Technology in Meat Processing Industry

Speaker 1 0:00

Welcome Barbara Webster from NMR .

Speaker 2 0:05

Well , thanks , Mark . Thanks for having me Very exciting .

Speaker 1 0:08

Excellent , barbara , I sort of stumbled across your technology recently I think it was a set-up to try and start to do live assessment and sort of got interested in the tech . And Emma Hazleton was wandering around Beefwick and I think could have been some of your team or maybe even you , I'm not sure , but anyway , we've sort of been stalking you for a couple of weeks so I thought it would be great to get on and have a chat about the technology and anything . We're always interested in , things that let us know more about our products and our industry .

Speaker 2 0:52

Maybe if we just start with you personally and talk about your career today and how you sort of ended up working in in automation and work meat processing and measurement and this kind of field . Yeah , very good . Um , so I started um . Well , basically my career has tracked the um path of technology commercialization . So if you're aware of sort of the key building blocks there it's R&D , money , markets , outcomes , money yeah , More money .

Speaker 2 1:10

And so I started at the lab bench doing R&D in material science and had a three-year stint over in Manchester doing my PhD in that area and then came back to New Zealand , moved into managing R&D , understanding the costs of doing that , managing for outcomes and how to get the best value out of your dollar .

Speaker 2 1:36

I then did a stint in trade and market access and that was really informative in terms of understanding the barriers to getting a product to market . So that was useful and all of that was time spent in government . So after that I then stepped out and went into the business world , and so that's again where I was managing R&D , but this time I was a member of the executive team and I got the opportunity to understand business strategy and implementation . And that was an automation engineering business , so hence the connection there and it exposed me to lots of interesting potential opportunities . And so then , after sort of running down that pathway , I finally stepped out into setting up a couple of startup businesses see if we can make it happen . So I joined forces with some colleagues and have for the last four years been working on a couple of startups , all linked to the meat processing industry , all engineering and all including automation .

Speaker 1 2:43

Excellent .

Speaker 2 2:44

Fantastic .

Speaker 1 2:44

Fantastic , fantastic , and I guess , yeah , the one we're talking about you can drop in the other startup as well . But yeah , I guess NMR have developed , or you and the team have developed , the Marble technology , which is trademarked as MARBL , and it's using nuclear magnetic resonance , which is I've sort of been doing a little bit of Googling but I'm relying on you to tell me more about it which is similar technology to any of us who've had an MRI , which is a big whirring machine which does all sorts of cool things to lining things up in your body . But I'm assuming NMR is similar . Can you give us a bit of a description on what nuclear magnetic resonance actually is ?

Speaker 2 3:21

Yeah , yeah , sure . And look , I'll keep it really simple as well , because I'm not the technical expert . We have two technical experts that could go into a lot of detail . So , yes , we're using magnetic resonance , and the Marble trademark name is linked to using magnetic resonance to measure fat . Yeah , cool In our case . Using magnetic resonance to measure fat yeah , cool In our case intramuscular fat . And in fact that's a lot of what they're doing when you have an MRI . And so when you have an MRI , you lie on that bed , you're in a big magnetic field and there's a whole lot of things that align protons . They all align , and so what happens is there's a pulse that disrupts those protons and then they relax , and the measurement of that relaxation is what we use to determine the amount of fat . So that gives you the overview .

Speaker 1 4:24

I suppose in depth , as most of us need to go .

Speaker 2 4:26

Yeah , yeah Well in an MRI they'll create an image and they'll do a lot of spatial work . We're not doing so much of that spatial work . Rather , we're just making the measurement in a volume of interest and then looking at the amount of fat that is in that volume of interest that we measure . So , yeah , that's the background to MRI , and then with our marble sensor , we've designed a specific style of sensor for our application . So , on MRI , you're lying in the middle of the magnetic field . For our application , so an MRI , you're lying in the middle of the magnetic field and with our sensor it's projecting a field into the meat or into the muscle . So it's a different design . And our point of difference is that we're designing these sensors for our specific applications . So for measuring carcasses or live animals . So the sensor sits up against the animal and then it projects that magnetic field in . We do a little pulse and then we measure the response and from that can ascertain the amount of fat that's in the meat or the muscle .

Speaker 1 5:40

Yeah , yeah , excellent . So I guess what we sort of know from what I've seen of I can't remember where we're at in MRI , I think it was HIP or somewhere and yeah , obviously fairly accurate or very accurate technology so of the options out there . Very accurate technology which means you can detect small things like marbling in meat at sort of somewhere between , I don't know , 1% and 10% or something . It's not a heap . And yeah , I guess from early days you would know that that technology was going to work . It would have been pretty clear that it had the capability to work . But then shifting it from something that's a concept that works into working out how to get IMF in carcasses to operate at chain speed and in processing conditions must have been a massive challenge .

Speaker 2 6:26

Indeed indeed . And so , yes , the technology itself is very mature , and so we were indeed sort of working with a technology where it was known and it's actually a standard method used for measuring fat . So we did have some idea of sort of the accuracy of the methodology before we started . So , yeah , then we had to take that and put it into sort of an industrial application , if you like . What I should have said actually about the sensor is the sensor itself , because it's got a whole lot of permanent magnets and it's very heavy , so they're 50 plus kgs . So another point of difference that we've had to put into our applications is combining the sensor with mechanical systems to hold it , manipulate it , and automation . So it's the combination of those things that give us our unique point of difference . Yeah , so anyway , um , you're going back to the application and putting it into processing conditions where we've got a leading product , um , which is measuring lamb carcasses , and so we do that hot at the end of the slaughter floor , and the reason we got into that was the Australians were particularly interested in that trait for measuring eating quality .

Speaker 2 7:59

So our starting point of going into this was market-led , and really that's an approach that we like to take so that we're getting the technology pulled through by the market , and in four years we've progressed from a prototype sensor to a production prototype system . So that was a unit that we installed temporarily in a processing plant and had that in so that we could calibrate and validate our measurements and the system and understand the accuracy of the measurements that we're making . So that work was done over a year ago in Australia , and then , since then , we've developed our first inline system and all the associated requirements of that , and we've had it in a processing plant and have worked through all the issues of cycle time , wash down conditions , durability , repeatability , all of those things . So we've now measured over 300,000 carcasses at line speed . I should say , though , that we gave ourselves an advantage we took a processing plant that was operating at the slower end of line speed , so working at four to five carcasses a minute , rather than the really high-volume ones that are around .

Speaker 1 9:25

Which makes good sense if you want to be successful , if you want success exactly .

Speaker 1 9:32

Don't go to the hardest first , and then obviously , if you can solve there , speeding up becomes a bit easier . So currently it's , I'll probably say 300,000 , that's a lot , and I guess one of the nice things about I guess one of the nice things about being heavy is that you would have had to automate , whereas if it was handheld you can kind of , and so that forced you down a line of development that others don't have to go down maybe , which is quite cool in the end , I'd imagine .

Speaker 2 9:59

Well , actually , yeah , I just mentioned that . It's actually an important point because the meat processing industry is labour constrained , yeah , and so they never have people around that can stand there . Yeah , and well , it's difficult to get people that can stand and do something . So , if you really want to add value , I think to have an automated measurement where you're just getting the data out and it's feeding directly into systems is the way to go . And you're right , because of the weight of our sensor , that took us down that path , but I think it's a key point of difference and adds a lot of value , yeah 100% and we'll direct people to your website .

Speaker 1 10:34

There is a video running with showing showing it working , um , and so we'll get people to have a look at that . Very good , so that's been in in carcasses and there's pretty clear why that's important and I think even that you know , I don't know , the at the pointy end where imf is highly is more important , and slow chain speeds okay , like you're not gonna . Maybe every carcass that gets killed is going to be sending into these premium markets or whatever . So probably . Yeah , it would be awesome once we have got everything flat out , but I think there's going to be value creation .

Speaker 2 11:06

Our goal is to get that up to every carcass being measured , and so , in fact , one of our current development projects is a twin sensor system that will measure at 10 to 12 carcasses a minute . So if we're achieving those line speeds , we're measuring every carcass .

Speaker 1 11:25

Yeah , awesome . Yeah , that's good . We don't need to settle for a second then . No , that's good . I guess one of the great things of that technology will be that you won't have to cut the carcass , you won't have to stick anything in it . It's after you've done it . No one knows you've been there , kind of thing . No one knows you've measured it .

Speaker 2 11:42

Yeah , very good point . The fact that it's non-invasive is extremely valuable , you know . It keeps the quality and the integrity of the meat and it allows us to take the same technology from carcasses and use it on live animals as well . So we're just putting the sensor against the surface , making contact , and then we're projecting into the meat , and that's a valuable trait in terms of our technology . That , combined with the automation , gives us a good point of difference .

Speaker 1 12:19

Yeah , particularly in a world where microbial contamination is the biggest thing that you're petrified about in a meat plant . Indeed , indeed , yeah . You've moved now to look at live animal assessment , which I imagine brings a whole new raft of challenges . How's that going ? And is that , I guess , is it just time now until it works well , is it already working or what's the future there ?

Speaker 2 12:44

Yeah , well , getting into this area again , it's been industry-led . So there was a call from the Australian industry . They were looking for technologies that could be applied for measuring live animals various traits , and so we put a hand up and said , ok , we'll have a go at IMF . And we were entering this after having done a lot of work with beef . So we measured marbling in beef and so we knew what we needed to measure , we understood the muscle that we were measuring in and the location on the animal and and we could design a sensor for the application . So we've only just really completed our first project , but it's been really successful . So we went in with a design specific for beef and applied it to cattle and have measured 300 odd animals and the data is looking really good .

Speaker 2 13:48

But it is early days and there's some refinements to do because that sensor is really big and hence the need for a counterbalance and manipulation system , and we know ahead of us there'll be some refinements and fine-tuning Based on our land work . It's probably . You know , we went from an alpha to a beta sensor . We'll probably be doing something similar to that . The biggest challenge , though , in live animal work is animal behaviour and the likes , and so that has been the new element . Yeah , yeah , so no , no , look it's looking really good . No , no , look it's looking really good . And you know , we hope ahead of us that , yeah , within a next revision we can potentially have it in the hands of early adopters . All going well .

Speaker 1 14:46

Excellent . I guess one of the other challenges , other than working with children and animals , is the need to move it around the place , because if you're going live , then I guess there'll be some feedlots or whatever where they put in a static system , but most times it'll be something that might move around for seed stock measurement or whatever . So I guess , depending on what the application is , but that will become a bit of a challenge as well , I imagine , if it's heavy and everyone needs to move .

Speaker 2 15:08

Yeah , so in our prototype work we took an existing crush and adapted it . We packed everything down into the crush and shipped it off and set it up again , and so you know it could be permanently installed or it could be a service offering where it's towed on the back of the vehicle and comes to a site , something like that . Yeah , cool .

Speaker 1 15:38

Throughout your career , you've been closer than most of us . I'm thinking in terms of what's already possible with sort of sensors and robotics and how that gets applied to the processing and differentiation of meat . What do you think ? Where are we going ? What's 10 years look like in terms of ? Like you've already said , labour is constrained . A major challenge is getting people in these plants to do the job is probably what stops us If it was automated . Machines don't need to sleep , they don't need to eat , they don't need to do anything . They can keep chipping away . What do you think ? Maybe 10 years is the wrong timeframe , but I mean I can't . Yeah , you can see this just always getting better and better and less humans involved .

Advancements in Meat Processing Automation

Speaker 2 16:17

Yeah , and look for me having been touching on automation and meat processing for some time . It's interesting to reflect back before sort of going forward . And it was actually the meat industry that was the first to come up with the production line concept of manufacturing . They were doing that before cars were being manufactured that way .

Speaker 1 16:45

Yeah , exactly .

Speaker 2 16:46

And so they were the early entr entrance to this whole manufacturing process . But the challenge of naturally varying product has constrained how the meat industry could develop . So if you now sort of think forward and you look at a car processing facility , it's virtually all automated , you know , with robots and vision systems and sensors . Leap forward , because our ability to cope with naturally varying product for automation applications is going to only become easier now with those tools . And so in the next , you know , 10 years , I think there'll be a massive acceleration of adoption of technologies .

Speaker 2 17:48

And look , there's already some very smart technologies being used in these meat processing plants . There's , you know , x-ray scanning systems and robotic cutting um , and machine vision is already being applied . And so I think , you know , sometimes there's these hockey stick curves of technology adoption , I don't know . It feels like everything's aligning such that , you know , in 10 years' time there's going to be a massive adoption of technology . And you know , I can see that some of the challenging routine sort of tasks that are done will be automated and staff will be being used for the skills that they have and their ability to be able to adapt and think , and so I can see quite a change in the role that automation and the likes of robots will have , as well as the role of people in meat processing . So in the future I can't see it being labour constrained as it is today .

Speaker 1 18:57

No , yeah , and that gets pretty exciting . You can just things like double shifts so you can get much more greater utilisation of investment and all this sort of stuff that really , at the moment , humans are the thing that stopped that happening .

Speaker 2 19:09

Yes , yes .

Speaker 1 19:12

Well , yeah , we're looking forward to all of that occurring , and I guess one of the challenges in ag innovation or any innovation but agriculture , particularly where you're developing something that's not going to be used by tens of millions , which you can sort of get away with in consumer technology but access to capital , has been a challenge for this technology and we don't have to go into this in any great detail , but I guess , just yeah , how the landscape could be better to make it easier .

Speaker 2 19:36

Yeah , look , as a typical Kiwi I'm going to say yeah , nah , yeah , good In that initially it hasn't been a constraint because we've had to prove the technology , prove the application and build the confidence up ourselves that we do have a viable product and that we can meet the market need . Now that we've done that , however , I see things changing and actually I think ahead of us . We're going to need capital to be able to realise the potential opportunities that we now sort of have formulated . So , yeah , it's a good Kiwi saying yeah , nah .

Speaker 1 20:22

It's a good one and yeah , I guess that's true once you can get investment to develop stuff pretty well in this or in both sides of the ditch .

Speaker 2 20:31

But it's that next level which makes it , and I should say that to this point we've been very lucky and we've had tremendous support from funding agencies that invest in new technology development so the likes of Meat and Livestock Australia , AMPC and the Sustainable Fibre and Futures Fund here in New Zealand so that has allowed us to get where we are without feeling like we're capital constrained . So that's been tremendous .

Speaker 1 21:05

Excellent , and we'll definitely give them a plug in the show notes as well . Yeah , because without someone tipping , tipping at the top , then nothing happens down the bottom . I guess in terms of expansion , which you've sort of just touched on , but yeah , I guess now that it is kind of working at chain speed , what's the ? So it's working in one plant , are we ? Yeah , is expansion in the ? Is that imminent or a few things to sort out before we go water ?

Speaker 2 21:34

Yeah . So for the lamb carcass grading , by this time next year I expect that we'll have a full system working line speed at , you know , on a large Australian plant , which will be a milestone itself . And then beyond that , the rollout of that particular lamb grading technology . I see it being closely linked to the MSA , sheep pathways uptake . So of course we got into this because they were after objective measurement technologies to be able to define key eating quality traits and so that sheep pathways rollout , like the beef MSA work should pull the technologies through and pick up the adoption of the technologies . So I can see it rolling out in line with the adoption of the likes of the machine pathways work in Australia . Here in New Zealand it might look a little different because it's not as structured in terms of the industrial landscape you know for key measurement traits . It's lots of companies , you know , doing it for themselves .

Speaker 1 22:53

So to say yeah , yeah , so just a quick discussion . Msa stands for Meat Standards Australia for those who are listening abroad and that's , I guess , a quality assurance game . Well , you get MSA rated and take a range of things into account and end up with a score that gives you some certainty around how that's going to eat or cook .

Speaker 2 23:13

Yeah , you're correct and look the other thing . So we've got the other applications of the technology also sitting there in terms of we can potentially hot-grade beef carcasses as well , and then we've got the measurement of IMF in live cattle and we're also having a go at looking at live sheep as well . So , yeah , we see that the lamb application is our leading application and then hopefully those other ones will be pulled through and trying to get market pull in terms of getting the technologies out there .

Speaker 1 23:53

Yeah , yeah , and there is definitely a lot of interest , particularly , as you said , in Australia . There's a lot of IMF spend . It's a big part of what we hear about in the genetic programs is people improving , looking to particularly the terminal breeds that got a bit off track in Australia . But , yeah , a lot of people using genomics to improve IMF and so we'll be looking down that line . So , yeah , live measurement technology would have a lot of people listening , interested , I'm sure . Is there other traits that magnetic resonance can do Like ? Is fat the thing that's easiest ? Or something like shear force ? I'm thinking that no other technology can measure other than when you take a bit and literally do shear force . Is there other things that we could pick up in that we haven't investigated ?

Speaker 2 24:36

Potentially that we could pick up and that we haven't investigated Potentially . So there's a lot of information that's hidden in the sort of response , the relaxation signal that you get back , and we have looked into it and there is more there to be discovered . The challenge is getting it out and doing something with it in the time that you have , when you're wanting to work at production speeds and so , for example , you know I don't know potentially tenderness , potentially water holding capacity and the likes . It's actually a very advanced sort of technique when you use it in different ways . It's just the way that we're using it at the moment . It's being used relatively simply and relatively directly for a single measurement , just for fat , yeah .

Speaker 1 25:36

I guess it's a bit of chicken and egg . Once you've got hundreds of thousands of carcasses being measured by it , then you can do other things with those carcasses to work out what the machine's telling you . But you need that flow through to make research easier , I suppose , and having a defined product like an IMF measurement will certainly get you down that line .

Speaker 2 25:53

Yeah , and as we do more work , we may well our key technology leads . Evan McCarney and Robin Dykstra have been working in this area of magnetic resonance for their careers , and they know how to dig out good information from an NMR signal , so they would just love to get their teeth into that .

Speaker 1 26:18

Yeah , excellent . Nothing like a couple of focused scientists to get into the detail .

Speaker 2 26:23

Yeah .

Speaker 1 26:24

Thanks very much , barbara . Great to have a chat and really looking forward to seeing that technology roll out . The last question well , sometimes we ask at the start , sometimes we ask at the end , but we've been asking guests lately what's the last thing you changed your mind about ?

Speaker 2 26:35

You're going to laugh at this , Well .

Speaker 1 26:49

I get some interesting answers .

Speaker 2 26:50

So probably because because it's work , I didn't want to be involved in in any projects to do with live animals .

Speaker 2 26:53

And here we are . Here we are working with live animals . The back story to that is one of my previous projects was a real challenge . So we worked desperately hard to do an automated system for dairy applications and it was hard and there were easy wins early on . But to transfer it through to a full commercial product presented us with some real challenges , all to do with sort of animals and natural variation of animals and movement of animals . So I vowed at the end of that that no , no more working with animals . And here I am three years later and we have a live animal project on the agenda .

Speaker 1 27:43

Excellent . Well , I'm sure the industry is glad you changed your mind on that one . And yeah , certainly here we've been trying to do some work and all the science lines up , but animal behaviour is the biggest challenge of any technology development that involves them .

Speaker 2 27:56

Oh yeah , so we're not alone .

Speaker 1 27:59

No , that's for sure . Excellent Barbara , do you want to give your other startup a plug ?

Remote Meatpacking Robotics Development

Speaker 2 28:04

So Mimeo Industrial is our other startup and it's looking at shadow robotics , and so we are well advanced with the development of an application that's picking and packing from a boning room belt and using the smarts of people to operate that system remotely . So very soon we will have two robots operating and packing , and they'll be operated remotely from an office .

Speaker 1 28:43

Perfect .

Speaker 2 28:44

Cool yeah . So going back to your question around , what does a meatpacking plant look like in the future ? I see a lot of those .

Speaker 1 28:53

Yeah , yeah , yeah , no , definitely , yeah , yeah , for sure , excellent , all right , barbara , I'll let you get on and come up with another idea and appreciate your time today and look forward to seeing that technology rolled out .

Speaker 2 29:04

Very good and , mark , thank you for the opportunity . As a little start-up , it's wonderful to be able to get out and talk about what we're doing .

Speaker 1 29:12

Excellent . Thank you very much .

Speaker 2 29:14

Great .