Hello everybody, and thank you for joining us tonight. I'm Theo from the webinar vet, and tonight we have Doctor Navvarinam, and he is currently a clinical farm animal vet and includes coordination of the West Point Pro Dairy Group. Previous roles included working in university higher education.
The pharmaceutical sector and in the global animal health industry. He has studied and consulted in several countries with the focus on dairy production. Other roles have included managing the health and running of a commercial calf rearing anti-unit and sheep flock.
Doctor Navarrinam has further responsibilities in being a current director for the Oxford Farming Conference, trustee for the Country Trust, and co-founder of the British Veterinary Ethnicity and Diversity Society. He also sits on the farm animal working group for vet sustain. Welcome, Doctor Navarine.
Thank you, thank you very much for, for, for that, and, it's a pleasure to be here. So I want to today talk about health monitoring technology on UK dairy farms. So as you said earlier, I'm a, I'm a vet, first opinion vet, and I work with vets all over the country, and my role has previously been mainly virtually 100% dairy, and most of my working life's been dairy.
I've got more of a diversified role now with beef and sheep as well, but my passion has been dairy farms and working with dairy cows. So Well, the plan for my session, so this, the aim of this webinar is really for the first opinion mixed vet, or the farm vet practitioner, and, and basically, and, and I'm not going to go into a huge amount of detail, but really the amount of detail that as a first opinion vet, it would be good to know, and especially with farmers, want to ask the sort of basic questions that you have an idea. So what I want to go through is the role of a vet, especially a farm vet.
Monitor and why we monitor them. I'll go through the, the vast array of different monitors and sensors out there, and really give you a sort of quick guide on, on what they're about, and what to look out for in them. And then I'll have a quick discussion about the future of this technology and what can go forward.
So the first thing is, what's the role of the vets? And there, there are many roles of the vets, and especially as a farm vet, we, we, we, we, we, we're empowered with lots of different ideas. The first thing is to be a trusted advisor.
So again, farmers rely on us to be there to support them in many different ways, and to advise them both on the health of their animals, but sometimes some of the business decisions as well. We need to be up to date. So again, farmers rely on us to know what's going on, to know what's available, and to be aware of, of not only vetting, but of the sector as well, because again, as a farm vet, we're not just vets, but we're part of the agricultural sector.
And so having that role and being in there means that we need to be able to provide that information. We need to be able to collaborate with the farm team. So, again, farmers have all sorts of people working on their farms, nutritionists, foot trimmers, AI technicians, and other other workers on the farm.
So. Again, if we can create vet led teams and work with all of those people around the table, then that will help us to, you know, work more efficiently, help the farmers support their business, and then also help us in our other role, which is to find the answer. And, you know, if farmers ask us a question, even if we don't know.
What, what we're really entrusted with is to find the answer somehow or some way. So, you know, having a way of finding the, the resources to get those answers is really important. And, and then what happens with is if we can provide all those services to the farmer, we can provide a better trust and a better relationship with that farmer.
And generally that's how this talk came about, and once I was in my practise and one of the young vets came to me and said, well, one of our farmers was asking about room and bonuses and which one was the best. And they said to me that they didn't know really what the differences were between room and bonuses, and in fact, what bonuses were out there. So I said right, what I'll do is I'll, I'll go and do the research and I'll produce a cheat sheet.
So when I was researching about room and bonuses, I actually thought, you know, and I created this cheat sheet, I thought I gave it to all the different members of my team. And, then I realised, actually, this would be really good if everyone in the sector knew, because actually the amount of knowledge in our sector is quite low when it comes to these things, because we always expect the farmer to find out from other people. But actually, you know, if we want to take the lead and, and be that trusted advisor, then actually it's up to us to find out that information as well.
So, the next question is, we start with what do we monitor and why do we need to monitor. So as vets, we're trained all about disease and treating and preventing disease. So if we take that as one example, what we need to know is, can we detect disease on the farm and monitor it over time?
The, there's other thing that's quite a big part of our sector is antimicrobial usage, and how do we reduce antimicrobials? So again, if we can detect disease early, then maybe there's a way that we can reduce the amount of antimicrobial use, which again, benefits the whole sector and society in general. So, monitoring disease and antimicrobial use is one reason why we need to monitor.
Production, which is basically the, the, the main aim of the farmer on how, you know, they need to produce food, they need to produce it efficiently. Again, if we can monitor, and we can detect problems early, we can reduce the amount of loss or pro time that food production's reduced, and also food production is is linked to disease. So again, you know, if we're monitoring production, then we can also monitor disease.
And that's linked to the business of the pro of the farm and the sustainability of the farm long term. Finally, it's vital if we want to monitor herd health and reduce, and and increase the amount of preventable ways of, of reducing disease. So forward planning is key to sustainability.
So if we can develop key performance indicators or standard operating procedures, then we can decide, you know, what we're doing now, how we can improve things and try and make positive changes. And that's all part of herd health planning, and a part of the bigger picture. So, monitoring has, is important for all these three things, and all three are related, but all three are all about.
Vetting, and all parts of vetting, and it's all about us working with the farmer to, to create a sustainable business and and and futureproof their business. And also improve what we're doing now, going forward. So We start with what do we want to monitor, because now we know why we want to monitor, what do we want to monitor.
At the end of the day, fundamentally, it's the dairy cow. She is doing everything for us, she's producing the milk efficiently, we're putting inputs into her, we're getting this output, we want to find out, you know. What, what, what we, we wanna know that she's performing efficiently and at her peak.
A cow has many processes going on, and as vets and farmers, we're tuned into trying to look for cues that show normality and cues that show abnormality. And, and so, you know, what sort of things are cows doing that can indicate normality and abnormality? And so there are, you know, when, when we're at vet school and when we're on the field, we look out for different things like are they grazing, are they ruminating, are they coughing?
How are they standing, how are they walking? So, as vets, we're very good at using our senses, and in fact we've got to rely on our senses being number one. You know, we can have as many different instruments and other contraptions to, to, to monitor cows, but really using our eyes to see, so again, breathing rates, mobility.
Using our ears to hear, can we hear any abnormal abnormal sounds, any creaking, anything like that from their breathing, their movement. Can we smell ketosis, you know, ketones in the breath, or anything abnormal, you know, from the cow, you know, maybe it's dung smells abnormal, maybe there are wounds that smell abnormal. And using, you know, our hands, our senses touch.
So again, the other, is it hot? Is it painful when we touch? When we touch the backs, can we monitor the body condition score and tell if it's high, low, or normal?
So again, you, on the cow, there's so many different things that we could look at. And again, as vets, we do look at, and a simple clinical exam goes through all those different bits and pieces, and, and, and we do that every day. But we've got to understand that these are limited, and they're limited in many, many ways.
But the first thing is they're limited because they're individual, you know, we can only do it one cow at a time, at one point. And, and, and the other thing is it's time point. So when we get called out to a farm, we can do these clinical exams and then we walk away.
So that clinical exam is based on that time point. We don't know what it's like in the afternoon unless the farmer tells us. We don't know what it's like the next day, unless the farmer tells us that we go out, but we could only ever monitor something at one time point by doing it physically.
And it's very subjective. So again, two different vets could go, and if it's something like touching, feeling, listening. Then two vets could come up with slightly different answers, so it's quite subjective on the vets, and so one vet might think that something is abnormal, whereas another vet might think something is normal.
So, again, you know, we are, we are trying to be as uniform as possible, but there is that subjectivity as well. So that's the, so that's the issue sometimes with just using our sensors to on a single cow. But what the cow really is, it's a walking amount of data.
Everything's a data point. All those different things I talked about were data points. So, again, you know, we can give a value to everything.
We can give a value to the body condition score, we can give a value to the breathing rate, we can give a value to the way they walk. And so once we have a value, we then have a starting point, and if those values. Change based on that median point, we can sort of say is something abnormal or normal, and by how much is it abnormal or normal.
And then we can become more consistent in how we discuss abnormalities and normalities. The other thing with these data points is then we can start to compare a greater number of animals with a consistency, and over a longer time period as well. And so then before we can detect short-term changes versus long-term changes.
And that's really really important going forward if we want then look for herd health in a consistent way. So knowing that a cow has these changes that we've always looked at in the past and we know can be, can, can indicate problems, and we know that there's data to be collected, the next thing is monitoring that data, how do we sense that data? And again, in, in farm animal practise, there are a number of different.
Monitors out there. And I'm gonna go talk through them in detail, in the next few slides. Now, before I go on, the biggest thing that I think young vets especially get overwhelmed with is the technology and the amount of monitors out there.
And actually, what are, how are they all fundamentally different and how are they all unique? And the simple answer is, there isn't a lot out there. If you go, for example, to a dairy show this year.
Go to the same dairy show next year. Go to the same dairy show in 5 years' time, and what you'll find is 95%, 98% will be exactly the same, OK? And actually that's the same with the technology.
Most of the technology, nearly all the technology is exactly the same, for example, a Rumenbolus, you know, they're all pretty much very similar, all very similar technology, but there might be slight differences, or they're using the same technology in slightly different ways or validated it in slightly different ways. So. Don't get too worked up or too overcomplicated on the number of monitors and the number of different things that people market them for.
Break it down and keep it simple. And I hope that when I discuss the next few things and the monitors, that's what I do, because again, there isn't a lot of technology out there. They're all very similar and actually it all depends actually on the farm and what you want from it, rather than what it'll, what it'll sell you or tell you.
So, when you're thinking about a monitor, and what monitors, the, the basics are is they look, find a variable and they look for changes over time. And that's basically what a sensor does. If the sensors in the neck or the rumen, or in the vulva, you know, it doesn't matter where it is, they're looking for a variable and then looking for changes over time.
And that's fundamental to any monitor. And then if that variable changes from the norm in any way. There's an abnormality, and then we need to work out what abnormality is.
So the monitors detect a problem, and that's what we said earlier, find a variable, look for changes, and then it needs to be analysed. And the analysis is, is, is what is actually the most powerful thing at the moment. The detection is less powerful, it's, you know, the technology hasn't really changed very very much, over the last, you know, 10 years or so, but the analysis has.
What we've started with is getting data and just. You know, working out statistics. Now, things like machine learning, artificial intelligence, all these things are looking at data and modelling the data and create, and, and the more data that's going in, it's creating more and more accurate artificial intelligence and machine learning and modelling.
And, and this is what is is key and we'll just talk briefly about this later. But It makes, it helps, but the most important thing, once you've got this data is then to use the data. The data is useless if it just sits there and isn't analysed and doesn't inform the farmer and the vet of how to move forward, how to make positive change.
And so if you use that data to make a positive change or to clarify that a change has made a positive difference. Then the, the whole monitoring and the whole data collection is useless. The important thing is, what are we going to use the data once we've collected the data.
And so as a vet, we know when we're going through the process, think about what we want to detect and why, but then how is it going to be analysed and how is it going to be used? So there are 3 types of technology out there, and according to Knight's classification, there are at cow, near cow, and from cow sensor systems. So the first group includes sensors placed on the cow and in the cow.
So on the neck, legs, body, vagina, tail, rumen, rectum, or the subcut tissue. The second group of monitors are interaction between the cow and its environment, so external camera systems, drone technology, weight measurement systems. And then the third group, which is probably developing a lot more now, is monitoring the physical and chemical properties of the products from the cow.
So think about meat, milk and things like that. So when you think about technology, they're grouped into one of these 3. And, you know, what we want to see is, are there any diseases or issues on the cow, based on the cow.
Are there physiological differences? So again, you know, is there any changes to the body, to the, to the, to the way the cow, the production of the cow, the way the cows, acting, working, you know, are there detections that we can make there? And the third thing is, are there changes or anything in the environment that's causing changes to the cow?
So again, can we monitor the environment in different ways to, to see if there are issues that need addressing, or to affirm that any management changes we've made or anything we're doing has been a positive impact on those cows. And so those are the three big things that we want to see. So, let's start with the different types of .
Technology out there. And, and what we wanna do, so let's start with the pier, and in the parlour. Cows are releasing milk, so again, looking at milk, and if we're thinking about disease, we're thinking about mastitis, clots in milk is a very simple one to do.
And yes, we can do all sorts of tests, handheld tests and such like. But another simple one's an inline mastitis detector, which is basically just a grid. And when milk flows over the top of that, any big clots are caught up.
So after a cow's been milked, a farmer can quickly see that, contraption on the left. And if there's lots of clots there, then maybe discard that milk, or that milk is very, is very mastictic, very clotty. The reason why it is, you can use other sensor, you can use other machines to detect.
But, you know, that's a very quick, easy, you know, that blue contraption can be pulled out, washed for the next cow, and so on. So nice and simple for that. For the other side, we've got an automatic version of the same thing.
So the milk flows across, and that contraption will measure the clottis of the milk and then tell you whether it's mastic, it's a problem or not. So again, you can just do it by visual observation, or you can use a digital detector. So that's a very simple Impala technology used to sense the mastitis in milk.
Now, if we upgrade that. We can have robots, and robots are really taking off at the moment. In some countries actually, they're getting rid of the robots because they feel like there's a huge amount of data, and how do they manage that, and also the way that you manage a robot is very different from a parlour, but.
Robots do provide a lot of technology, so I want to give an example of a robot, here, it's a Leli astronaut. Now there are other good manufacturers out there, producing some brilliant robots as well, it was just that, I'd recently, Read about and seen a Lily Lily, so I wanted to talk about it. And this lily is, is one of the top of the range ones, one of the brand new ones.
And some of the stuff they talk about that this, this machine can do is, for example, for udder control, it has a somatic cell counting machine, and it can check, check the temperature for every quarter, for every milking. So, you know, the cow's milk has been checked for somatic cell counts and the temperatures checked. For repro, again, this machine can then, Communicate with tags, so again for reproduction, it can detect as well, the floors can have a weighing scale as well, so when the cows walk on to get milked, there can be a weighing scale.
Also, the machine can communicate with collars and all sorts. The machine also detects the fat, protein and lactose, and that kind of indication of health as well, and nutritional health of the animal as well. So, you know, again, fundamentally, milk is being extracted and this machine can analyse the milk and can give us an indication about mastitis, nutritional health.
And other indicators, as well as, you know, doing the whole milking process. This information can then be beamed to be analysed and be sent straight to the mobile phone of a farmer. The farmers don't even have to be standing next to the cow to find out if the cow's aesthetic or not.
So that's like future technology that here now. Monitors, there are so many different monitors out there that do so many different things. Cattle behaviour is influenced by the productivity, reproductive and diseased state of the animal.
So therefore, changes in the behaviour of the cattle can be used to assist in predicting and detecting health problems in dairy cattle. So sensors allow the behaviour of the complete herd to be monitored over 24 hours a day. So over long periods of time, non-invasively and without affecting the normal behaviour, so the cows can just carry on as they are.
Their use is being increased by both researchers and farmers. And with increasing herd sizes, it's difficult for producers to allocate sufficient time to observe cattle. So these technologies help assist in identifying individual cows that need attention, plus monitoring the behaviour, health and welfare of the herd at the same time.
So here I've got 4 different examples of monitors that are on cows. So I'll start with the feet monitors, which are the bottom left. Leg sensors can record lying, standing, movement, and stepping behaviour.
And their use can be useful because it's found that severely lame cows spent longer time lying down and had a longer duration of lying bouts compared to cows that are not lame. Also, prior to calving, cows that suffer dystociia reduced their transitions from standing to lying positions more frequently than cows with eusocia. Also, masitic cows have reduced lying times, a higher number of daily lying bouts, and took more steps than control cows.
Cows with hock and knee injuries lay down for less time each day than cows without lesions, and cows with yonies lie down for even less time. So several leg-based sensors which record line behaviours have been validated by comparing their data with live observations, and that's how they validate these things. So looking at, looking at these leg sensors, they can tell you about those different diseases purely on the amount of time cows are rising and standing or or lying.
Net monitors, assess cattle welfare and fertility, and they do that by monitoring the time spent ruminating. The time spent ruminating is a key indicator of health, as cats that are ill or injured eat less and therefore ruminate less. So You know, if you estimate the time spent ruminating, that might be a huge way of estimating the herd health welfare.
The onset of stress, for example, also is accompanied by changes in behaviour, such as a drop in time that animals spends ruminating. So these neck collars, along with the, the, feet collars are accelerator meter-based sensor systems. So the ones on the legs detect movement, the ones on the neck collars detect rumination.
Then what happens is with the neck collars, that an algorithm is used to combine the parameters of rumination with the activity levels to generate a health index, and a healthy cow gets an index of 100. But if the index falls below a certain threshold, and sometimes it could be up to about 86, then that will indicate to the farmer through a, an app or something that there's an issue such as metabolic to digestive, mastitis or metritis disorders. The next thing I'll talk about is ear tags, so you can see on the right bottom, that's an example of an ear tag.
And again, ear tags use acceleration centres to pick up vibrations from cow's ears. So the data from the movements is then collected by receivers, which are mounted in the cow shed and then analysed in real time to see whether cows are eating, resting on heat, or ruminating. So ear movement is normally consistent but irregular when the head is down when eating.
And there's a different movement pattern when the cow is ruminating. So again, the differences of when the movement when the cow's eating with the head down to ruminating, has been validated and there are, and, and they can be detected. Some ear tags can also detect body temperature, so they can flash, for example, red, if an animal has an increased body temperature, which could indicate a fever, which farmers could then, when they're walking past the line of cattle, they see these red flashing, indicators, they can go and look at those animals in more detail.
So again, all accelerator metre. Technology, just put in different places. Actually, if you wear a Fitbit or any of those sort of watches, it's the same technology.
Say accelerator metre technology in those sensors to detect movement and changes. And it's just the difference is, is how you validate those movements to being abnormal compared to normal. On the top, there's on the right, there's a temperature vaginal contraption that you stick in the vagina monitors body temperature, and on the top left is something called mucor.
And it detects the changes in the tail. Movement, indicating when we're close to carving or not. Rumen bonuses, are, are, are growing and the use of rumen bonuses are growing, and it's probably one of the only things that we can do in cow at the moment, because with chips, the difficulty is when cows are sent for slaughter, how do we make sure that we've removed those chips?
And, and that's the difficulty of getting chip technology to be, to be validated and used, in food producing animals. But boluss, they sit in the rumen, and normally. Like there are lots of bots and bonuses in the market and when you Google, you'll find lots of bonuses, but fundamentally they're all very, very similar.
They normally automatically take readings of pH and temperature every minute, for example, and then they store the average value of the pH and temperatures over a 10 to 15 minute period. So they'll regularly be taking the temperatures and pHs, but then they'll give an average every 10 to 15 minutes. And they store this data in a base station or cloud.
Some, bonuses also detect activity levels as well, which would be maybe a little bit of an extra. So, you know, the, the, the most they'll detect is pH temperature activity, the minimum they'll detect is pH and temperature. And really, you know, it doesn't matter which bolus you use, those are basically the three different things that they can detect.
The pH detection is a nutritional sensor. So when you feed an animal, it directly influences the rumen pH. So those rumen pH those rumen pH sensors can detect feeding composition and feeding management and help support that.
The temperature part of the rumen bolus detects, reflects drinking and feed intake as well. And it also coincides with increased, and so when an animal drinks or takes in food, its body, sorry, when an animal drinks, takes some food, its temperature naturally decreases slightly. So when the rumen bonuses detect that, that could indicate one or the other.
But also the rumen bonuses can detect an increase in body temperature, which could increase, which could indicate. Disease The other thing that it could be detected is also e stress because body temperature slightly rises in east stress. So what the companies do is they try and validate to see how much of an increase in body temperature is and over what period indicates e stress and what indicates fever, and the same way, how much of a decrease and over what time period indicates drinking, and indicates feed intake.
And then the companies can then market it for those different uses. The Rumen bonuses can also detect activity, and the changes in activity can be, can detect whether there's a rumen issue. So a decrease in activity of the rumen could indicate that there's an inflammatory condition, or there's a lack of food being intake.
So there may be a diet problem or or a management problem. So that's what the, the, the, the, so they can do those three things, activity, temperature and pH and those are the different things they can tell. And different companies will have different claims based on those three things.
Things to be aware of is bonuses don't always sit in the room and sometimes they can slip to the reticulum and so can be, can give false readings. And the future is, can these bonuses detect biomarkers in the rumen, or can we put something in the rumen to detect biomarkers? Another big issue, especially with bonuses, but with technology in general, is the amount of validation.
Only 14% of bonuses on the market have had some sort of, some form of external validation. So again, you know, and, and bonuses especially are the least validated form of technology onal available at the moment. So be very aware of farmers buying bonuses with lots of promises, you know, we need, you need to know where those promises have come from and what studies have been done.
And actually the research is so poor that, you know, it's still gonna take a while before we get, we get better, better data and better use of these bonuses. Nose bands are, are unique. They're not very common and not very in this country especially, but they are available.
And for example, this one, it comprises of, you know, this, a nose ban sensor, a data logger with an online data analysis, which goes to the cloud. The nose band sensor is consists of a glycol-filled silicone pressure tube and it's got a built-in pressure sensor placed in a casing of adjustable polyethylene halter, which sits over the bridge of the nose of the cow. So when the cow chews, the curvature of the nose band is altered by the jaw movement and it exerts a pressure change on the pressure tube.
So when the pressure sensor. Changes, it allows the jaw movements to be recorded. And it can quantify and classify those jaw movements, depending on the pressure peaks.
So, you know, the pressure peaks can change and can be characterised by whether the animal is ruminating, eating, drinking, or doing some other activity. And therefore can detect, you know, whether something is higher, lower, or as it should be on those different things. So those are on sensors, and again, like I said, there aren't a lot of them.
The technology is very similar among the different manufacturers, but it's just sticking to the basics and understanding that this is the limitations of those sensors, this is what they can do, but it's about getting the data and using it. Video is growing and automated video analytics is, is, is a big part of the sector now. The behavioural data has traditionally been collected using direct visual observations or video recordings.
But visual behaviour observations are time consuming, labour intensive, subjective and open to interpretation. So we need to think about ways that we can detect automatically. Grammes are a good thing and they can be used in lots of different ways.
So for example, we can look at the cow. And, and, and monitor the cow, monitor the behaviour of the cow, monitor the physiology of the cow, and, and use it in that way. And I'll talk a little bit more about that in a minute.
We can look at feeding, for example. Again, we can put cameras over feed bunks in sheds, and we can monitor how the cows are eating, where they're eating more, where they're not eating, so we could look at the, the amount they're eating, whether they're sorting, and whether there's areas where cows are getting access to food. And again, you know, with food being one of the biggest inputs into a farm cost wise, and being essential for production, we need to be able to monitor and, and, and do that.
The other thing we can do with data is. Look at movement. So again, we can look at how much individually are cows moving, are cows lying down for the, you know, 14 hours that we'd like them to lie down, or are they standing, not doing anything, or how much time are they eating?
We could also look at them in a setting as well, in the shed. So again, are cows grouped in one place or are cows evenly distributed? And if they're in one place, is that because the place they're not at got poor ventilation?
Is it too hot? Is there bullying going on that's preventing cows moving properly? Is the design of the shed not adequate?
So again, using cameras to monitor cows, to watch cows is really important and see if they're doing the things they should be doing, resting and eating, they make two main movements, especially if they're indoors. So, using, using cameras, we can take that to the next level, rather than just filming a shed, we can analyse that video. So again, there's all sorts of technology coming out there and the use of computers to analyse those images, to analyse data points or pick pick points, and, and monitor the changes in those points over time.
And the important thing about this is it needs good identification of those animals. So we can look at a herd and look at general, but if we want to pick out individual animals that need more attention, that need help, the cameras have got to be able to identify animals and alert people based on that that single animal, as well as looking at the herd and making sure that as a group we're trying to get what we're trying to achieve. The cows with poor body condition are at significantly greater risk of developing lameness, and that lameness in itself is closely linked with weight loss and other productivity related issues, so lower fertility and lower milk yield.
So using cameras that look above the animal, that can take photos of animals walking past them, and again, Having what, what is normal, what is what we're meant to be, and therefore what does this animal look like? Again, for these cameras to work really well and for the data to be valid, we want good animal identification. And also we want to be, the, the data has to be compared over time.
And finally, we need to know the stage of production of these animals, cos again, the body condition of score of animals can change over their lactation and during the dry period. And so therefore make, you know, if we, if we, we need to know are animals actually at the right stage of body condition or are are there abnormalities, and this is important as a, as if we're looking physically, we could know that ourselves. But if a camera's looking at it and trying to give us the answers or.
The reaction, then the cameras have also got to be learning about the differences in animals and also the difference in breeds, again, some breeds have slightly different confirmations to other breeds that can look in different body condition scores. Another example looking at locomotion, so again, we can look on side on, and we can have pick points on the legs, and then the cameras can analyse, are those, are the feet moving in the right way, you know, are the legs moving in the right positions, are the legs in the right angles, and are there differences? We can look from above, is the body moving right?
Are the animals walking, you know, the intras or placing their feet correctly. But we can also use infrared cameras to detect lesions such as early stage interdigital dermatitis. And so the, the, you know, it's not just colour cameras, but we can use infrared and we can use other things as well.
The interesting thing about cameras, especially looking at feet and is. That they need to work in harsh conditions and that's the limitation of these cameras, you know, in warm, wet, acidic, slurry, and also they might get banged, they might get knocks, so the cameras have got to be sturdy. But, so, but fundamentally, the cameras don't have to be anything special.
So if they, if they're put in a special casing in a special way, then actually you can buy even a GoPro would be fine for these sort of things. A, a basic digital camera will do the job as long as it's protected and looked after in the right way. And then as long as the, the computer analytics is able to, to use that data to create the meaningful, .
Outcomes, then actually it's quite a simple technology to. Instigate and use on farms, and again, this technology is increasing, but what's increasing is not the actual technology, but what we use with the images that we we're getting from that that technology and how we're analysing it and creating meaningful answers. The next thing, really, you know, there, there are limitations to looking at one technology, monitoring one thing, because again, it is, it can mean that we're gonna lose some of the sensitivity and specificity.
We might not get a full answer, so remember I said body condition scoring's great, but we need to know the production stage and other things about the animal, the feed and all these sort of things. So what we could do is maybe combine some of the technologies together. So here's a really good example.
It was like in a, in a study where they were looking. At carving, time to calving in beef and dairy cows, but rather than just using one piece of technology, they combined three bits of technology to create the, increase the accuracy of the data they were getting. So they used collars, tail mounted sensors, and then they used locating devices, which were sort of camera systems on the shed.
And again, by combining all the different data, when they were getting answers, it was a much more accurate because it was based on 3 different systems rather than just one. That increases the accuracy and validity of the data. So, technology's great, and I've gone through the main bits of technology out there, and to be fair, they're probably one or two little things that I haven't mentioned, but generally speaking, this is pretty much most of the technology available to farmers, to sense changes from normal, to detect health and welfare, factors for their cows, their herd, their environment.
But there are limitations, so cost is gonna be a big thing. You can't just monitor everything, you can't have every bit of technology, because again, that's a lot of data. What does it mean?
How does it change everything? Does it make a difference if you know a bit more to then what you're going to do? Everything has to be maintained, so there's an ongoing cost with technology, not just purchasing of the equipment, but purchasing of the analytics and of the data.
So, again, you know, when picking technology, what do we need to know, why do we need to know it, and how are we gonna get that from as cheap as possible and most efficiently. Sensitivity and specificity of the different technologies is important. Like I said earlier, the validity of these technology is quite poor in general, and especially for room and bonuses, which have some of the poorest validity out there.
But again, I will talk a little about sensitivity and specificity. Combining technologies increases that, but again, we need to know the limitations of what we're doing and, and what we want from it. Validation, I've just mentioned it earlier.
There is an issue, but again, over time, and the more farmers that take this on and the more data that's been generated, that will only increase and improve over time. Finally, the practicalities of farm. Not all farms are indoor herds, not all farms are completely outdoor herds, and not all farms have the same breeds.
Some farms are lucky enough to have huge big buildings, some farms are making use of older buildings that have been modified. So think about technology and think about how does it work for your farm. Some people are using rotary parlours, some people are using .
Robot parlours, your robot milkers, you know, don't do the, does the technology fit with what you're doing? Does it fit with your lifestyle, does it fit with your budget, and, and actually, you know, who's gonna monitor. It, who's gonna service it, how is it going to be looked after.
So before engaging with any sort of technology, think about, does it work for your farm, and just because someone else has got that 10 energy doesn't mean that you, it will work 100% for you. It might be something slightly different that would be better for you. When it comes to sensitivity and specificity.
They are the measures of a of a test's ability to correctly classify if an animal is having a disease or not having a disease. So sensitivity refers to a test's ability to designate an individual with a disease as positive. So a highly sensitive test means that there are a few false negative results, and thus fewer cases of disease are missed.
The specificity of a test is ability to designate an individual who does not have the disease as negative. So a highly specific test means there are a few false positive results. But it's not feasible to use a test with low specificity, for example, for screening.
So because many animals with the disease will screen positive and therefore receive unnecessary workups because they haven't got the disease. It's desirable to have a test or to have a sensor that's highly sensitive and specific, but that's not possible. So what you need is a trade-off.
And, and that's, that's the reality. Because there are some animals that will be normal, some will be abnormal, but some will form in part of the grey area. So choices must be established when you're choosing, you know, if you want something with a high sensitivity compared to a high specificity.
Data is the king of everything of of this, you know, the sensors are only detecting the problem, but what do we do with what they've detected, and the data is what then helps create that change, that action, that impact that we want to make. But data has to follow a few different things. It has to be unified.
So, you know, raw data can come from lots of different sources, but it can be incomplete, it can be duplicated, it can be conflicting. So it has to be unified for analytics to be able to use it and to make it more reliable. So unifying data means merging the different sources of data into one database.
So we get a total accurate view of, you know, the sources of where the data's come from. And, you know, if lots of different sources are, are independent and we put it into one, then we've unified it and then we've got one place to look at. Flexibility is where, you know, data can be flexible, so it depends on who, who analysed it, which company analysed this one compared to which company analysed that one, which programmes analysing it.
And so there are variables to that, and different features to the data. So what we need to do with the data is we need to clean it. And so what we do is we clean it, take away some of that variability, take it, and, and therefore make it a bit more easier to evaluate and record, because it's one type of source.
It needs to be accessible So, again, you know, the people who need to analyse it need to get hold of it. So, you know, if the data is on one system, but it's not allowed to go into another system, so therefore we, we end up with having two or three different systems. Data gets lost, data gets less useful, less more difficult to use.
So, therefore, you know, people lose faith in it. But if it can go into one system, if both the farmer, the vet vet, advisors, such like can get access to that data so they can help use it, then it becomes more powerful. If it's in a format that can be transferred to another system, then that helps.
Data then. Has decisions have to be made on data, even if it's a decision to stick, to carry on, because the data's saying that, then do it. We've got to, the data has to be a part of the process of making decisions, otherwise the data has no power, has no purpose.
And the biggest thing going forward is ownership, who owns the data? Is it the data companies, is it the farmers, is it the milk buyers who owns the data, and that's a debate that will happen in the future. So what does the future look like?
For example, the monitors I talked about are limited in what they can detect, but if we can find out those variables, they could mean other things, then those monitors have bigger uses. For example, neck mounted activity coll collars can be used to detect heat stress, and there's a paper that has found that that rumination. Can be detected the heat.
But again, we need to know what the temperature was like that day and connect it with that rumination, and then we can say, yes, it was heat stress, rather than just looking purely at the heat column saying, you know, it's a combination of technologies of the future. Better machine learning, artificial intelligence, data control, you know, these are the things of the future and, and are growing. You know, environmental concerns, looking at sustainability growing, so methane monitoring, and even contraptions like this which can mage the methane, but also turn it into more environmentally friendly gases, or less harmful environmental gases.
So technology is changing and we're growing with that as well. And also the requirements of the sector's changing, so being more sustainable, environmental friendly is, is, is creating a demand for technology and sensors to help monitor that. So, I've done a very quick overview of all the technology, and I hope you found that interesting.
I've kept it very simple, and, my take home messages are there, there is a lot of technology out there, but there's not a lot of variability between the technology. They're all based on very simple, fundamentals. There might be lots of different manufacturers with lots of different claims, but it all comes down to very simple sensing.
How we analyse that data is essential, and making decisions on that data is essential. There's not a one size fits all, so explore them, think about the farm system and what's best for that farm, and, and, and choose a technology that way and. Just keep looking out there, talking to manufacturers, and, and, and, and, and being aware of, of what's coming along, and, and as vets, if we do that, then farmers can rely on us to help them, support them, and futureproof their business going forward.
So thank you very much. Thank you, Doctor Namarrina and thank you all of the people who are going to join this webinar. Good night and see you in another webinar.
Bye-bye.