YT YouTube2Text

YouTube2Text → Transcripts → Dr. Dennis Buckmaster & Dr. Jacquelyn Boerman: Digital Tools for Dairy Feed

Dr. Dennis Buckmaster & Dr. Jacquelyn Boerman: Digital Tools for Dairy Feed — Transcript

by The Feed Science Podcast Show • by Wisenetix · 5,753 words · 818 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00uh I do like to on this topic of
0:02uniformity to parse the problem into two
0:05parts and one part is uniformity among
0:09batches and so that is you know to this
0:12morning versus tonight or today's
0:15morning versus tomorrow's morning versus
0:17the next morning they may not be uniform
0:20even on average if the person loading
0:23the mixer doesn't get exactly the right
0:25amount according to the
0:27[Music]
0:35recipe. Welcome to the feed science
0:38podcast. I'm Adam Fenholles coming to
0:40you from North Carolina State University
0:42in the prestig department of poultry
0:44science and with animal food
0:46manufacturing program here on behalf of
0:48Wiseetics. Thanks for joining us. We've
0:50got two guests joining the podcast
0:52today. Both are from Purdue University.
0:55We have Dr. Jackie Borman. She's an
0:58associate professor in the department of
0:59animal science and Dr. Dennis Buckmaster
1:03who is a professor in the department of
1:05agricultural and bioengineering at
1:07Purdue. Jackie, Dennis, thanks for
1:10joining me. Thank you for the
1:11invitation. Great to be here. Yeah,
1:13absolutely. If you guys wouldn't mind,
1:15and Jackie will start with you um
1:17looking at my screen here, you're on the
1:19you're on the top window, so you get to
1:20be first. If you wouldn't mind,
1:22introduce yourself um a little more.
1:24what your background is and and what
1:26you're doing there at Purdue and then
1:27we'll uh follow up with you Dennis as
1:29well. So my background is I'm originally
1:31from New York and I grew up on a dairy
1:33farm and then I went to graduate school
1:36at University of Illinois doing dairy
1:39nutrition and then a PhD at Michigan
1:41State also doing dairy nutrition. After
1:45uh graduating with a PhD, I worked for
1:47Cargill Animal Nutrition for a little
1:49bit uh doing technical consulting and
1:52then I have been at Purdue for about
1:55seven and a half years working mostly in
1:59dairy nutrition and management research
2:01extension and teaching. Gotcha. All
2:03right. Uh well, I grew up in northeast
2:05Indiana on a farm. We also had a short
2:08line equipment business and
2:10interestingly that was all about
2:12livestock feeding systems, silos,
2:14conveyors, TMR mixers and things like
2:16that. So I'm an egg engineer from
2:18Purdue, went to grad school at Michigan
2:20State and the first half of my career
2:22was at Penn State University doing
2:25primarily forage systems work including
2:28harvest storage and feed delivery, TMR,
2:31particle size, things like that. Uh
2:34about halfway through my career I moved
2:35to Purdue. I've been here just almost 20
2:38years now. Um, and here focus a little
2:42bit the same and then sort of made a
2:43hard turn toward things in digital
2:45agriculture and the work that Jackie and
2:48I are doing together with some others.
2:49Sort of a kind of a unique blend of
2:52forages and feeding and digital a. So
2:54really excited to be involved.
2:56Excellent. Yeah, that's uh we we've got
2:58sounds like we've got a great um kind of
3:00breadth of expertise here from the from
3:02the uh animal feeding side and and
3:05backgrounds all the way going back to
3:06farming and equipment and digital and
3:08dairy and it's how we we should have a
3:10great conversation. Um I think what
3:12we're going to spend most of our time
3:13talking about over the next 20 minutes
3:15or so um is what we do at the TMR. And
3:20so we've got a number of different folks
3:21that listen to the the podcast and um
3:25you know we know that we have folks that
3:26are listening that are in the typical
3:29compounded feed mill type of industry or
3:32academia. We've got folks that listen
3:34that are involved in liquid feeding and
3:36premixing. And we have folks that are uh
3:38doing things related to the uh to the
3:41the feeding happening there at feed lots
3:44and at dairies using the TMR. So, if um
3:47you guys wouldn't mind uh and uh Jackie,
3:49I guess we'll let you do this. If you
3:51wouldn't mind describing to our audience
3:53just briefly what a the TMR is for those
3:56that um might not live in that that
3:59world. I'll start and then I know Dennis
4:01has a lot of experience with this as
4:03well, but uh TMR stands for total mixed
4:06ration. So, it's going to be a
4:08combination of forages and byproducts,
4:11minerals, grains, etc. And uh dairy cows
4:15are going to be fed a TMR every day. So
4:18um most of the time with our high
4:20producing lactating cows on an asfed
4:23basis, they're going to be fed well over
4:26a 100 pounds and some may be approaching
4:29150 pounds of feed that they're being
4:31fed. So, because it is this combination
4:34of so many different ingredients, that's
4:36really why we want the cow to be able to
4:39consume the same uh ingredients and
4:42nutrients for each bite of feed she's
4:45consuming. So, that's kind of why we're
4:48leaning towards evaluating uniformity of
4:50that TMR. Absolutely. And if and Dennis,
4:53if you wouldn't mind, describe kind of
4:54what the actual um the the the physical
4:58nature of of what a TMR is doing versus,
5:01you know, kind of our our inside a feed
5:03mill feed mixer uh type of a situation,
5:06right? So in the TMR because we have
5:08forges and concentrates, uh there's a
5:10large variation in the sort of the
5:12structure and the feel of those
5:14components, the the length of the
5:16particles. You know, in some TMR mixers,
5:18you toss a entire large round bail in
5:21there, and the mixer is supposed to chop
5:22it up into appropriate sizes for the cow
5:25to consume. Uh, and at the same time,
5:28like mix it perfectly, uh, or at least
5:31near, you know, at least uniform enough.
5:34Uh, so it you have some feeds in that
5:38ingredient blend that are very dense and
5:40small particle sizes, and you have other
5:42particles in there that are relatively
5:44light and fluffy. And so there's a term
5:48called overmixing. I don't know that I
5:50really like that one, but uh like you
5:52can do too much. You can reduce the
5:54particle size too much. And so you know
5:57the from the nutrition aspect uh we need
6:01the right physical and chemical traits
6:04in each of those bites that that Jackie
6:06is talking about. Well, at least within
6:08the meal, we want the right blend
6:11chemically and physically there. Yeah,
6:13you got Yeah. Yeah. I'm I'm I'm already
6:15enjoying this. You're talking my
6:16language. When when I hear someone say
6:18they don't really like the term
6:19overmixing, I I get excited because I
6:21hear that one, too. But of course, as
6:23you pointed out, in the TMR world, the
6:25mixer is also kind of a processing
6:28machine as well. So, in theory, we could
6:30overprocess. We're probably not unmixing
6:32by overmixing though, which is certainly
6:35certainly important. Um, and then also
6:37that idea because we talk about this
6:39with the different species and we've
6:41talked to folks here on the podcast that
6:42do poultry and swine and multiecies and
6:44dairy cattle and we talk about
6:46uniformity and and the fact that
6:50nutritionally speaking. Uniformity is
6:52important. We want good uniform feed.
6:55But we can be a little different if
6:57we're talking about a day old chick
6:58eating grams of feed and a lactating
7:02dairy cow eating hundreds of pounds or a
7:06100 pounds worth of total thing. Okay,
7:08we we don't have to have exactly the
7:10same amount of uniformity there. But
7:12knowing that we want as much as we can
7:14reasonably get, why does the the
7:17uniformity matter to the cow? And and I
7:20know Jackie, you already talked about a
7:22little bit, but a little bit more from a
7:24nutritional perspective, what's the what
7:26are the problems if she's not getting a
7:28uniform feed? Yeah. So, for this dairy
7:32cow, which we're expecting her to do a
7:34lot for us, so she's producing high
7:36amounts of milk and um from as a
7:40nutritionist, again, it's not a chick
7:42that's eating just grams, but we are
7:44balancing some nutrients at the gram
7:46level for her. So those ingredients are
7:49going to be really expensive. So they're
7:51things like rin protected amino acids
7:54that we want to make sure that every cow
7:56is getting a similar amount or at least
7:59um based off of her intake a similar
8:01amount. Um so and that's going to have
8:04an effect on things like her milk
8:06protein output or milk fat or just milk
8:11in general. So we want to make sure that
8:13there is uniformity. We're also a lot of
8:16times feeding the rin. Um, so we want
8:20consistency in that rin environment. We
8:22don't want a lot of variation to occur.
8:25Uh, so that the outputs from the rin
8:28like volatile fatty acids, like
8:29microbial protein are consistent
8:31throughout the day. This cow, we're
8:34asking her in some cases to grow, but
8:37we're definitely asking her to produce
8:38milk every day. Sometimes to also grow a
8:42calf. Uh so there's a lot of things that
8:44we're asking her to do every day and she
8:47needs to have consistent nutrition in
8:49order to make that optimal. Absolutely.
8:52So we we've talked a little bit about um
8:54the different ingredients and the forges
8:56and whatnot. And you talked about
8:58something there that you you we are
9:00concerned about like you said down on
9:01the Graham level some of the
9:03additives. Those additives are generally
9:06going to come as part of like a larger
9:08supplement, right? where we're going to
9:10be we're going to have a supplement that
9:11contains a lot of our expensive
9:13additives that we're buying from a feed
9:16mill or we're manufacturing at a
9:17different location and then we're
9:19combining that grain supplement with
9:23those additives and whatever else with
9:25what else in the TMR. A lot of times
9:28that's going to be like a carrier like
9:30some corn grain or soybean meal that is
9:33going to add the the mass that you need
9:36really in order to mix appropriately.
9:38But something that is a little unique on
9:40the ruminant side is that we're doing
9:44the mixing at the farm. Uh so we so
9:48every day, you know, a person is out
9:50there mixing multiple diets for cows.
9:53So, we don't have maybe the expertise of
9:57someone who's working at a feed mill and
9:59has the training and has the equipment
10:01um to make sure and and is doing quality
10:04control on how uniform they're doing
10:07because we're doing it on farm and it's
10:10every day but there's not a lot of tools
10:12in place for them to evaluate the
10:14performance that that they're getting
10:16out of the diets they're mixing.
10:18Absolutely. Okay. So with that in mind,
10:21knowing that we want to have as much
10:25uniformity as we can reasonably get with
10:27the equipment we're using and the amount
10:28of time we have and the ingredients that
10:30we have at our disposal and knowing that
10:32it's being a lot of the the mixing
10:35that's happening is being done on farm,
10:37bringing in supplement, adding fiber
10:39sources, adding forage sources, maybe
10:41some energy sources, etc. How do we at
10:45least currently commonly evaluate the
10:48TMR and and to know whether or not we're
10:51doing a good job mixing? Okay, I'll take
10:53that one. Uh there would be several ways
10:56to do it. You know, there are sort of
10:58research standard methods of putting a
11:01particular chemical or a salt in there
11:03and taking samples and then uh
11:07calculating uniformity by standard
11:09deviation or coefficient of variation.
11:11often uh I do like to on this topic of
11:15uniformity to parse the problem into two
11:18parts and one part is uniformity among
11:22batches and so that is you know to this
11:25morning versus tonight or today's
11:28morning versus tomorrow's morning versus
11:30the next morning they may not be uniform
11:33even on average if the person loading
11:36the mixer doesn't get exactly the right
11:38amount according to the recipe
11:41But it's not always their fault either
11:43because primarily those forages might
11:46vary.
11:47Particularly silage out of a bunker silo
11:50depending on the climate. If it rains or
11:53it's been hot and dry, the moisture
11:55content of the forages can change from
11:58day to day. And so because these rations
12:02are balanced on a dry matter basis, but
12:04it's mixed on an asis is or what basis
12:08uh depending on you know the talent or
12:11the aptitude of the person doing the
12:13mixing they may make some adjustment
12:15that's approximately right. It can
12:18they'll never do it perfectly right
12:19because they can't get data fast enough.
12:21Uh but that you know that is one of the
12:24sources of variation. Uh so we would at
12:26least be able to evaluate the among
12:29batch like is the person who's doing
12:32that
12:33job being conscientious and if it calls
12:36for 2,000 pounds of this did they put in
12:392,000 or 1,800 or you know 20 250 or
12:43something like that. So that's one. The
12:45other one would be within batch
12:48variation. So even now I've put the
12:51things in this blender, this mixer, and
12:54if I don't run it long enough, well then
12:57I'm going to have well, you know, hot
12:58spots in energy or protein or that sort
13:00of thing. If I ran it too long, I like
13:03as you mentioned earlier, I might
13:04process that forage and cut it up a
13:07little bit too much. Well, there is
13:08going to be some time. That's sort of
13:10the sweet spot. It's been long enough,
13:12go ahead and deliver it. Don't waste any
13:14more fuel and machinery maintenance by
13:16running the mixer 20 minutes when 10
13:19minutes would have sufficed also waste
13:21labor like who wants to do that. Um so
13:25the uniformity within the batch I don't
13:28know that many like there are some
13:31companies and uh consultants who would
13:33facilitate that analysis but I don't
13:35know that universally farms actually do
13:38it. Uh but certainly one way that we'd
13:41advocate doing it is the particle size
13:43analysis. It's a very physical
13:46uh thing. Take a sample, shake it out,
13:48weigh some masses, and you see it even
13:51if you you don't regularly analyze
13:54uniformity. It's it provides a teachable
13:56moment regarding uniformity. Again, my
13:59guests today are both uh faculty members
14:02at Purdue University, Dr. Jackie Porman
14:05and Dr. Dennis Buckmaster. All right.
14:07So, I think we've done all of that in
14:10the first half of our discussion to get
14:11to now the second half of the
14:13discussion. Um, would you guys like to
14:15describe the project that you're working
14:17on using digital agriculture, uh, video
14:21analytics, things like that to figure
14:23out if there's a way that we can kind of
14:25in real time measure and report out on
14:29this TMR uniformity? Yeah, there's
14:31really three objectives that I think we
14:33can talk about here for the project. And
14:36the first objective is to evaluate that
14:40uniformity. So we have a project where
14:43we have cameras mounted above a mixer
14:45wagon and then before the mixing process
14:48starts we are taking video of what it
14:52looks like in the mixer wagon. So it's
14:54directly above looking down at the
14:57contents of the mixer wagon. And then
15:00after it's been mixed for half of the
15:02duration that it's supposed to be mixed,
15:05we stop the augars and we look down
15:07again and we evaluate the uniformity.
15:10And then at the final before it's being
15:12fed, we evaluate the uniformity again.
15:15How we're evaluating uniformity is we're
15:18selecting a hundred. It could be a
15:20thousand. It could be however many. But
15:22we're selecting regions of interest
15:24within the mixer wagon and we're saying
15:27are these regions similar to one another
15:29or are they different? So we're doing
15:31every comparison between all of the
15:34regions. It results in a lot of data. It
15:37results in like a lot of data to then
15:40analyze and figure out what to do with.
15:43And um we're working on this project in
15:45conjunction with Dr. Amy Reedman who has
15:48the video analytics expertise. And then
15:51um we have a student who's gaining a lot
15:53of that skill set as well. But so that's
15:56one of the projects that we're working
15:57on to try to evaluate uniformity again
16:00within that mixer wagon. And one of the
16:03key things that I think we're trying to
16:05do that is difficult to achieve is
16:08giving information to someone before it
16:11is unloaded because there's not a lot of
16:14information that the farmer can have
16:17before feed is unloaded. So it's from
16:19that perspective of above looking down.
16:22Um and we're comparing it to some of the
16:26measurements that Dr. Buckmaster already
16:28talked about about particle size
16:30distribution and um the individual
16:33nutrients across a feed bunk. So that's
16:36kind of our comparison. So another
16:38project that we're working on is also
16:40with video and um trying to use a
16:43different video camera to predict how
16:46much feed cows are consuming. Uh so
16:49that's somewhat related to this as well.
16:52And then the third one is really more of
16:54Dr. Buckmaster's expertise about
16:57integrating all of this data because we
16:59know that dairy farms are creating a
17:02tremendous amount of data and our
17:04project is creating even more of it for
17:07them. So we need robust tools in order
17:10to be able to pull in this information
17:12so they could have access to it in a
17:15meaningful way to make decisions.
17:17Excellent. So what then um Dennis are
17:21the the challenges with that that
17:23integration and pulling all that data
17:25together? Um you know we we talk about
17:27similar things about in the feed mill
17:29side of okay we've got a a potential
17:32very significant big data source if we
17:35wanted to measure everything and in a
17:37lot of some cases we can in some cases
17:39we can't but the question still remains
17:41well what are we going to do with it
17:42anyway and I I imagine that's the
17:44challenge here as well. It very much is
17:48uh so I'll start with the end in mind.
17:50The goal here is if we could know on an
17:53individual animal basis how you uh first
17:57that we delivered to her to her bunk a
18:01uniform blend and how uniform it was.
18:04That was part of it. Uh if we could know
18:06on an individual animal basis how much
18:08she is eating when she is eating it then
18:11we can tease out a behavior. we could,
18:14you know, sort of uh check off some
18:18anomalies like this cow usually eats
18:20this much by this time of the day and
18:22she didn't. Like I'm not really worried
18:24yet, but then it gets to tonight and she
18:26still hasn't. Okay, maybe there's a time
18:29for an alert to fly, you know, via text
18:31or email or something like that. Uh so
18:35that's one of the elements. The other
18:36one is really feed efficiency, feed
18:39conversion efficiency. And you know, as
18:42Jackie mentioned, of course, they're
18:43producing milk, but it's not just feed
18:45to milk because depending on their stage
18:47of lactation, they may be growing a calf
18:50or recovering some body condition. So
18:53the in the whole of it we would want to
18:57know exactly how much did she eat
18:59exactly what did she eat? What is her
19:02rate of gain? You know not necessarily
19:05by the day but this week or this two
19:08week period how is her body weight
19:10changing? What is her milk production by
19:13the day? Because in that
19:16instance, excuse me, if there is much of
19:19a a change daytoday in milk production,
19:21that should kind of raise flags like
19:24something is up
19:25here. Well, that data comes from
19:28assorted sources. And even in the in
19:30what we now have as a Purdue animal
19:33sciences data ecosystem,
19:36uh there are multiple pieces of software
19:39that collect different aspects of the
19:42dairy data. Could be a tracker on her
19:44anklet. It could be a uh you know a
19:47twoing monitor. Can be these video
19:49cameras that we're speaking of.
19:51Certainly the milk production by the
19:52quarter by the milking event you know
19:56morning, evening or even three times a
19:58day. the integration of that
20:03data particularly when you want to get
20:05it to the individual animal level and to
20:08be able to do some calculations about
20:10number 17 and number 18
20:14uh this is complex I like to use like
20:18it's a silly quote but there are many
20:19variables in the equation of life well
20:22there are a lot of v we're talking about
20:23the life of a cow and there are so many
20:26variables uh some of them you know we
20:29wouldn't necessarily need by the minute
20:31but actually knowing the environment uh
20:35the temperature and humidity in the barn
20:38that can give me insight as to why the
20:40whole group ate a little more or ate a
20:42little less today and so I on one hand
20:46we would want to know individual animals
20:48but they are a part of a group and like
20:50what's the group
20:52doing all of this can in the end give us
20:55insight of oh they consume more or less
20:58less than what we expected. Maybe
21:01because we changed forage source, like
21:03we switched from one silo to the other
21:06on that day. Uh well, that's helpful to
21:10know. So now I'm not worried about a
21:12disease or sickness or something like
21:14that cuz you know it it is what it is
21:17and maybe next time we'll adjust in a
21:19more smoother transition instead of an
21:22abrupt change in feeding like all those
21:24kinds of things. So the the challenge is
21:27the data integration because the data is
21:30coming from different platforms or
21:33sources.
21:35Deep in the jungle, nature holds the
21:37secrets to healthier, happier
21:39[Music]
21:40chickens. Science is unlocking those
21:42secrets with Opient, the revolutionary
21:44lighting innovation for poultry farming.
21:46By mimicking nature, opient gradient
21:48lighting creates a unique floor level
21:50lighting gradient that improves feed
21:52conversion by up to four points, slashes
21:54energy use by 70%. And keeps flocks calm
21:58and stress-free. That means better
22:00animal welfare and a more efficient
22:01farm. Want to learn more? Opient
22:04gradient lighting is now available
22:05exclusively through your local Valco
22:07dealership or email
22:10[email protected] today.
22:14What are you guys specifically finding
22:17also
22:19about using using video analytics in
22:23this case as far as I'm thinking both
22:25the environment it's being used in as
22:27well as um what we're what we're trying
22:31to look at and and I've got in my mind
22:32and I'm no expertise on any of this but
22:34I've got in my mind the fact that okay
22:36we're we're going to be outside we're
22:38going to be dealing with dust and rain
22:40and environmental conditions like that
22:41and we're also dealing with, you know,
22:43it's it's not like we're mixing
22:46um, you know, bright red basketballs and
22:49bright blue marbles. We're we're we're
22:51doing relatively small stuff and it's
22:53all going to be some sort of, you know,
22:56uh, natural shade of yellows, oranges,
22:59greens, browns. So, what is what has
23:01that process been like trying to I I
23:03suppose train the the systems on what
23:08they're actually looking at? I think
23:09that when we started out, you know, we
23:12write a grant and we have all of these
23:13ideas of what is possible and then you
23:16get the data and you're like, oh, you
23:18learn a lot as you access the data. So I
23:21what and it's really been a lot of
23:23expertise of Dr. Reedman, but what we've
23:26learned is that there's different color
23:28spaces that we can explore. So some are
23:30going to um be a little bit more uh
23:33sensitive to things like shadows. So
23:36maybe we should avoid that color space
23:39and go towards one that that shadows and
23:41lighting conditions have less of an
23:43effect on. And you're right, it is brown
23:46and green in color a lot of times. So
23:50we're initially looking at um the pixel
23:53intensity. So we are going to uh look at
23:57color from uh on a spectra and then
24:00figure out uh where it's falling. Um but
24:04there's been some other uh instances
24:06where we're exploring like maybe some
24:08different things so different next steps
24:10and we've been working with Diamond V
24:12who has quite a bit of expertise in this
24:15space of doing Tamar audits um as well
24:19but we know that mixer one of the issues
24:22with mixer wagons is that maybe they're
24:24not run at a higher high enough RPMs to
24:27get the augers moving fast enough so
24:29feed just kind of sits and stirs rather
24:31than mixing. Um, so there could be pe
24:34places in the mixer wagon that are going
24:36to be more critical to evaluate than the
24:39overall mixer wagon. But yeah, we're
24:42learning a lot. Um, dust,
24:45manure, all the wetness, like all of
24:48these things aren't great for video
24:50cameras. Um, but I think that there's
24:52some tools that we can utilize to try to
24:54minimize the risk for that. And we
24:56really are relying on a lot of expertise
24:58from people who have done video
25:00analytics for a long time. just a wee
25:03bit. Uh we're trying to keep this
25:05certainly on the affordable range. So
25:07we're using red, green, blue cameras.
25:10However, for those that are familiar
25:12with imagery analysis, there are things
25:15called multi-spec or hyperspectral
25:18imaging, they can actually see layers of
25:21color that the human eye can't. And so
25:24there is a potential that hyperspectral
25:27would tease out, you know, what is
25:29really different about a piece of
25:30soybean meal versus a part of the corn
25:33stock that is to me looks to be the same
25:36color, but maybe hyperspectrally it's
25:38quite different. So uh we haven't
25:41pursued that just because at as of today
25:44it's pretty cost prohibitive. like we
25:46could do it in research but uh we want
25:49this to be practical on the farm and at
25:52least at the moment I don't know that
25:53hyperspectral imagery would be practical
25:56on the farm but it also does hold
25:59potential like let's go back to uh
26:02forage harvesters now uh at least one
26:04brand uh can give you nutrient analysis
26:07on the fly while you're chopping the
26:09corn silage or the alpha alpha how are
26:10they doing it hyperspectral imaging so
26:13it's uh it's back to you know few
26:16decades ago. N IRS, near infrared, near
26:21infrared uh spectroscopy. Uh that's the
26:25sort of thing. Well, maybe it will
26:27someday be practical because not only
26:29can I say, yep, it's uniform. I can also
26:31know and this is the net energy of
26:33lactation. This is a the degradable
26:36crude protein value of what we're going
26:38to put in the bunk in five minutes.
26:40Yeah. No, that's I think that's an
26:42excellent excellent set of points
26:44because you know and and for the
26:46audience that's familiar you mentioned
26:47NI you know we've gone in the feed
26:50industry from okay NI was was obtainable
26:53only by so many and it was only benchtop
26:56and then it became more obtainable on
26:58the benchtop by more and more folks
27:00where we see it in a lot of different
27:01places and then we start seeing inline
27:04NIR that can be used and it's still too
27:07expensive for most facilities to have it
27:10all over their facility. But we're
27:13getting closer and closer to that. And
27:15then we're taking it another step in
27:17some instances where we're not even just
27:19using it to measure nutrients. We're
27:21figuring out, well, if you grind a
27:24sample to different particle sizes, the
27:26NI looks at it differently. It's why you
27:28got to be so careful when you grind it
27:29before you put on the sample. Oh, well,
27:31that means particle size has an impact
27:33on it. Can we use this to measure
27:34particle size? Can we use this to
27:36measure uniformity? And it sometimes
27:39feels in the moment like a very slow
27:41burn, but you know, you look back 10
27:44years later and go, "Oh yeah, 10 years
27:46ago, I'd have said we can never afford
27:47to do this." And now someone figured out
27:50how to build that same camera for, you
27:52know, next to nothing. And it's it's
27:54that part's always really interesting.
27:56So, kind of to I think both of your
27:58points, um, you kind of have to to get
28:02interested in what's going and and
28:03actually that's the the question that I
28:05want to to I think close out on and we
28:08always try to do something at the end
28:10that that gets into professional
28:12development or resources, but I I think
28:14you guys could answer a really
28:16interesting question and it's it's not
28:18just specific to um your particular
28:22project, but overall, and we see this in
28:24academia and it's I imagine it's true
28:26also in the industry. It's easier, it's
28:29always easier, and that's true for in
28:32many cases funding and publishing and
28:34finding students and everything else.
28:36It's always easier to kind of do the
28:38next um iteration of the basic science
28:42thing that we already know, like, okay,
28:44let's let's let's let's feed some more
28:45cows and see how they grow better or
28:47whatnot. But you guys have chosen to do
28:48something that's, you know, kind of
28:50highing fruit. you've chosen to do
28:51something that's that's high risk, high
28:54reward from the standpoint of we have no
28:56idea if this can even work. What what's
28:59that like in the in in the environment
29:01in which you work? Like like what's
29:03exciting about that? And and what would
29:05you say to others who are
29:07considering, you know, lowrisk,
29:11lowreward versus that high risk, high
29:14reward kind of thing of of really
29:15getting into the the future. I guess we
29:18have on this project and we're also very
29:20interested in the adoption of this. So
29:22we have another collaborator Linda
29:25Feifer from a sciences education and
29:27communication. So in this project we
29:30have faculty and students from at least
29:32four different departments animal
29:34science a and biological engineering
29:36electrical computer engineering and this
29:39uh social sciences of adoption in
29:43ASC going forward in terms of
29:45professional development all of us need
29:48to be really good at working in teams
29:51when I know my language and my field but
29:54I don't know yours but I kind of need to
29:57for us to make prog
29:58progress and we've had many
30:01conversations where something was really
30:03clear in one in one mind
30:07uh and kind of misunderstood or maybe
30:11even not understood at all in another.
30:13And so that ability to cross talk among
30:17disciplines is
30:20critical. You have to want to do it. And
30:23we are blessed uh with Jackie leading
30:25this and the team that we have that
30:28everybody does want to do it like you
30:30know there is good energy uh that also
30:33is a I think for the grad students
30:36particularly and the undergrads for them
30:38to witness that collaborative energetic
30:43approach like I think this is just a
30:45really good culture to build but we're
30:49building that interdisciplinary culture
30:51So, you know, the use of tech and the
30:55application of it, that sort of thing.
30:57Great. And I would agree. I think that
30:59this has been just a privilege to be on
31:03this team and to learn. So, I I find
31:05myself thinking more like an engineer
31:07than I ever have. And I'm a
31:11I'm a nutritionist by training. And a
31:13lot of what I had done was incremental
31:16change to like you know just get down to
31:19maybe understanding requirements more
31:21and that is progress but it is not giant
31:24progress all the time. So this has been
31:28a little bit uncomfortable based off of
31:30what I was trained to be uh because we
31:33are trying something that may not work.
31:36And when I feed a cow, I know that like
31:38I know what I fed them and I know what
31:40the responses are and you can always
31:41write something about what happened. But
31:43when we're doing this, it's a lot of
31:45like designing systems and trying to,
31:48you know, it's iterations of things to
31:51try to get it to something that's usable
31:53and it doesn't always work the first
31:54time and then you have to redo. So it
31:57does take a different mindset, but I do
32:00feel like it's been just wonderful to
32:02learn about like how other researchers
32:05work and how we can make progress by
32:08understanding our needs a little bit
32:10more and our capabilities. Yeah,
32:12absolutely. And and and as the host, but
32:15someone who's also interested in these
32:16kind of things, I'm going to echo what
32:18you both just said. It's it's exciting
32:21to go do the hard stuff. It's exciting
32:24to go do the stuff that people haven't
32:25tried to do before. You will not be
32:28successful doing that if you aren't
32:30uncomfortable and if you don't have a
32:33good team. You just you it won't work.
32:35And I think you guys have both just
32:37explained that very very nicely. So I
32:39appreciate that and I I hope it's a
32:40lesson that the audience can take as
32:42well. I very much enjoyed our
32:44conversation. It's uh it's been a very
32:46quick half an hour. I think you're I
32:47think the project you guys are talking
32:49about is uh it's quite exciting. it um
32:51it's one of those things that you know I
32:53imagine could significantly have an
32:55impact on the industry but also will
32:57have applications beyond what you're
32:59just doing in this specific instance
33:02looking at TMR. You figure out how to do
33:03this, it will apply other places and I
33:05think that's that's really really cool.
33:06So I appreciate both of your time today.
33:08Thank you so much. You're very welcome.
33:10My guests today have been uh two faculty
33:14from Purdue University. Dr. Jackie
33:16Borman is an associate professor in the
33:18department of animal science and Dr. Dr.
33:19Dennis Buckmaster is a professor in the
33:22department of agricultural and
33:23biological engineering. Thanks again on
33:26behalf of Wiseetics. I'm Adam Fenolds
33:29from North Carolina State University.
33:31Thanks a lot for listening. Thanks a lot
33:32for watching. Please feel free to like,
33:34subscribe, leave us reviews. Let us know
33:36what else you'd like to hear about. And
33:38until then, we'll talk to you the next

About this transcript

This page contains the full transcript of Dr. Dennis Buckmaster & Dr. Jacquelyn Boerman: Digital Tools for Dairy Feed by The Feed Science Podcast Show • by Wisenetix, generated from the public captions YouTube serves with the video. The transcript has 5,753 words across 818 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.