16:04.620 the data efficiently of these systems and put 00:16:04.620 --> 00:16:06.840 it together in a system that could be usable 00:16:06.840 --> 00:16:12.879 for decision making. Now, at the same time, because 00:16:12.879 --> 00:16:18.799 I'm going to be completely open as well, this 00:16:18.799 --> 00:16:22.120 part of the integration of the data, we never 00:16:22.120 --> 00:16:26.159 expected it would be so much time and effort 00:16:26.159 --> 00:16:31.629 demanding. And that's actually what has been 00:16:31.629 --> 00:16:36.009 the bottleneck of the system. That's the part 00:16:36.009 --> 00:16:38.509 that is taking more time and effort to do it. 00:16:38.730 --> 00:16:46.190 But then after that, we need the analytics. And 00:16:46.190 --> 00:16:49.210 actually, even before the data was available, 00:16:49.730 --> 00:16:52.570 we wanted to demonstrate that the analytics are 00:16:52.570 --> 00:16:56.629 important. and provide value to the system. So 00:16:56.629 --> 00:16:59.870 we have published a number of papers showing 00:16:59.870 --> 00:17:02.669 that, demonstrating that once we have the data 00:17:02.669 --> 00:17:05.390 available, this is all what we can do with this 00:17:05.390 --> 00:17:08.349 data. This is what it's going to provide, this 00:17:08.349 --> 00:17:11.069 integrated data. Not only this integrated data, 00:17:11.089 --> 00:17:14.890 but this pipeline, constant flow of integrated 00:17:14.890 --> 00:17:18.630 data. It was interesting. I think about a month 00:17:18.630 --> 00:17:20.349 ago, maybe a little more than a month ago, I 00:17:20.349 --> 00:17:23.890 saw out of Europe that they were establishing, 00:17:24.539 --> 00:17:26.940 I'll call it guidelines or protocols for how 00:17:26.940 --> 00:17:29.720 data should be shared on the farm. And so that 00:17:29.720 --> 00:17:32.960 you can find a unified data set, which given 00:17:32.960 --> 00:17:34.960 what you've been doing, if you could ever be 00:17:34.960 --> 00:17:37.799 able to have a unified way to transfer data would 00:17:37.799 --> 00:17:39.920 make life a lot easier. I hope we're getting 00:17:39.920 --> 00:17:42.259 there here in North America quicker, but I know 00:17:42.259 --> 00:17:44.460 we're not there yet. Actually, yeah, we're not 00:17:44.460 --> 00:17:48.140 there yet, but it will be coming hopefully soon. 00:17:49.880 --> 00:17:51.839 One of my questions for you, and that's one thing 00:17:51.839 --> 00:17:55.660 to talk about this on a level, Maybe if you could 00:17:55.660 --> 00:17:57.480 give me an example. Let's say if we had a cow 00:17:57.480 --> 00:18:01.839 that came up with early mastitis, how would that 00:18:01.839 --> 00:18:04.519 work in the dairy brain, or how would the system 00:18:04.519 --> 00:18:07.160 handle that if a cow in one system was showing 00:18:07.160 --> 00:18:09.380 it, or how does that system work that way, say, 00:18:09.480 --> 00:18:13.920 for a cow with early mastitis signs? Actually, 00:18:14.019 --> 00:18:19.539 we would go even before that. We have a system, 00:18:19.599 --> 00:18:25.500 for example. We have work with data. of the integrated 00:18:25.500 --> 00:18:31.380 data in the dairy brain ecosystem that will give 00:18:31.380 --> 00:18:36.240 us a chance of a specific animal would have higher 00:18:36.240 --> 00:18:39.720 probabilities of having mastitis during the first 00:18:39.720 --> 00:18:42.599 lactation, for example, right? We can use genetic 00:18:42.599 --> 00:18:47.299 information on the animal. We can find information 00:18:47.299 --> 00:18:50.140 of previous events during the raising of the 00:18:50.140 --> 00:18:54.539 animal. And that gives us a chance compared to 00:18:54.539 --> 00:18:58.859 the mates if this animal would have higher probability 00:18:58.859 --> 00:19:03.579 of having mastitis. On top of that, while she 00:19:03.579 --> 00:19:06.960 is starting the production life, as we know the 00:19:06.960 --> 00:19:10.319 mastitis will not likely happen before the production, 00:19:10.579 --> 00:19:12.980 when she starts the production life, whether 00:19:12.980 --> 00:19:15.299 it is the first lactation or the second or third 00:19:15.299 --> 00:19:18.660 or later lactation, we have additional information 00:19:18.660 --> 00:19:22.339 that is very constantly flowing to the system, 00:19:22.380 --> 00:19:25.500 which is the milking, basically, right? In addition 00:19:25.500 --> 00:19:29.119 to previous events. I mean, the fresh period, 00:19:29.279 --> 00:19:31.480 the transition period is very important, and 00:19:31.480 --> 00:19:34.079 we put that together as well into the system. 00:19:34.119 --> 00:19:37.119 But one important thing we have found, for example, 00:19:37.200 --> 00:19:42.940 is the very specific information we would have, 00:19:43.019 --> 00:19:46.000 the data we have of the milking. In every milking 00:19:46.000 --> 00:19:50.250 we have, the speed of the flow of the milk we 00:19:50.250 --> 00:19:53.349 can have even the conductivity of the milk when 00:19:53.349 --> 00:19:57.009 it's been uh milk and and if you put on that 00:19:57.009 --> 00:19:59.869 together other information like the i mentioned 00:19:59.869 --> 00:20:03.529 before the genetics previous events and we can 00:20:03.529 --> 00:20:06.529 have additional characteristics of the earth 00:20:06.529 --> 00:20:09.710 that will also impact the animal we can have 00:20:09.710 --> 00:20:15.579 days before the probability of an animal getting 00:20:15.579 --> 00:20:18.460 sick. We can have a very good prediction. I'm 00:20:18.460 --> 00:20:22.160 talking about more than 90 % of accuracy that 00:20:22.160 --> 00:20:25.339 will give us a flag, say, this animal thinks 00:20:25.339 --> 00:20:30.460 it's going to get sick next week. So if we receive 00:20:30.460 --> 00:20:35.859 that, we can take it, right? We can do some things 00:20:35.859 --> 00:20:38.880 on the farm. At least the first thing would be 00:20:38.880 --> 00:20:44.519 try to find the animal. and try to monitor that 00:20:44.519 --> 00:20:47.200 animal more closely to see what's happening. 00:20:47.579 --> 00:20:51.319 There may be some alternatives to try to prevent 00:20:51.319 --> 00:20:59.180 that through nutrition interventions, or we can 00:20:59.180 --> 00:21:03.160 do prevention medicine as well. That could be 00:21:03.160 --> 00:21:09.599 done. So I think we should think even in that 00:21:09.599 --> 00:21:13.539 direction in which... The same as in humans. 00:21:13.740 --> 00:21:17.140 I think that's something that's kind of in the 00:21:17.140 --> 00:21:19.940 next generation we should try to prevent before 00:21:19.940 --> 00:21:24.059 it happens. We should not try to cure what has 00:21:24.059 --> 00:21:27.119 already happened. If it has happened, it will 00:21:27.119 --> 00:21:31.319 also give additional information of what we could 00:21:31.319 --> 00:21:35.960 do to try to figure out what would be the best 00:21:35.960 --> 00:21:38.539 outcomes we could have based on the data once 00:21:38.539 --> 00:21:44.599 again. I'm having a little feedback. I don't 00:21:44.599 --> 00:21:50.779 know if that's me or... Let me see. I can try 00:21:50.779 --> 00:21:52.980 to turn mine down a bit. Maybe it's maybe me. 00:21:53.700 --> 00:21:59.720 Okay. Any better? Yes, it seems like. Okay. Now, 00:22:00.019 --> 00:22:01.940 I did mention before, and I've had experience 00:22:01.940 --> 00:22:05.180 in the past, you're talking about many systems 00:22:05.180 --> 00:22:09.089 having to work together. But you're also talking 00:22:09.089 --> 00:22:11.369 about many competitors because, you know, you 00:22:11.369 --> 00:22:15.289 have Lely or DeLaval and Gia and all the different 00:22:15.289 --> 00:22:18.650 milking systems. And then you've got like Dairy 00:22:18.650 --> 00:22:21.630 Comp and all these different systems. And you're 00:22:21.630 --> 00:22:25.190 wanting to feed it all into your data hub. How 00:22:25.190 --> 00:22:27.750 did those conversations start and how did you 00:22:27.750 --> 00:22:30.630 get them to start working with this or how has 00:22:30.630 --> 00:22:34.509 that been going? Yeah, that's a very, very good 00:22:34.509 --> 00:22:38.640 question and a very good. I do think in a certain 00:22:38.640 --> 00:22:41.920 way we need to provide value to every player 00:22:41.920 --> 00:22:45.380 into the ecosystem, starting from the farmer 00:22:45.380 --> 00:22:48.920 to any one of the companies involved, right? 00:22:49.119 --> 00:22:55.059 I do think there is commercial economic interest 00:22:55.059 --> 00:22:59.319 of companies, which is understandable that they 00:22:59.319 --> 00:23:01.660 would like to dominate the system. We have seen 00:23:01.660 --> 00:23:07.430 this in other industries. are Apple user or you 00:23:07.430 --> 00:23:11.650 are a Windows user, kind of like that. I mean, 00:23:11.690 --> 00:23:15.069 in a certain way, it's happening. There are companies 00:23:15.069 --> 00:23:20.190 that have more entry into the market level of 00:23:20.190 --> 00:23:23.589 the farms, and it would not be rare already now 00:23:23.589 --> 00:23:26.390 that they will choose a new technology based 00:23:26.390 --> 00:23:28.990 on what they already have there to see if there 00:23:28.990 --> 00:23:32.349 is compatibility among those systems, right? 00:23:33.150 --> 00:23:36.930 Staying on that, I mean, in our experience within 00:23:36.930 --> 00:23:40.049 the DERIC project, one thing that has been key 00:23:40.049 --> 00:23:43.829 is the fact that we have the buyout of the producer. 00:23:43.910 --> 00:23:46.329 The producer is the consumer of the technology, 00:23:46.549 --> 00:23:52.589 and they have the say to share the data, right? 00:23:52.710 --> 00:23:56.009 And so they have the capability to say to the 00:23:56.009 --> 00:23:59.769 company, whether it is DERICOMP or whether it 00:23:59.769 --> 00:24:05.910 is GIA or Baumatic, I want to share my data with 00:24:05.910 --> 00:24:09.630 these researchers, in this case, to work on this 00:24:09.630 --> 00:24:16.630 project. And so there it comes another layer 00:24:16.630 --> 00:24:20.430 of complexity yet, because the companies, in 00:24:20.430 --> 00:24:25.490 our opinion, in our experience, they will comply 00:24:25.490 --> 00:24:28.470 with that request. They will try to do their 00:24:28.470 --> 00:24:34.089 best in general. However, in our experience, 00:24:34.170 --> 00:24:37.609 once again, they can make it very easy to consume 00:24:37.609 --> 00:24:40.289 the data, or they can make it extremely difficult, 00:24:40.549 --> 00:24:47.390 actually, right? So, I mean, sometimes for us, 00:24:47.490 --> 00:24:50.589 the ideal situation would be they would provide 00:24:50.589 --> 00:24:53.529 APIs, which is these application programming 00:24:53.529 --> 00:24:57.940 interfaces. We use nowadays every day in our 00:24:57.940 --> 00:25:00.720 smartphones to do a bank transaction. We are 00:25:00.720 --> 00:25:03.799 sharing data. And even though that's very sensitive 00:25:03.799 --> 00:25:08.380 data, it can be securely transferred. Our health 00:25:08.380 --> 00:25:12.960 data, we can actually specifically say, I want 00:25:12.960 --> 00:25:15.740 to share my data from this clinic with this pharmacy, 00:25:15.920 --> 00:25:19.900 and it is securely done without no problems throughout 00:25:19.900 --> 00:25:24.000 API. So what I'm trying to, the point I'm making 00:25:24.000 --> 00:25:28.470 is, It is doable. Technically, it's possible 00:25:28.470 --> 00:25:34.349 and it is happening as we speak. Now, if you 00:25:34.349 --> 00:25:37.269 go to the daily industry and you look for APIs, 00:25:37.750 --> 00:25:43.670 there are not many. I would start saying there 00:25:43.670 --> 00:25:46.609 are not many companies that have, I mean, I would 00:25:46.609 --> 00:25:49.069 say the big majority, they will not provide API. 00:25:49.390 --> 00:25:54.960 The ones who are... having some APIs development 00:25:54.960 --> 00:25:59.740 are in the development stage, in our opinion. 00:25:59.940 --> 00:26:02.339 I was thinking as well, entering into this, I'm 00:26:02.339 --> 00:26:05.000 not a programmer or computer scientist. I was 00:26:05.000 --> 00:26:10.240 thinking it is binary. Yes, API or not, it's 00:26:10.240 --> 00:26:13.539 not like that. It's a spectrum, right? So I would 00:26:13.539 --> 00:26:16.359 say even those companies that are developing 00:26:16.359 --> 00:26:20.059 APIs in the data industry are not at the level 00:26:20.059 --> 00:26:23.700 of usability that we would have. wanted to have. 00:26:23.900 --> 00:26:27.640 So we have created internally our own APIs with 00:26:27.640 --> 00:26:31.380 the data we extract, right? But it's a system 00:26:31.380 --> 00:26:34.740 that's still very fragile because whether we 00:26:34.740 --> 00:26:40.319 receive a CSV, in some cases we receive spreadsheets, 00:26:40.319 --> 00:26:44.039 we have extracted data even from PDF documents, 00:26:44.240 --> 00:26:47.079 for example, and you know that's very fragile 00:26:47.079 --> 00:26:51.019 because if in the spreadsheet they change a cell 00:26:51.519 --> 00:26:54.599 location, it will screw the whole thing and we 00:26:54.599 --> 00:26:58.420 need to do an investigative work to find out 00:26:58.420 --> 00:27:01.759 where the problem was and try to recover and 00:27:01.759 --> 00:27:08.960 redo the system, right? So those are challenges 00:27:08.960 --> 00:27:13.200 that we are still living and I think we need 00:27:13.200 --> 00:27:18.839 to overcome those. And I think one way to do 00:27:18.839 --> 00:27:21.980 that is What I mentioned at the beginning, I 00:27:21.980 --> 00:27:25.160 think if everyone feels confident that there 00:27:25.160 --> 00:27:28.440 is going to be value in the system for each one 00:27:28.440 --> 00:27:31.180 of them, it would be a much easier proposition 00:27:31.180 --> 00:27:36.819 to everyone, right? Yeah, you mentioned at the 00:27:36.819 --> 00:27:39.079 beginning another important thing. I'm part also 00:27:39.079 --> 00:27:42.019 of groups that are working on standards like 00:27:42.019 --> 00:27:47.839 ICAR or IDF in the dairy industry. working on, 00:27:47.880 --> 00:27:50.819 for example, in this case, the one I'm more involved 00:27:50.819 --> 00:27:54.720 in is about standardization of data from health 00:27:54.720 --> 00:27:57.539 monitoring systems like sensors. And, I mean, 00:27:57.599 --> 00:28:01.240 it's very interesting. And ICAR is now a big 00:28:01.240 --> 00:28:03.640 institution and has the standard there. They 00:28:03.640 --> 00:28:07.200 have a GitHub. But that's one thing. And the 00:28:07.200 --> 00:28:10.160 other thing is which companies are actually using 00:28:10.160 --> 00:28:17.390 them, right? That's a different story. are willing 00:28:17.390 --> 00:28:20.430 to take the time to understand those and implement 00:28:20.430 --> 00:28:24.089 on their own systems, right? And if we go even 00:28:24.089 --> 00:28:27.950 further, the other thing, I mean, we have talked 00:28:27.950 --> 00:28:30.569 about a lot of companies and we agree on that, 00:28:30.690 --> 00:28:38.509 but also there should be another layer of data 00:28:38.509 --> 00:28:41.210 education, I would say, at the farm level as 00:28:41.210 --> 00:28:45.579 well. I do think farmers could play a role. into 00:28:45.579 --> 00:28:48.359 the the system mostly in the cases that they 00:28:48.359 --> 00:28:52.099 have the opportunity and the capacity to enter 00:28:52.099 --> 00:28:55.980 their own data in some way or shape right like 00:28:55.980 --> 00:28:58.420 for example mastitis that we discussed earlier 00:28:58.420 --> 00:29:01.700 they can name different names to the same disease 00:29:01.700 --> 00:29:06.119 they don't have a standard even inside the farm 00:29:06.119 --> 00:29:10.039 everyone that records a mastitis event will mark 00:29:10.039 --> 00:29:13.900 differently in the system and even the same person 00:29:13.900 --> 00:29:16.859 across time will have different names, that makes 00:29:16.859 --> 00:29:19.599 a nightmare in order to process the data later 00:29:19.599 --> 00:29:23.079 on, right? And you've been doing it for several 00:29:23.079 --> 00:29:26.799 years now. Can you tell me what the experience 00:29:26.799 --> 00:29:28.859 on -farm has been with it and how it's been going? 00:29:30.059 --> 00:29:34.579 Yes, I think we had a very good reception. I 00:29:34.579 --> 00:29:38.500 think farmers understand what the idea and the 00:29:38.500 --> 00:29:44.150 concept is. and also the potential value of it, 00:29:44.309 --> 00:29:49.589 right? So we haven't been able yet to scale that 00:29:49.589 --> 00:29:53.150 up, but the farms we have worked with, actually, 00:29:53.269 --> 00:29:56.529 we have been able to show that value out there. 00:29:56.650 --> 00:29:59.230 And that has been, I think, the most exciting 00:29:59.230 --> 00:30:03.450 part of the dairy brain because there has been, 00:30:03.609 --> 00:30:07.150 and still there is a lot of work which is under 00:30:07.150 --> 00:30:09.549 the hood, right? You don't see. basically results 00:30:09.549 --> 00:30:12.250 on all these scripts just to get the data ready 00:30:12.250 --> 00:30:17.410 for. Once that gets out there into a system that 00:30:17.410 --> 00:30:20.089 is very, I mean, an online system in which the 00:30:20.089 --> 00:30:22.569 farmer can see real time what's going on the 00:30:22.569 --> 00:30:25.730 farm, we can calculate a simple thing like income 00:30:25.730 --> 00:30:29.279 or feed costs. If you go to a farm and you ask 00:30:29.279 --> 00:30:32.579 for feed efficiency or income or feed cost, they 00:30:32.579 --> 00:30:35.359 will say, yeah, I know this is my average, but 00:30:35.359 --> 00:30:38.779 do you know group by group, even cow by cow, 00:30:38.859 --> 00:30:42.519 if that's possible? And every day, that's a different 00:30:42.519 --> 00:30:44.539 story, right? They will say, oh, I can calculate. 00:30:44.640 --> 00:30:47.099 I can download here, download there, and do this 00:30:47.099 --> 00:30:50.640 and do that. But that's very different than habit. 00:30:51.130 --> 00:30:55.150 ready and see the evolution of that day to day. 00:30:55.289 --> 00:30:58.509 So, for example, in the farm, we were able to 00:30:58.509 --> 00:31:03.369 pinpoint even the small change in the silage, 00:31:03.470 --> 00:31:06.289 how that's impacting feed efficiency and income 00:31:06.289 --> 00:31:11.069 immediately. So you can even prevent or you can 00:31:11.069 --> 00:31:16.309 take action to counteract what's going on at 00:31:16.309 --> 00:31:18.609 those points in time, right? Because you are 00:31:18.609 --> 00:31:21.819 seeing almost at the time. things are happening 00:31:21.819 --> 00:31:25.380 on the farm yeah it was interesting i was uh 00:31:25.380 --> 00:31:28.920 looking and it was farm fit data so st's product 00:31:28.920 --> 00:31:33.059 um where they had the boluses and the real -time 00:31:33.059 --> 00:31:35.660 data and you can see long before you ever would 00:31:35.660 --> 00:31:38.000 see it in the bulk tank when a cow maybe went 00:31:38.000 --> 00:31:40.579 off feed or a cow did something like that and 00:31:40.579 --> 00:31:42.519 you have that data right then but the challenge 00:31:42.519 --> 00:31:44.380 and this is where i think the dairy brain comes 00:31:44.380 --> 00:31:46.559 in and other ones like that is the challenge 00:31:46.559 --> 00:31:48.900 is it's in one spot unless someone's monitoring 00:31:48.900 --> 00:31:52.630 it 24 -7, but they're also having to monitor 00:31:52.630 --> 00:31:55.329 all these different spots all at once. It's very 00:31:55.329 --> 00:31:58.730 hard to stay on top of it all. And that's where 00:31:58.730 --> 00:32:00.990 I think this integration is where really the 00:32:00.990 --> 00:32:04.410 future is, because what is the limitation or 00:32:04.410 --> 00:32:06.430 what has been the technical challenge to stopping 00:32:06.430 --> 00:32:08.970 you guys from being able to scale and work on 00:32:08.970 --> 00:32:16.069 a large form? I mean, I think there are challenges 00:32:16.069 --> 00:32:22.019 on the technical part, for sure. which are uh 00:32:22.019 --> 00:32:28.140 doable they are uh we can overcome those i think 00:32:28.140 --> 00:32:31.339 the the technology is improving so fast that 00:32:31.339 --> 00:32:35.019 they should be i mean i'm thinking on for example 00:32:35.019 --> 00:32:38.900 even from connectivity on farms uh internet access 00:32:38.900 --> 00:32:43.339 that's changing fast and they are having more 00:32:43.339 --> 00:32:48.000 and more uh access to that. That helps a lot. 00:32:48.559 --> 00:32:52.160 From the updates on the systems on the farms 00:32:52.160 --> 00:32:55.220 as well, the computer systems, we have had challenges 00:32:55.220 --> 00:32:59.339 with using different Windows systems in different 00:32:59.339 --> 00:33:02.240 computers on the same farm, for example, and 00:33:02.240 --> 00:33:06.240 losing connectivity when they are doing updates 00:33:06.240 --> 00:33:11.119 on the system. But those things are... Well, 00:33:11.160 --> 00:33:13.599 that's – and then I think we have different versions 00:33:13.599 --> 00:33:16.180 of the software of the company in different forms 00:33:16.180 --> 00:33:20.259 as well. But those, I think, are – we can overcome 00:33:20.259 --> 00:33:26.680 without no much problem. We have the scripts, 00:33:26.920 --> 00:33:31.160 and we expect hopefully the upcoming year to 00:33:31.160 --> 00:33:34.640 release as an open source, actually. That's kind 00:33:34.640 --> 00:33:38.750 of – we have decided that – I mean, this is the 00:33:38.750 --> 00:33:43.970 other thing. We live based on grants because 00:33:43.970 --> 00:33:47.049 we are an academic institution without commercial 00:33:47.049 --> 00:33:52.269 purposes, right? So that's one distinction. It 00:33:52.269 --> 00:33:56.390 has its own benefits, but it has its drawbacks, 00:33:56.509 --> 00:33:58.950 right? I mean, the companies have another type 00:33:58.950 --> 00:34:00.930 of interest, but they have another power as well 00:34:00.930 --> 00:34:04.750 to do this kind of thing, right? So, and basically 00:34:04.750 --> 00:34:14.690 we have come to, in our grant process and that 00:34:14.690 --> 00:34:23.230 kind of stops what we are working on. However, 00:34:23.670 --> 00:34:26.690 I think... there is a benefit of doing this from 00:34:26.690 --> 00:34:28.730 that standpoint of view rather than from the 00:34:28.730 --> 00:34:31.090 commercial side. And I think there are initiatives 00:34:31.090 --> 00:34:33.190 in the commercial side, in the industry side 00:34:33.190 --> 00:34:36.769 that are worthwhile as well to look at, right? 00:34:37.349 --> 00:34:39.829 But we understand those will have a different 00:34:39.829 --> 00:34:42.889 interest and a different way to share whatever 00:34:42.889 --> 00:34:45.650 they are learning, for example. But on top of 00:34:45.650 --> 00:34:47.650 that, the other thing I think we need to keep 00:34:47.650 --> 00:34:50.559 in mind in all this process is uh what's called 00:34:50.559 --> 00:34:54.079 data governance in general i mean more on the 00:34:54.079 --> 00:34:57.079 on the soft side not the technological part but 00:34:57.079 --> 00:35:00.440 data ownership uh confidentiality of the data 00:35:00.440 --> 00:35:03.760 uh i never uh thought we would spend so much 00:35:03.760 --> 00:35:07.519 time and time and effort thinking and working 00:35:07.519 --> 00:35:10.539 on that and actually we gather a special group 00:35:10.539 --> 00:35:15.380 to discuss this this kind of things how is the 00:35:15.380 --> 00:35:19.230 best way to think about data ownership on the 00:35:19.230 --> 00:35:23.210 farms together with the companies the confidentiality 00:35:23.210 --> 00:35:26.829 of the data how we protect the data of the farmers 00:35:26.829 --> 00:35:31.929 and how we process the data into this system 00:35:31.929 --> 00:35:34.050 so that that's something that we need to keep 00:35:34.050 --> 00:35:36.650 in mind as we move forward as well i don't think 00:35:36.650 --> 00:35:43.099 this is fully solved in other domains but It 00:35:43.099 --> 00:35:46.059 is workable. I mean, we are working every day 00:35:46.059 --> 00:35:49.400 on that, right? I mean, but that's something 00:35:49.400 --> 00:35:54.960 I think we need more express guidelines in the 00:35:54.960 --> 00:35:57.619 data industry. You know, it's been very interesting 00:35:57.619 --> 00:36:00.079 because we've written about it a few times. It's 00:36:00.079 --> 00:36:02.239 who owns data, and it's in many different ways. 00:36:02.599 --> 00:36:05.260 You know, if you take the obvious genomics, the 00:36:05.260 --> 00:36:08.059 genomics comes from the data that farmers provided, 00:36:08.280 --> 00:36:11.659 and yet... now they pay to get their information 00:36:11.659 --> 00:36:14.539 back so that's always been a hot topic but we 00:36:14.539 --> 00:36:18.679 also have it in robots even transitioning the 00:36:18.679 --> 00:36:20.760 same robot system from one farmer to the next 00:36:20.760 --> 00:36:23.760 generation and the actual fees to transition 00:36:23.760 --> 00:36:26.699 who owns the data and who has access to the data 00:36:26.699 --> 00:36:29.599 from the same farm just from one generation to 00:36:29.599 --> 00:36:31.820 the next there's been headaches so this whole 00:36:31.820 --> 00:36:35.380 who owns data is a very challenging subject but 00:36:35.380 --> 00:36:37.099 it's a subject we're going to have to get pretty 00:36:37.099 --> 00:36:40.869 clear on because Pretty much everything is focused 00:36:40.869 --> 00:36:43.369 off of using the data these days and who owns 00:36:43.369 --> 00:36:46.550 it and who has access to it and how do we use 00:36:46.550 --> 00:36:49.690 it. It's going to be a very interesting conversation 00:36:49.690 --> 00:36:52.769 as we go forward, as you know probably better 00:36:52.769 --> 00:36:56.590 than I do. Another interesting subject I saw 00:36:56.590 --> 00:37:00.210 that I've come across on your research was more 00:37:00.210 --> 00:37:04.250 of the part about how higher producing cows actually 00:37:04.250 --> 00:37:07.150 have a smaller carbon footprint per unit of milk 00:37:07.150 --> 00:37:10.559 or for milk production. And I know a lot of people 00:37:10.559 --> 00:37:14.239 always think that bigger is or higher is a bad 00:37:14.239 --> 00:37:17.420 thing. But why would you say that? How would 00:37:17.420 --> 00:37:19.639 you explain to people the kind of sustainability 00:37:19.639 --> 00:37:22.460 paradox that higher producing cows actually are 00:37:22.460 --> 00:37:26.519 better for the environment and why? Well, yeah, 00:37:26.599 --> 00:37:30.019 I think that's a very good point you're raising 00:37:30.019 --> 00:37:34.659 there, because what we should be interested in 00:37:34.659 --> 00:37:39.369 is what we call the. the the emissions intensity 00:37:39.369 --> 00:37:43.650 right which is related to the the value product 00:37:43.650 --> 00:37:47.510 we have from the production in this case milk 00:37:47.510 --> 00:37:51.650 right so i i think it makes all the sense in 00:37:51.650 --> 00:37:55.429 the world to measure the emissions per kilo of 00:37:55.429 --> 00:37:59.750 milk produce right not the total production because 00:37:59.750 --> 00:38:04.030 we do need that milk to be consumed by humans 00:38:04.030 --> 00:38:07.550 right so it is a nutritional product uh we have 00:38:07.550 --> 00:38:11.489 a huge demand so what we want to have is the 00:38:11.489 --> 00:38:15.449 least emissions for each unit of that product 00:38:15.449 --> 00:38:19.090 right now we should recognize and we should actually 00:38:19.090 --> 00:38:22.130 mention this i think it's important for this 00:38:22.130 --> 00:38:26.570 kind of discussion that a higher producing cow 00:38:26.570 --> 00:38:33.239 will consume more feed and will produce overall 00:38:33.239 --> 00:38:40.039 more amount of greenhouse gases, right? Anthetic 00:38:40.039 --> 00:38:43.599 methane and also probably the manure methane 00:38:43.599 --> 00:38:50.199 will be higher, total. But if we measure that 00:38:50.199 --> 00:38:54.119 per unit of milk produced, it will be less, right? 00:38:54.280 --> 00:38:58.880 So there is a marginal gain by producing more 00:38:58.880 --> 00:39:05.530 per cow. And thus, anything we do, whether it 00:39:05.530 --> 00:39:10.050 is at the farm level, at the herd level, or at 00:39:10.050 --> 00:39:14.670 the specific cow level, to improve feed efficiency 00:39:14.670 --> 00:39:19.929 will be a huge benefit to the environment. That's 00:39:19.929 --> 00:39:23.130 for sure. So we can do, I mean, at the cow level, 00:39:23.369 --> 00:39:25.789 we can, nowadays, we have the opportunity to 00:39:25.789 --> 00:39:29.690 select cows for... feed efficiency residual feed 00:39:29.690 --> 00:39:35.329 intake in canada there is the methane um methane 00:39:35.329 --> 00:39:39.949 uh trait uh in the us soon likely and actually 00:39:39.949 --> 00:39:42.570 here in in the department uh there are people 00:39:42.570 --> 00:39:45.809 working on on this and unlikely there would be 00:39:45.809 --> 00:39:48.969 a trait so you can select animals that are more 00:39:48.969 --> 00:39:52.530 efficient in the production with respect to emissions 00:39:52.530 --> 00:39:57.699 but uh Even without having that, we can do a 00:39:57.699 --> 00:40:01.019 lot of stuff to reduce the emissions per unit 00:40:01.019 --> 00:40:04.139 of production, which is being more accurate in 00:40:04.139 --> 00:40:06.719 the feeding of the animals, using nutritional 00:40:06.719 --> 00:40:11.280 grouping that for sure decreases the emissions 00:40:11.280 --> 00:40:13.840 per unit of production. We are going to have 00:40:13.840 --> 00:40:16.639 more feed efficiency overall on the herd, and 00:40:16.639 --> 00:40:19.619 we're going to reduce our emissions per unit 00:40:19.619 --> 00:40:23.659 of milk being produced, right? So, for sure. 00:40:24.059 --> 00:40:28.920 Everything we do to improve the productivity 00:40:28.920 --> 00:40:32.760 of the animals will have a direct impact on reducing 00:40:32.760 --> 00:40:37.559 emissions. And this relates directly to methane 00:40:37.559 --> 00:40:44.280 or greenhouse gases, but that's also true for 00:40:44.280 --> 00:40:46.739 other emissions like nitrogen or phosphorus to 00:40:46.739 --> 00:40:50.630 the environment. It's interesting because from 00:40:50.630 --> 00:40:52.530 the research I've seen and some of it's from 00:40:52.530 --> 00:40:58.670 yours as well is that typically the most efficient 00:40:58.670 --> 00:41:02.610 or sustainable producing cow is usually your 00:41:02.610 --> 00:41:06.190 most economically profitable cow as well. They 00:41:06.190 --> 00:41:09.329 kind of go hand in hand. And I link that back 00:41:09.329 --> 00:41:11.630 to your dairy brain because when we have all 00:41:11.630 --> 00:41:14.190 our data systems together, we probably have a 00:41:14.190 --> 00:41:16.929 clear understanding of just who those cows are 00:41:16.929 --> 00:41:21.070 in the herd. that really are are truly both profitable 00:41:21.070 --> 00:41:25.369 and providing the most sustainable options but 00:41:25.369 --> 00:41:29.110 right now i don't know if we truly on a mass 00:41:29.110 --> 00:41:32.369 scale have a data insight to see it for most 00:41:32.369 --> 00:41:34.590 farmers but i do think it's very interesting 00:41:34.590 --> 00:41:38.050 where how those two correlate yes and we will 00:41:38.050 --> 00:41:42.070 eventually we will get there yes yes yes hopefully 00:41:42.070 --> 00:41:46.079 so yes yes Another one I was looking at, and 00:41:46.079 --> 00:41:48.500 it was another piece of your research I saw that 00:41:48.500 --> 00:41:50.519 I've actually looked quite a bit is your beef 00:41:50.519 --> 00:41:56.000 on dairy research. And in that, I saw the strategy, 00:41:56.079 --> 00:41:59.780 the simulation you guys ran with many different 00:41:59.780 --> 00:42:03.079 options and how you said that technically beef 00:42:03.079 --> 00:42:05.699 on dairy can be a mathematical trap. It can be 00:42:05.699 --> 00:42:09.039 right into an issue. Can you describe what that 00:42:09.039 --> 00:42:11.639 trap is or how that works or what you were finding 00:42:11.639 --> 00:42:16.360 showed? I think what that would refer to is the 00:42:16.360 --> 00:42:22.800 fact that, I mean, there has been this new economic 00:42:22.800 --> 00:42:26.360 alternative for dairy producers, which I think 00:42:26.360 --> 00:42:29.320 is great. Because they can, in addition to produce 00:42:29.320 --> 00:42:33.840 milk and salvage value on the animals that are 00:42:33.840 --> 00:42:38.019 discarded, they can produce also off springs. 00:42:39.199 --> 00:42:41.699 crossbred animals that have a very high value 00:42:41.699 --> 00:42:44.940 in the market. I mean, without going to the reasons 00:42:44.940 --> 00:42:47.219 why is that, I mean, the main reason is because 00:42:47.219 --> 00:42:49.699 we have a high value for those crossbreds, right? 00:42:49.980 --> 00:42:53.179 So there is an opportunity and looking at the 00:42:53.179 --> 00:42:55.659 markets in the future, the opportunity seems 00:42:55.659 --> 00:42:59.460 to be here to stay at least for the foreseeable 00:42:59.460 --> 00:43:03.019 future. Okay, so that's good. But I do think 00:43:03.019 --> 00:43:09.000 at some point there has been It has been overdone. 00:43:09.300 --> 00:43:12.519 There has been, okay, we look at the opportunity 00:43:12.519 --> 00:43:17.460 and we were having a better reproduction performance 00:43:17.460 --> 00:43:23.599 and we used too much deep semen and we entered 00:43:23.599 --> 00:43:26.320 into the problem that I think now we are coming 00:43:26.320 --> 00:43:29.900 out, which was having not enough replacements, 00:43:30.039 --> 00:43:33.440 right? Now, overall, as an industry, we produce 00:43:33.440 --> 00:43:37.889 too many crossbreds. and we didn't produce enough 00:43:37.889 --> 00:43:41.590 replacements to maintain our herds on the farms, 00:43:41.730 --> 00:43:45.010 right? And then that has become to the issue 00:43:45.010 --> 00:43:47.309 that now if you want to buy a replacement, it's 00:43:47.309 --> 00:43:50.550 extremely expensive, and you don't find it in 00:43:50.550 --> 00:43:55.889 the market indeed, right? Now, our tool, I think 00:43:55.889 --> 00:44:00.429 in that sense, is very informative, and it is 00:44:00.429 --> 00:44:02.750 very good that it can give you an opportunity, 00:44:02.869 --> 00:44:06.010 not only... to see what's the proposition that's 00:44:06.010 --> 00:44:10.110 going to give you most net return, but also the 00:44:10.110 --> 00:44:13.869 proposition that will still keep producing enough 00:44:13.869 --> 00:44:17.429 replacements on your herd, which I think is important. 00:44:17.789 --> 00:44:21.269 And I think normally that's the rationale of 00:44:21.269 --> 00:44:23.750 the producer. You want to produce within your 00:44:23.750 --> 00:44:25.989 farm enough replacements. I mean, there may be 00:44:25.989 --> 00:44:28.670 farmers that are willing to buy replacements 00:44:28.670 --> 00:44:32.440 from outside, but that's... An uncertainty and 00:44:32.440 --> 00:44:36.059 that's a risk only some are willing to take. 00:44:36.239 --> 00:44:40.579 Right. So and the other thing I would mention 00:44:40.579 --> 00:44:43.760 in that and that specific tool and that line 00:44:43.760 --> 00:44:48.119 of work is the fact that normally the use of 00:44:48.119 --> 00:44:51.619 beef semen, angus specifically, for example, 00:44:51.820 --> 00:44:56.820 will go hand in hand with the use of sex semen 00:44:56.820 --> 00:45:00.269 as well. So we can use more sex semen to our 00:45:00.269 --> 00:45:05.269 top animals to gain replacements out of them. 00:45:05.690 --> 00:45:09.429 And so we have enough. And at the same time, 00:45:09.449 --> 00:45:12.190 that will give us a space to use more beef semen 00:45:12.190 --> 00:45:15.489 on the rest of the animals to make value out 00:45:15.489 --> 00:45:19.010 of them. So I think that's very important. So 00:45:19.010 --> 00:45:24.599 we did that analysis. One thing I think just 00:45:24.599 --> 00:45:27.579 since we're talking about to mention is if you 00:45:27.579 --> 00:45:33.119 do that one -time analysis today, your herd demographics 00:45:33.119 --> 00:45:37.260 are going to change nine months from now when 00:45:37.260 --> 00:45:39.460 you're having the calvings from that breeding, 00:45:39.679 --> 00:45:43.420 right? And then the recommendation is you need 00:45:43.420 --> 00:45:46.380 to do this constantly until you reach a certain 00:45:46.380 --> 00:45:49.739 level of stability on the herd to make sure you're 00:45:49.739 --> 00:45:52.280 going to have those replacements. that you need 00:45:52.280 --> 00:45:55.099 we do have additional work for example working 00:45:55.099 --> 00:45:57.619 with these roofers that we call that long term 00:45:57.619 --> 00:46:00.119 that we can do exactly the same analysis but 00:46:00.119 --> 00:46:03.920 it's going to give us what's your best proposition 00:46:03.920 --> 00:46:08.239 today to in the long term to still have a stability 00:46:08.239 --> 00:46:11.179 on enough replacements and the best economic 00:46:11.179 --> 00:46:15.099 proposition you know i think it's interesting 00:46:15.099 --> 00:46:17.280 because when you look at the beef on dairy or 00:46:17.280 --> 00:46:21.099 the replacement heifer issue we have right now, 00:46:21.179 --> 00:46:23.099 you know, we're dealing with decisions that were 00:46:23.099 --> 00:46:26.579 made back in 2023. So it's going to take a little 00:46:26.579 --> 00:46:28.679 while. And I know that I looked at yesterday's 00:46:28.679 --> 00:46:31.440 USDA data, and they show that, you know, they 00:46:31.440 --> 00:46:35.539 expect 3 million heifers this year, but then 00:46:35.539 --> 00:46:41.170 you factor in the mortality rates and all different 00:46:41.170 --> 00:46:43.369 issues that have cut that down. We're probably 00:46:43.369 --> 00:46:45.250 looking at about two and a half million. And 00:46:45.250 --> 00:46:47.989 just to keep current production, we need 2 .8 00:46:47.989 --> 00:46:54.730 million. So our current shortage is going to 00:46:54.730 --> 00:46:57.690 continue for at least another year or two, because 00:46:57.690 --> 00:46:59.769 I think it's only now people are starting to 00:46:59.769 --> 00:47:02.670 realize that decision. But with biology and the 00:47:02.670 --> 00:47:05.130 biological cycles and raising and all those times, 00:47:05.230 --> 00:47:08.030 you know, it's two to three years before the 00:47:08.030 --> 00:47:12.079 market. correct itself. We're still going through 00:47:12.079 --> 00:47:17.059 that. You need to forecast. Andrew, I would need 00:47:17.059 --> 00:47:22.579 to end a hard stop at the hour because I have 00:47:22.579 --> 00:47:26.099 another meeting coming up. I'll just have one 00:47:26.099 --> 00:47:29.380 more question then we'll finish up with. You 00:47:29.380 --> 00:47:33.699 have built a lot of tools online and you offer 00:47:33.699 --> 00:47:36.099 them free or some with a subscription to it. 00:47:36.650 --> 00:47:39.210 What would you tell producers about what tools 00:47:39.210 --> 00:47:41.489 are there? How can they use them? Why they should 00:47:41.489 --> 00:47:45.730 use them? Stuff along that line. Yeah, I invite 00:47:45.730 --> 00:47:49.730 everyone to visit the website and you are going 00:47:49.730 --> 00:47:53.090 to have a very good feeling of these tools. I 00:47:53.090 --> 00:47:56.650 mean, we have tried to do this very easy to use 00:47:56.650 --> 00:48:00.130 and very practical for application on the farms 00:48:00.130 --> 00:48:04.019 and all the tools. Almost all the tools have 00:48:04.019 --> 00:48:07.539 come by questions from producers or decision 00:48:07.539 --> 00:48:14.000 makers. So we have that into account very heavily 00:48:14.000 --> 00:48:18.199 and very seriously in order to make tools that 00:48:18.199 --> 00:48:22.480 we think are very useful to the farm. And as 00:48:22.480 --> 00:48:25.539 you mentioned, they have been for free for a 00:48:25.539 --> 00:48:30.050 long time. We needed to go to a subscription 00:48:30.050 --> 00:48:33.150 mode that's going to start actually in 2026. 00:48:33.630 --> 00:48:36.769 But still, anyone who wants to use and has the 00:48:36.769 --> 00:48:42.170 need, we can actually entertain those considerations. 00:48:42.329 --> 00:48:45.809 So I invite you to go to the website and contact 00:48:45.809 --> 00:48:48.610 me if you want to use the tools. I would be glad 00:48:48.610 --> 00:48:52.110 to entertain that. That was Dr. Victor Cabrera, 00:48:52.250 --> 00:48:56.230 University of Wisconsin -Madison. If you want 00:48:56.230 --> 00:48:58.650 to explore the free decision tools his lab has 00:48:58.650 --> 00:49:01.969 built, replacement optimizers, beef on dairy 00:49:01.969 --> 00:49:05.309 simulators, feed management calculators, head 00:49:05.309 --> 00:49:08.849 to dairymgt .info. We'll drop the link in the 00:49:08.849 --> 00:49:10.989 show notes. The full feature article that goes 00:49:10.989 --> 00:49:13.429 with this episode is live now at thebull .com. 00:49:13.690 --> 00:49:16.789 Search Dairy Brain and you'll find it. If this 00:49:16.789 --> 00:49:18.849 episode made you think about how your own data 00:49:18.849 --> 00:49:22.130 systems connect or don't, we'd love to hear about 00:49:22.130 --> 00:49:25.300 it. Drop us a line at editor. at thebullvine 00:49:25.300 --> 00:49:27.539 .com. This has been the Bullvine Podcast.