 It's me who is saying thank you to you for giving me this opportunity to give an industrial perspective on the research and here at Arla I'm responsible for you could say overall responsible for our university research collaborations and we have around 80 projects running at any time, collaboration projects with universities. So science is really you could say the basis for our work and for our products. I would like to hear give you a few examples of how detailed scientific input is relevant to our products within this area and methodologies that we are together on these days. And furthermore I'd like to bring your attention to the vast amount of unknowns and difficulties in extrapolating fundamental science into industrial applications partly due to the scale difference. You probably know and hopefully appreciate the qualities of Lerpheck butter and milk fat is probably the most complex natural fat blend. So and it is of course behind the rich taste that we have in butter. So there are approximately 400 different fatty acids in milk fat. Milk fat crystals are known to what they are important and fundamental to the qualities of butter. And they are well described here are some figures from Kaufmann showing that alpha crystals are formed initially when your flash cooling milk fat and then over around an hour they transition into beta-marked crystals and hardly any beta crystals are formed. By the way in margarine, plant-based margarine it's very different and beta crystals can form and cause brittleness and graininess of the product. But it's not like that for milk fat and this kind of detailed knowledge of course helps us when developing and troubleshooting our products. So if we have brittleness and graininess and that the water for instance is not really kept in the product as it should be then we can reason that this is not caused by a transition from beta-marked into beta crystals. The root cause must be found elsewhere. And like this figure shows, well I borrowed it from the web page of the University of Gulf showing a schematic representation of butter structure which contains both crystals and fat globules and fat that is the continuous phase after phase conversion and also droplets of water. So it's a very complex you say higher order structure that we have here and that is where we should look for causes of product defects then because it's not the form of the crystal that is something wrong with. Of course there's quite a lot still to troubleshoot to find out what is the defect but it could be growth of crystals or it could be interactions between crystals, fat, water droplets. So there are many other places to look but at least when you can rule out something you are one step further ahead of finding out what's the problem or developing a better product. Then however when we are doing science we are often trying to simplify, purify, keep everything but one thing constant and as controlled as possible. So when studying milk fat crystallization for instance this is usually done using and hidrus milk fat but in reality we have water, salt, proteins, fermentation metabolize and so on and so forth in our products they are much more complex than those pure systems model systems. And also when we are scrutinizing a sample scientifically it's usually a very small sample for instance here when loading something into analytical machinery it's in the micro liter volume range whereas foods are higher volumes and production scale is of course huge and for instance butter at RWA producing it at three sites globally and the production amounts to more than 100,000 tons per year. So well that's why I gave rise to my title here is a matter of scale also because the control we have at lab and the control in production scale are also on different scales. So we are cooling down, we are heating, we are centrifuging, heating again, mixing, pumping, doing phase conversion, adding ingredients, heating again, storing, packaging, transporting and then at some point the consumer takes over and well doing not so scientific temperature cycling in the fridge and on the dining table. So the life of the product is a very long journey through a lot of different conditions and hopefully still the consumer can enjoy a great product but how do we get a better understanding of all those structural changes during the processing intended and not so intended conditions that the product is meeting during its life and what might be the importance of the heterogeneity that we have in raw materials and processing conditions well just imagine the scale if we are pumping could be 10 or 50 cubic meters of liquid be that milk or cream into production or storage tanks it takes us hours to fill a tank or empty a tank. So there might be quite a difference between the first milliliter and the last milliliter entering or exiting that tank. So that's a dimension that we usually do not look into at lab scale at all but what does this this difference is what do they actually mean how important are they how do we study them how do we scale this down into the lab and address it I think there's an entire world of complexity that we are not really oftentimes trying to address well you can think about this a lot and please do so now turning to proteins milk proteins and looking at the majority of the milk proteins the caseins they're known to form these super molecular structures composed of both the four casein families but also nano clusters of phosphate calcium phosphate and the structure and function of the casein micelle as a carrier of both protein and calcium phosphate for bone growth is I would say one of nature's wonders and it's also been the the the topic for scientific debate on structure for for many many years I'm not sure we have reached consensus now but at least we have reached quite good models that explain all the phenomenon that we are experiencing working with caseins but there's a lot that still is not described and understood in detail just some examples here a graph from you at all at allies showing how temperature affects the solubility of the caseins so in low temperature you have higher solubility of caseins and you have a lower stability and likewise at higher temperatures the caseins are not soluble but but but but more born in in in this colloidal state in the micelle so well we know this but the kinetics is not really well described yet and it's also a bit difficult because some of it happens fast and maybe something happens slowly and how do you really capture this likewise if we're looking at at the effect of of heating and cooling and and looking at here is the skim milk and again the relative change over time when you're heating well I just need my pointer here I think and so when heating you you have this well change in pH and in temperature of course and then also turbidity but then when you're cooling well temperature and pH are following the same way as when you heat it but but the turbidity it's much slower returning here at this treatment of skim milk returning to the same turbidity as it started by but if you are if you're heating it not to 20 degrees but to 40 degrees it seems to maybe not return or at least not in the timescale studied here so there is a there's a there's a change here so of course it's complex because you well you're actually really to notice this system you will need to study it at a rate of different parameters so it's it's quite a huge experimental space that that we we don't have a full model of how this works but still you can utilize parts of it there's an example by by Schaefer with a process that you can use to to partly purify beta-casin from the from the case in micelles beta-casin is is diffusing out of the case in micelle partly at low temperatures so by microfiltrating and diafiltrating at high temperature first you can purify more or less the case in micelles and that when you then store them at cold temperature a lot of the beta-casin is leaving the micelle and going into solution and so if you microfiltrate and micro diafiltrate again you can wash out the beta-casin partly and in this way and by by by full control of time and temperature you can optimize a production of of beta-casin and beta-casin depleted casein but here you are controlling time and temperature really well to do this and well another thing that's happening if you look at calcium also when when when concentrating and this is a work by by Coretic present here I guess somewhere in the audience and and leave back from 2014 and if you're looking at at the the calcium leaving the case in micelle depending on pH depending on on whether the sample was heated or not before the study and then concentrating it you get very different responses of calcium leaving the micelle and going into solution and knowing how important calcium is then you can start wondering what this can be used for and and not least which effect it might have in in all kinds of dairy production processes again here in the lab you have a well-controlled system working with few milliliters exactly controlling the temperature time and everything as as you do in the lab but then in in in the industrial reality we are collecting milk and farm it is collected over over sometimes several days at the farm before it's collected by by our tankers and of course the milk is cooled when it is stored at farm but we say one tank one farm tank of milk has many different ages because there are many milkings going into the same cool storage so the time stored cold will be a mixture over time so in one tanker leaving from a farm you will have milk that has been stored at cold temperature for different times now mixed then that instance the dairy and it is again stored maybe it is separated it is re-mixed standardized to different fat contents and pumped around and heated and and stored again and so on and so forth going into a lot of different processes yielding the different products it's just to say we do not have the kind of control over time and temperature that you have in the lab at all so so that's a an extremely high complexity if you would really want to understand exactly what is happening so you're just understanding the raw material going into the process here and well so when we're doing production sometimes we are using you know several silos of milk for one production and and it means that there will be quite a heterogeneous raw material and some phenomena might be based on the mixture you have of raw material in your tank what are actually the the histories of each micelle in such a silo are there some of the blends and mixtures here that are that are beneficial or detrimental to the to the product what about if you're storing it storing it cold and then filtering it then you will lose calcium would that be good or bad well depends on the product and the process following you can easily maybe have too high calcium in serum and you will get fouling and clogging of of heat treatment systems and yeah we also know that calcium ions are of course very important for protein-protein interactions so it will affect a lot of products likewise beta-casin the most mobile you can say casin leaving the micelle it's also caused known to cause bitter taste quite easily so so taste might be you know a variable due to this lack of control you could say so but there's a huge scale that makes it more complex than the way we usually study in in the lab so this is just to say what is the reality of industry and maybe bringing some attention to studying some of the phenomena happening on that scale bringing that back and looking at the structures and structural changes happening under these conditions on those timescales and with those complexities of size that we have in industry and well that was actually the talk I wanted to to give you as an input into the workshop here thank you Peter maybe you can close down your presentation so we can see some faces yes great so we have some thank you first for the insight in some industry perspective that's always good to have and we have some questions here in the chat first from Tommy in some products you mix milk fat with vegetable vegetable oil like in the Swedish product Brigotte that means this means that lipid diversity changes in such is such mixed products less challenging in terms of maintaining product quality well I guess maybe I should jump in Peter yes thank you I had a feeling that you might want me to I was thinking very much about you well I wouldn't say that it's a smaller challenge when we start manipulating the triglyceride composition it's a different challenge so basically it brings us to trying to understand you could call it oil binding or interactions between liquid oil and milk fat and it's not just a challenge during production it's very much also a challenge in consumer situations suddenly it becomes more of an important point to study structures at slightly elevated temperatures like the temperature that the product would have on a at a dinner table so basically we have fewer building blocks to make the crystal network and that makes the product maybe more fragile in some aspects so definitely understanding these interactions between different kinds of triglycerides is a very important one for us great we have another question here with scaling there is automatically a higher negative impact when failing and we all know that quality will fail over time a modern industry will also invest in high tech equipment with which increases the competence needed in all areas of the business how do you cope with this transition I'm not sure I really did you understand the question I can elaborate on the question if you want yes please yeah well I have a background in the in the medtech industry and I know from from that period when I worked at this company producing incontinence products we saw this transition from having equipment in the production from when you were adjusting the older equipment then you can use like wrenches and that kind of stuff to steer the material in the production line but then the more modern machines required more technically skilled staff which was quite hard to find and the people who were more knowledge in programming and so on to be able to run these machines and I believe that it's like the same transition in every industry going from yeah old machines to more modern machines yes I think that well we are moving in you know having more and more automated processes which of course increases the control but I think that also a lot of the machinery just the just the amount of of stainless steel that we have as our production infrastructure and the investments in that it gives us definitely some limits we cannot just you know move forward and have the most advanced equipment at any time we need to also to it's simply too too expensive so we need to wear down our equipment and productions equipment before we are investing in you and improved ones but we're trying to you know modernize our production system but it happens over time gradually and I think we are increasing in control we are also more and more trying to incorporate sensors in the production systems and we are collecting a vast amount of data so we are also looking forward to improved methodologies in in big data analysis because you know we're collecting data but but getting knowledge from the data from production data that's that's a very difficult task that we are I don't think we are addressing that enough I think there's a lot to learn from it if if we could address it the right way I think a lot will happen over the next few years in that field yeah thanks okay let's move on to the next question do you see any role for neutron x-ray scattering in our last R&D process for sure yes but I cannot right away point exactly to you know just giving you the perfect project to run but but for sure just I mean just looking at the case in my cell and and and the huge complexity that it has and how it affects our products for sure there will be a lot to learn and a lot of control that we could have if we understood the system better so of course it doesn't necessarily require neutron scattering but but but just looking at the structure of of the case in my cell definitely scattering techniques are relevant to study them and likewise with the fat crystals you could say if we can develop more into being able to handle more complex samples and still get you know useful data from analyzing a more complex sample then it would better you could say address the complexity we have in a in a real product because there's still quite a long distance as I pointed out here in my presentation from from you know a simplified system and into the real thing so bridging between that will definitely make it easier to address you can say industrially relevant questions in a way where you embrace like that complexity I know it needs to be broken down but we but that's I think where we really need to work together to try to to mature this area and another one here what are complex samples to you well it depends it depends but one example here would be looking just at milk fat without water in it I mean then you could have you're looking into cream or you could need to process cream it could be processed into butter and then you could even then have mixtures like like the blends used for for spreadable products so you would go into you could say higher and higher complexity and likewise looking at milk understanding milk or maybe skim milk as simplified but again skim milk isn't simple you have more than 300 different proteins and I mean and and 400 fatty acids still in trace amounts and other compounds so it's and then you could say a lot of science has done on maybe purified whey proteins but you mean that's that's far away from the complexity you're having the you could say the natural raw material being milk but then again when you're processing it into cheese you have a completely different thing so so for you could say you have simplified models of different foods dairy products so but but going into really trying to get a really pure substance to study this oftentimes the starting point but then we need and oftentimes it's needed to understand you could say the the building block but it's just there's a lot of work to do to embrace more and more complexity moving towards the foods that we are actually eating thank you for that I have one final question I'm I'm actually known as the worst summer worker in dairy industry ever I produced hundreds of liters of strawberry yogurt without strawberry once but I'm not only a producer also a heavy consumer so if you just elaborate a bit what how will me me as a consumer how will I notice the increased focus on these technologies in the future I think for a big part you would hardly notice because I think we are not so good as human beings to know but noticing small improvements we just take them for granted and I think we we as a food industry we are we are quite good at at avoiding consumers experiencing product failures and we're doing our best to avoid that I think we will be better and better at avoiding that and better and better at securing high quality all the time reducing food waste so I think it it will matter for you could say for the planet and maybe maybe the quality of the product will also increase to some extent of course you you can also envision new innovative products coming out of this but I don't know I think of course being part of the research and development department I should understand here and say that it's not so important to innovate but I mean we are also very conservative when it comes to eating habits so I think in general maybe it's more important for us to be sure that the food is good of high quality all the time than it is to have something new mind-blowing food to eat I mean a lot of dairy products are very appreciated as traditional products so securing that we are producing high quality all the time I think it is very important but it is fun if we can if we can understand some structure formation that can make us make new new food structures I mean milk chips or whatever that that's that's only fun and and great of course we should we must do that too thank you