The Business of Dairy

Are Automatic Milking Systems Profitable in Australia?

November 01, 2021 NSW DPI Season 1 Episode 6
The Business of Dairy
Are Automatic Milking Systems Profitable in Australia?
Show Notes Transcript

Commercially operating automatic milking systems (AMS) have been in Australia since 2001. While this system currently makes up less than 1% of the Australian dairy industry, positivity in the industry is currently seeing many farmers consider AMS as part of future investment in their businesses, with a number of farms installing them in the coming 12- 18mths. 

This month we speaking with two of the “Milking Edge” project team (Juan Gargiulo and Jessica Bell)  who have done excellent research into the profitability of AMS systems compared to conventional milking systems in our Australian pasture based environment and developed world leading, freely available learning modules and decision support tools for interested farmers and industry players to help them in their consideration, understanding and investment process.

Useful resources related to this podcast:

Online training modules and AMS Integrated Management Model (IMM) 

Register for the AMS Farm Showcase.

This podcast is an initiative of the NSW DPI Dairy Business Advisory Unit

It is brought to you in partnership the Hunter Local Land Services

Please share this podcast with your fellow farmers and colleagues and feel free to contact us with suggestions or comments via this email address

Further NSW DPI Dairy channels to follow and subscribe to include;

NSW DPI Dairy Facebook page

DPI Intensive Livestock Twitter feed

NSW DPI Dairy Newsletter

Access the show transcript

Produced by Video Lift

The information discussed in this podcast are for informative and educational purposes only and do not constitute advice. 


The Business of Dairy

Episode #6 Transcript – “Are Automatic Milking Systems Profitable in Australia?”



Sheena Carter: Welcome to the Business of Dairy podcast. I'm your host, Sheena Carter, development officer with the New South Wales Department of Primary Industries dairy team. Each month, I speak with industry people generous enough to share their stories, knowledge and skills with us to help you in the increasingly complex area of farm business management. Commercially operating automatic milking systems have been in Australia since 2001. There are currently 43 AMS farms operating in Australia and current positivity in the industry and an interest in technology is now seeing more farmers consider AMS as part of future investment in their businesses, with a number of farms about to install them in the coming 12 to 18 months. 


This month we look at the history of AMS in Australia, whether the systems are profitable and learn about resources available for industry to investigate the systems in more detail. A unique three year project called Milking Edge, which has been a joint initiative of the New South Wales DPI, Dairy Australia and DeLaval, has undertaken extensive research and development work to help support farmers and the industry in their successful investment and operation of these AMS systems. 


The project is being led by Nicolas Lyons from New South Wales DPI, and this month we speak with two of the Milking Edge team, Juan Gargiulo and Jessica Bell. Juan has been involved through his Ph.D. in research into the profitability of AMS systems compared to conventional milking systems in our Australian pasture-based environment. He has also developed a world first decision support tool to help in the understanding of the system and its performance. Jess has been involved in the development of the very first freely available online learning modules and courses for those who are considering or operating AMS, helping them to understand key farm management areas. 


Thanks for joining us today on the podcast, Jess, it's great to have you as a colleague, but also your expertise in the AMS field is invaluable. So I'd just like to start off the podcast painting the picture of what the automatic milking system­­ looks like in Australia. What is the history in Australia? When was it introduced to Australia? Why people took on automatic milking machines? What were some of the drivers and those kind of things? Can you give us a bit of an insight, paint the picture for us? 


Jessica Bell: Thanks, Sheena, and thanks for having me on today. Probably the first thing to note about AMS is that it's not a new technology. Prototypes were first developed in the 70s, with the first commercial installation occurring in the Netherlands in 1992, and it was only nine years later that it arrived here in Australia in 2001, and it was the first time the technology had been used in a pasture-based system. Historically, AMS is used in smaller indoor systems, often owner-operated, better situated to high producing cows and in areas where there was a stable milk price, but labour shortage. So that's what we often saw in European settings. But as popularity of the technology continued to grow in the northern hemisphere, interest from pasture-based systems in Argentina, in New Zealand and of course here in Australia began to grow. So, four Lely A3 robots were shipped here to Australia in 2001 and installed on a commercial pasture-based farm in Gippsland. And since then, adoption, whilst slower than originally expected, has been consistent and we now have 43 operating AMS farms across the country, in all dairy regions, with an additional five installations that we know about currently going in at the end of 2021 and early 2022, and approximately another 60 farmers seriously considering the technology. 


Sheena Carter: That's a strong degree of interest out there, Australia wide, which is which is good to see. What were some of the reasons in Australia for that early adoption, do you think? 


Jessica Bell: Yeah, so we’ll hear from Juan later on in the podcast, but it was actually part of his study that farmers were asked why they invested in AMS initially, and the top three responses were flexibility, availability of labour and the dairy required upgrading. And we've since seen those three reasons for investment continue to grow. And I think one of the beauties of the dairy industry here in Australia, and the beauty of the technology, is the great diversity that we see, and the technology is able to be a part of. 


So, whilst the AMS systems here in Australia are still predominantly pasture-based, currently about 85% of farms, AMS farms, are still pasture-based. But we've seen continued growth in the housed systems and within that there's even more diversity. The smallest operating farm has two robots milking approximately 100 cows right up to the world's largest pasture-based system, which is operating 16 robots and milking 900 cows. There is businesses operating one AMS dairy. There is businesses operating multiple AMSs, dairies and businesses operating AMSs dairies alongside conventional systems. So there is great diversity within the niche market of AMS. 


Sheena Carter: Interesting. And I guess some of that adoption might be, you know, we often think about the labour, obviously labour within Australian dairies is a very costly component, it’s the second largest cost category that we see generally, or mostly, in dairy businesses, with about up to 24% of total costs. But there's also, I think from, you know, my looking around the country and seeing some farms that are adopting AMS, some of that reason seems to be that succession, the next generation coming into the business as well. 


Jessica Bell: Absolutely Sheena, succession planning has certainly been a major consideration for a lot of farmers considering the technology. We've also seen greater interest in farmers that are considering how AMS fit with indoor or TMR systems. And over the duration of the project, we've completed some surveys of industry providers and farmers, and one survey conducted back in 2015 by my colleague Juan showed that approximately 60% of Australian farmers and 80% of service providers considered that AMS was the fourth technology, with increased expected adoption over the next 10 years. And we actually ran a recent survey as part of Milking Edge that looked at how Australian dairy farmers are investing capital back into their business. And half of the respondents indicated that they would be looking at replacing or seriously upgrading their milking shed in the next five years, with half of these respondents expressing that they would consider installing AMS when the time came. Over the past 12 months, the Milking Edge team has picked up significant interest in AMS, particularly across New South Wales, Victoria and South Australia. 


Sheena Carter: It's great to see and I guess it's exciting times for the industry both, you know, to see that confidence in the industry generally, to start investing back into businesses, but particularly in the area of this type of technology. How old is the oldest unit in Australia? Do we see the units lasting 15 years? What's our longevity? 


Jessica Bell: One of the beauties about these systems is that they can be unbolted and re-bolted into a different, on a different farm, or as part of an expansion. So, we've actually had a really successful second hand market for robots in Australia. The first commercial farm that started in Australia in 2001, they actually retired from dairying altogether and sold the farm, but 2009 is our longest operating AMS currently. But what we have found, Sheena, is around 35% of those current AMS farmers have either already increased the number of robots on their farm or upgraded their system to the newer model, or the later model of robot, and approximately 50% of current AMS farmers have, or expect to continue, to expand their operation in the future. 


Sheena Carter: So there's good, there's confidence in the system and in the technology, obviously, if we're seeing that, which is fantastic. And you've mentioned, we've you know, the investment going ahead looks positive with lots of interest. Five coming up to be installed in the next 12 to 18 months or so and then stronger interest ahead, which is great to see. And I guess this all leads us to consider, are these systems profitable, and we'd hope they are if people are, you know, putting their money in investing in them? So I might just switch to you, Juan. You've done quite a bit of research in this area and investigating the profitability of AMS systems in the pasture-based system in Australia. Could you tell us a bit about the study, the work that you've done to try and understand the performance of these systems, both physical and financial performance? 


Juan Gargiulo: Yeah thanks, Sheena, for having me here on the podcast. So, as you mentioned as part of the Milking Edge project, I was in charge of a research, where we basically did conduct the research around the economics of AMS. So, there were few studies around the economics that used farm data – they were mainly based for confined AMS systems. There were no studies for pasture-based system available. The only available used models and assumptions. So this is really the first study with commercial farm data in the whole world. So what we did was, as part of the Milking Edge, we collaborated with Victorian consultant, Dan Armstrong, and collected economic and physical data from 19 AMS across Australia during five years. So it's a lot of data. It's only 19 farms, but it's more than 30% of the actual population of AMS in Australia. 


Sheena Carter: And that was using DairyBase, or Dairy Farm Monitor method? 


Juan Gargiulo: Yeah so, what we did was we use that methodology from DairyBase and Dairy Farm Monitor. So what we did was, we compared that database with the Dairy Farm Monitor, or we compared with the Dairy Farm Monitor results on that. So what we did actually, was we have AMS across different regions, and of course, there are a lot of differences between regions and also between herd sizes for example. So what we did was, we only compared AMS across similar regions with conventional farms and within similar herd sizes, to enable a fair comparison between systems. So actually, this study used only the three first years of AMS data, so we compare actually 14 AMS with 100 conventional farms from the Dairy Farm Monitor project. In general, the results of what we found was that the physical performance, for example, indicators like milk production per cow, or pasture utilisation or labour efficiency for example, were similar on average between both systems – between AMS and CMS. 


The main thing, of course, was that the profitability measure, for example, as return on total assets or EBIT, were also similar on average between both systems. But what we found was differences in the cost structure. Variable costs were similar in others, but there were differences, for example, in power cost or electricity costs – we found that overall were higher for AMS. 


Sheena Carter: ­­So I guess, just to context, in our variable costs, for our listeners: our variable costs, we’re really looking at those herd costs, the shed costs and the feed costs. So here we’re talking about the shed costs, some differences within those, with the high power costs.


Juan Gargiulo: Yeah. So within shed costs, the electricity or power costs were slightly higher, but this is a small proportion of the total cost, of course. What we found also was that herd costs were lower for AMS. This was because, as the system collects a lot of data on a daily basis, or on a cow basis, the farmers spend less on herd testing, so usually they tend to have a lower herd cost. The main cost category, that is feed costs, actually was similar between AMS and CMS. 


Sheena Carter: Right. Just to go back on the herd costs and the lower costs, you know, people aren't doing the herd testing because they don't need to, they're getting all of that, a lot of that data and more through the system. Do you think, this is a bit of a tangent, but do you think the AMS systems are attracting a certain type of people? To me, I would suggest that they’re data lovers. They love looking at the numbers and understanding or digging into the numbers. Do you think that would be a fair comment? 


Juan Gargiulo: Yeah, I think so. So, most of the farmers that milk with AMS are attracted to data and like to make decisions based on the data. But broadly, it’s not the only thing that explains why they adopt AMS or why they like AMS. There are other issues like lifestyle, labour, and flexibility. Yeah, in general, I think they love data.


Sheena Carter: Data and technology. Yep, sorry. I’ll let you carry on with your costs. So our variable costs…


Juan Gargiulo: We were talking about variable costs, which we found that overall were similar, with some differences in some specific costs. And what we found was in the overhead costs, in general, they tended to be higher for the AMS, mainly due to the depreciation costs. 


Sheena Carter: Yeah, so with our overhead costs, again, we've obviously got – these are more of the fixed costs within a business – but things like repairs and maintenance, and labour. So we've got cash costs, insurance as well. But we've also, when we're doing a true economic analysis, we're looking at some non-cash costs and this is what you're referring to with depreciation – so, that sort of loss in value of equipment over time. The other non-cash costs that we'd be looking at in a business analysis would be the cost of imputed labour, so generally that non-paid family labour that we see in the business. So the depreciation, I guess, is higher because we're investing in this kind of technology. 


Juan Gargiulo: Yeah, that's true. Usually, the initial capital investment is higher for AMS. So depending on the lifespan, you usually have higher depreciation costs. But as you mentioned, it's not cash that the farmer pays – it's a non-cash cost. Also, what we found was that on average labour costs were similar, but some farmers were really compensating those higher depreciation cost with more labour efficiency and lower labour costs. But on average, we found that labour costs were similar. That was quite surprising for us. 


Sheena Carter: So I guess if they've got the system operating well, I guess is what you're saying, and they are getting those labour efficiencies, that higher depreciation cost in those businesses can be can be offset. 


Juan Gargiulo: Yep. And actually, we found that the farmers that were really achieving high labour efficiency were offsetting those higher depreciation costs. In some cases, higher repairs and maintenance costs. 


Sheena Carter: I was just thinking also, in terms of, you know, while we're talking about offsetting costs, if we go back to our energy use and our shed costs, I'm certainly aware of a couple of farms that have implemented solar systems as part of their system, and I would assume that would help offset those increased power costs. 


Juan Gargiulo: Yeah. So, AMS is well suited for solar because the usage, yeah, the usage across the whole day is more stable and not fixed as on a conventional farm. So, it's well-suited for, better suited for solar systems, and many of the AMS are investing in solar now at the moment. So yeah, it's something that could be used to offset those power costs. But as we mentioned, it's true that they are higher but it's a smaller proportion of the total costs in comparison to others like feed costs or labour, and as you mentioned, that is the second largest category, and the depreciation as well. 


Sheena Carter: Yeah, that's right. I guess if we look at our feed and labour costs as total costs in the business, it's going to be 70-75% of total cost. So some of these other costs are a smaller proportion. And on that, you know, repairs and maintenance, perhaps not a huge cost in the scheme of things, but from the point of view of managing the system, we do see slightly higher repair and maintenance costs within these robotic systems, just because of their continual need to be kept operating efficiently, I guess. 


Juan Gargiulo: Yeah, I mean, certainly in most cases, most of the farmers, due to the warranty of the robots, need to commit to a servicing to keep that warranty in the first years. So probably that's not the case for many conventional farms, so that could explain those differences partially. Also, because the technology, of course, has a lot of moving parts and lenses and arms that usually are more expensive, of course, to replace. 


Sheena Carter: Fantastic. So, I guess we're seeing through this study, essentially, you have found that these AMS systems have just as much capacity to be as profitable as conventional milking systems. There's some slight differences within AMS in their cost structures, but overall, the opportunities are just as good as conventional milking systems. And obviously we see, you know, large degrees of variation in profitability in conventional milking systems as well. So regardless of system, there's always a range of profitability. But in general, you would conclude that AMS can be just as profitable as conventional. 


Juan Gargiulo: Yeah, yeah, that's true. And also another thing to add is that we dug a little bit deeper on why some farmers were achieving greater performance and profitability in AMS, and what we found, it's actually that the ones that were more efficient were utilising the robot more efficiently. One way of measuring the robot efficiency is by measuring that milk harvested per robot per day, it's one of the main KPIs, and we found that that KPI is highly linked with profitability, with a return on total assets and EBIT also. And the other important thing is the labour efficiency, of course it makes sense because it's highly linked with the depreciation efficiency in which the capital is utilised and also the efficiency in which labour is utilised. So those two things are key. There is a large opportunity, we found a lot of variability between AMS and the ones that were achieving higher labour efficiency and higher robot utilisation were the ones that were achieving really high profitability. 


Sheena Carter: When we're talking about high robot utilisation, what does that look like specifically? You know, what sort of numbers are the top, top farms doing? 


Juan Gargiulo: For example, the average was around 1100-1200 kilos of milk per robot per day. But some farmers were achieving 1500 kilos and that's a really large difference, on average in the whole year. So that’s milk sold to the processor on average of the whole year. And that difference of 200 or 300 kilos makes the difference in profitability. 


Sheena Carter: So obviously, there's a couple of factors there. It's either, you've got cows that are doing good per cow production and/or you're getting lots of cows through the robots. 


Juan Gargiulo: Yeah, actually we explored that and we found that the main thing was not the production per cow there, it was more, how many cows farmers were milking per robot. So, we found that the largest opportunity there was to try to increase those cows per robot to drive higher milk harvested and then higher profitability. So those were some of the things we found in that study. 


Sheena Carter: And I guess that's probably a topic for another podcast, how you drive that, you know, drive the cows through the system to get high throughput in terms of cow numbers. But yeah, it is very interesting, that efficiency space, and I think we should probably lead into some of the tools that have been developed through the Milking Edge project to help farmers, in sort of understanding AMS systems, but also if they're considering investing in these systems. The project has developed a number of tools and resources to help farmers understand this. Jess, we might just flick to you if we can for a minute on that, because you've been involved quite heavily in some learning modules, online learning modules, that have been developed for the industry. Can you explain what these modules are and how we can access them? 


Jessica Bell: Thanks, Sheena. So you're right, Milking Edge as a training and education project has developed numerous tools and resources available to farmers and service providers. One of the major resources is the one you've mentioned, our online learning modules, and these modules are based on eight key farm management areas and cover topics such as reproduction, herd health, economics and then some really specific AMS related areas such as incentives. So, how we get cows to traffic around a system without human intervention, for the most part, and daily farm routines – what life on an AMS looks like. The beauty about these modules is that they are free and available 24-7 to anybody within the industry. They're designed to be a self-paced learning tool so that users can work through the topics that are most relevant to them and to their situation. 


If you were a farmer looking to invest in this technology or learn more about this technology, you would probably want to work through one to eight of those modules. If you were a milk processor or a veterinarian that had clients or farmers that were considering AMS or operating AMS, you might be more interested in the topics of ‘milk quality’ and ‘herd health’ or ‘reproduction’. So, the beauty is you get to pick and choose what topics you do, and within those modules, how deep you go into that information. We have further links and additional case studies and additional reading that you can all choose to do on top of the base module. They take about an hour to get through, for most people, and are available on your laptop or on your smartphone or on your iPad. So, easily accessible. 


Sheena Carter: And where do people, I mean, we can put a link in our podcast show notes, but where do people go to access, to find these modules? 


Jessica Bell: So, the easiest way is either to jump onto our Facebook page and follow the link from there or on our DPI dairy website page, and there is a link – it's a two minute registration which will then put you on to a Tocal College learning platform that we call Canvas, and it sits on there. You have access to it indefinitely once you enrol and you can review the modules as many times as you like. We actually have had over 200 user’s signup and register to view the modules, and commercial AMS farmers are using these modules as part of their new staff induction. So, any staff that come onto a farm that may not be familiar with the technology can work through these modules and get that base understanding. 


Sheena Carter: Excellent. Really, really good. Yeah. And also you've, this year, you've developed and run a couple of online courses with farmers and service providers? 


Jessica Bell: That's right. So, we took the base, or the foundation level learning from the modules and developed a more tailored specific training course on AMS. So, as far as we know, this is the world's first dedicated training course for AMS and we have had a very successful year in running three courses. The course runs for 10 weeks, you work through those eight online learning modules – one a week, and then we also have farmer led Q&A discussions. So, the ability to talk to current AMS farmers and gain their insight and experience into key management areas on-farm. Users work through practical exercises as part of the course as well. One of which is Juan’s tool, which I know we will discuss later. By the end of the third course, which will finish at the end of November, over 65 people will have completed the course. It is an industry recognized course and there is a certificate of completion at the end of it. So, a valuable thing to have in the résumé when working with AMS technology and farmers.


Sheena Carter: And what’s… who have been some of the audience that have been taking this up? It's obviously people that perhaps are just curious, farmers, service providers, who's in the mix?


Jessica Bell: Everyone that you've mentioned, Sheena. Right from dipping their toes in the water and learning the basics of AMS, up to farmers that are currently installing the technology and looking to get support with that commissioning phase and a real deep understanding of what to prepare themselves for once they start training cows to the robots. Course three, which is currently underway now, over half of the participants are farmers, either seriously considering the technology or they've made that investment, and they want now to ensure that they're as prepared as they can be. 


Sheena Carter: Excellent. So, and we've started to mention your tool, Juan, that's a decision support tool to help people, I guess, drive efficiencies and understand, what would you say, simulate production to see how they can be more profitable using this? Can you outline, or explain in much better detail than I just tried to do, your decision support tool? 


Juan Gargiulo: Yeah, sure. So as you mentioned, there were no tools available for pasture-based system to kind of help farmers that were operating AMS, or farmers that were considering investing in AMS. There were no tools that allow farmers to tell them to simulate the scenario or optimize their performance or better understand the system. So, what we did here was, we use all the data collected in the study that I mentioned previously, like 19 farms, five years across Australia. And also, we integrated that data with another set of data that we collected from another project that is called the International AMS KPI Project, where we monitor 28 robotic farms from Australia and other countries from overseas – New Zealand, Ireland, Argentina and Chile. And we put all that data together and we developed some models and we develop a user interface, a web-based user interface. So, we created this tool. So, basically, this tool, as I mentioned, allows farmers to simulate the scenarios, optimise the system, benchmark against real data, and the beauty of this tool, of course, is that it's based on real farm data so farmers can simulate things and see how they are performing, or that simulation in comparison to real data. So, it's really useful thing. 


Sheena Carter: Yeah, that's fantastic. And that has been recently released online. Can you tell us what sort of engagement you've had with it so far? 


Juan Gargiulo: Yeah. So, before releasing it online, we also tested the tool with 11 farmers, also from different countries. They mentioned, or their feedback was, that the tool was really easy to use and useful for them. And after that, we launched the tool, actually in late August, and in the first two months we got more than 300 users that used the tool from more than 25 different countries. So, a really good engagement. 


Sheena Carter: Yeah, that's fantastic. Very widespread. And obviously it's meeting a need if there's no such thing out there in the market currently, and as you say, with real data. 


Juan Gargiulo: Yeah, and we not only see that this is useful for farmers, but also it has been used, for example, in training courses that Jess conducted, or we use it, for example, in the university, with students for pracs. So, it's got more applications other than just for farmers.


Sheena Carter: So, yeah, that's brilliant. And again, we'll put a link in the show notes for people that are interested so that they can click on it and play around with it themselves and get an understanding around how it works and what it can tell you about operating one of these systems. Anything further to add from either of you in terms of tools and resources or events the project is running? Obviously, the project wraps up, the Milking Edge project, wraps up at the end of this 2021 calendar year, but the resources will still able to be accessed. Anything else in the pipeline between now and the end of the year? 


Jessica Bell: I just wanted to call out, or highlight Sheena, our AMS farmer showcase, which is happening on the 30th of November, and it is a day-long webinar. But the idea is that you don't sit there and listen for the entire day, but you join the farmers, or the topics covered that are most relevant to you or of most interest to you. So, we will have farmers from all over the world share their story, how they're managing their system, what challenges and opportunities they've faced, which will be really, really interesting for both farmers considering the technology, farmers operating the technology, and service providers wanting to get a better understanding of the systems. So, that's on the 30th of November and details will be on our Facebook page. 


Sheena Carter: Fantastic. Well, look, thank you both very much for your time and your knowledge today. I think you've done a fantastic job in your journey with the Milking Edge project, and it really is ground-breaking, you know, research and development of tools and support for farmers and service providers that are either considering investing or are actually operating these systems and helping the industry to understand how they can get the best out of this. So, I think it's wonderful work and congratulations to you both and thank you, Jess, for being part of today and thanks, Juan. Much appreciated. 


Jessica Bell: Thank you, Sheena.


Juan Gargiulo: Thank you, Sheena.


Sheena Carter: Thank you for listening to this month's The Business of Dairy podcast, produced by the New South Wales Department of Primary Industries Dairy Business Advisory Unit. This series is also brought to you with funding and support from the Hunter Local Land Services. This month's show notes contain a transcript of the podcast and a link to the Automatic Milking Systems page on the New South Wales DPI dairy website. Here you will find links to the online learning modules and the AMS decision support tool called the Integrated Management Model. Also, the AMS Farm Showcase mentioned by Jess, will be held online on Tuesday, 30th November 2021, featuring Australian and international farmers speakers using the technology. It is free and the link to register is also available in the show notes. 


We'd love you to share this podcast with your networks and feel free to send any feedback or suggestions for future episodes to You can also subscribe to our New South Wales DPI Dairy Facebook and DPI Livestock Twitter feed and view, or subscribe to our quarterly DPI Dairy newsletter using the links provided.