Welcome to "AigoraCast", conversations with industry experts on how new technologies are transforming sensory and consumer science!
AigoraCast is available on Apple Podcasts, Stitcher, Google Podcasts, Spotify, PodCast Republic, Pandora, and Amazon Music. Remember to subscribe, and please leave a positive review if you like what you hear!
Gern Huijberts is the global quality director for Cargill Cocoa and Chocolate, with responsibility for businesses in Europe, the US, Brazil, West-Africa and Asia-Pacific. In this position, he manages the global quality organization for food safety, quality assurance and quality control as well as regulatory affairs.
In previous roles he has worked at international food companies like Heineken, H.J. Heinz and FrieslandCampina in the fields of Research & Development and Quality Assurance. Throughout his career he has always had a keen interest in sensory analyses, both as an active member in sensory panels as well as from a research perspective. He is very interested in artificial intelligence and machine learning for applications in quality assurance of food products.
Gern Huijberts holds a MSc degree in molecular biology from the Wageningen University and a PhD in biochemistry from the University of Groningen in the Netherlands. He is married with two children and lives in Doorn near Utrecht in the Netherlands. In his free time, he likes to take long walks with his wife in the local national park and he plays guitar in a rock band.
Transcript (Semi-automated, forgive typos!)
John: Gern, welcome to the show.
Gern: Thank you.
John: Thank you for being on the show. It's really a pleasure. So, Gern, we've known each other for a long time. I mean, I think maybe 10 years or so, now at this point, but, you know, maybe for our listeners who haven't had a chance to meet you, it would be good if we could start with your journey into the food industry and then into sensory consumer science. How did you get to be, well actually, I think it's an extremely fun job working on chocolate.
Gern: Right. Well, actually, I started in beer. After my PhD, I joined Heineken as a researcher, you know, there was an R&D department and so we did research on fermentation and obviously when you work at the brewing company, sensory is very important. So I was able to join their sensory panel. And actually, that was the first time I got into contact with the sensory sciences. And it was really very good with actually being paid to drink beer. So the sensory panel of Heineken is quite interesting. So I started out as R&D functions with Heineken, did a couple of jobs with Heineken, then switch to H.J. Heinz where I got into the quality assurance type of work. I was doing quality coordination as we called it for Heinz tomato ketchup where obviously also sensory was very important. So, you know, we looked at all kinds of quality aspects of tomato ketchup starting growing of tomatoes, the variety of tomatoes and the different types of processing, different types of ketchup you can make from that. And their sensory was very important. You know, you have the secret Heinz tomato ketchup spice mix. And so, you know, that was the best kept secret within Heinz, I suppose. And so we did a lot of work and that's actually where we met when we were looking at different types of maybe new flavors of ketchup and how we could develop that and that was also where I think maybe developed an even more professional interest in sensory sciences. I remember the workshops that I followed given by yourself on landscape segmentation analysis, and it really sparked my interest a lot. So with Heinz, I did QA and later I joined the R&D Department where I was also responsible in their new innovation center to set up the sensory research part where we had the pleasure of setting up like consumer panels, expert panels, you know, really exciting time. After I left Heinz, I joined FrieslandCampina which is a big dairy company in the Netherlands, business-to-business where I was responsible for quality and regulatory and food safety. And then I went to work for Cargill. And for Cargill, it was cocoa and chocolates. I didn't know a lot about cocoa and chocolate when I started. But it turns out to be a really very interesting. The supply chain is very interesting. It's an agricultural product with all the special things that come right seasonality, different varieties, different origins. So it's a fascinating field. And then bottom line, of course, you know what, chocolate, you know, people buy that because it tastes good. That's why people buy chocolates. So, you know, sensory here is extremely important. Quality is extremely important. I don't think a lot of people know Cargill for like, say, the cocoa and chocolate because it's all business to business. So you cannot buy like a Cargill chocolate bar in the supermarket. With Cargill supplies, like all of the big players like Nestlé, Mars, Mondelēz, they are customers of Cargill. So that is how I got into, let's say, quality of cocoa and chocolate.
John: Yeah. So there's one thing I want to ask you about in two big areas, I want to talk to you about. One is our shared interest in machine learning and then other of course is just the spice, the whole question of chocolate sourcing and the whole supply chain, very interesting. So let's maybe start with the machine learning piece and talking about, you know, what you see in kind of short term and then long term is the most promising application in machine learning and then I think with the time we have left, let's talk a little bit about the supply chain.
Gern: Alright.
John: What are your thoughts about machine learning within in our field?
Gern: You know, I think that artificial intelligence machine learning opened up huge possibilities to do all kinds of analysis that we were not able to do in the past because it's about processing a lot of data. And when I was like in my previous positions, you would measure quality of products and you would measure a few parameters like maybe the color or the bitterness or the sweetness. And the basis of that you would try to sort of manage the quality of those products. These days with all of the analytical techniques that we have available right now, like nor scanning, for example, but also electronic noses. So we have a lot of analytical techniques that will give you a huge amount of data and possibilities of data processing these days. And all the techniques that have been developed in the past 20 years are so powerful that there is a lot of things that you can do now, which were simply impossible in the past. So what gets interesting is if you were able to, let's say, make predictions on how you could optimize your processing conditions based on the varieties of beans that you can buy or varieties of beans which are available, you know, there are seasonality things. So, you know, sometimes the harvest in West Africa is a little bit different than the year before because the rain came three weeks later or it was a little bit colder or a little bit warmer. These things affect like the characteristics of the cocoa bean than using data analysis and artificial intelligence, you could maybe predict how should we say, optimize the production processes in order to get the same sort of quality and questions like that, you know, which are not so easy to do without the big data possibilities. Those are interesting things. This is just one example.
John: Right. But I think it's very interesting on because I think that oftentimes when you have processed food, you've got this desire, people tend to think about, okay, well, what can we manipulate on the analytic and the formulation side in order to achieve different sensory outcomes? But you have the kind of inverse problem, but maybe the converse problem in the sense that you actually want consistency.
Gern: Exactly.
John: In the output and you have this variety, you have your variabilities on the input and you're trying to react to what you know about the input to essentially stabilize things to get better.
Gern: Yeah. And interestingly enough, it’s funny you should say that because when I was working for Heineken, that was exactly what they were trying to achieve. Make Heineken can taste the same everywhere in the world where they have different breweries, different water qualities, different mold qualities. And still you want to make the same beer. The same with Heinz ketchup that should taste the same all over the world where you have different tomato varieties and different factories, different countries, different climates, and you still want to make that same product. So that is the power. That is actually what is important for like big global brands. They don't want to have these, let's say differences in taste. They want to have the same taste.
John: Yes, right. And then if there are differences, they controlled them.
Gern: Exactly.
John: It is interesting because I think, you know, at least in America, the mainstream beer industry gets a lot of criticism from the craft brewers. Right? Because brewers, you know, there's a lot of beer snobbery in this kind of thing. That actually a company like Miller is solving a much harder problem. It's much, much harder to brew millions of gallons of beer and have it taste the same to brew small batches and make it interesting.
Gern: Well, you know, it's a little bit on how you look at it. When I was the manager of the pilot brewery of Heineken, we would compete with the big breweries for, let's say, the best Heineken beer. There was a sort of internal competition. And, you know, until all of the beer were tasted by the sensory panel and they would get the score. All of the breweries had to participate in that. I was managing the pilot brewery and we would participate in that, too. And so but then, like the guys from the big brewery who say, for you, it's easy because you can put all of your attention, you can put all of your people on that one specific brew. And then I said, yeah, but for you it's easy, but you can blend everything away so we can do that because, for us, it's a one-off. It's one small vessel. If it's wrong, it's wrong. And so sort of the big industry is all about blending and mixing and that is also how you do achieve consistency.
John: Yeah, that's interesting. There's always a bit of the grass that is always greener at somebody else's job so it's easier. Okay, well now let's talk related point about the supply chain because I think it's really interesting the measurements that are being made along the way. You know, I didn't know before we started talking about this fascinating process going from the beans being grown down predominantly in Africa. Isn't it the case that?
Gern: Correct. Yeah.
John: Take us through this whole process. I think this is really quite interesting to learn about.
Gern: Well, you know, the cocoa beans, they grow in pots, on cocoa trees and cocoa trees, they're in a specific region of the world. So that's like West Africa but also like the northern parts of Brazil, Asia-Pacific. So it's an area around the equator which is suitable for growing cocoa. And so normally there are two harvests. So you have the main crop, which is in like the period October to, let's say, December and then there's the mid-crop, which is around February to April. Mid-crop typically has a poor quality than the main crop. So what do people do is they harvest these pots and then they ferment them. So they actually did. They put them on a pile, put banana leaves over it and just leave it for a couple of weeks. So that's where the fermentation starts. And after that, they take away the beans from the pot. So they break the box, take out the beans, then the beans dry. Normally they dry in the sun and eventually, that's the process of what they do. So then the farmers collect the beans, put them in jute bags, send them to a corporation, send them to like a place where they can sell the beans, and that is where middlemen and companies buy the cocoa beans. And then from that, you either get local processing or export to like big customers in the rest of the world. So countries like Ivory Coast and Ghana, they do something like 70% of the total volume of cocoa and then the world comes from those countries. So most of that is exported. There is local processing. So we also have two facilities, one in Ivory Coast and one in Ghana. And in that area, you see also the big competitors, they also have the processing facilities. But a lot of the beans also get transported to Europe in our case. And most of that goes to Amsterdam and interestingly, a historical fact, Amsterdam has been like the major cocoa port in the world from, let's say, the 18th century. So there's always been a lot of transport and also cocoa processing north of Amsterdam. And that is where you still have the also cargo has the big factories over there. So Amsterdam is a central point in, let's say, the cocoa trade. So we have our processing facilities over there. So the cocoa beans are converted to what we call cocoa liquor, and that is basically grinding. So it's separation from the husks, sort of as the shell, as we call them from the nibs, the inside of the cocoa beans. And then you have a process of roasting and conditioning to get the right color. And then it's from that on, it's milling and mixing where you get when it's really fine, it turns into like a thick syrup. That's what we call cocoa mass that is converted into cocoa powder and cocoa butter. You press it. So you basically pressed the butter from the cocoa powder. Cocoa powder can be used for all kinds of applications. So you can make chocolate from it, but you can also use it for cocoa drinks. You can make toppings for everything where cocoa is used. Cocoa butter is used for white chocolate. And then, you know, when you have the cocoa powder and cocoa butter, you can start making chocolate. And again, you know, that is mixing and grinding that's what chocolate making is all about. So it's mixing cocoa powder, maybe additional sugar. If you want to make milk chocolate, you'd have to add some milk components to that. You can add flavors. You can add like nuts, hazelnuts, whatever you like, and that to the chocolates.
John: As part of going that far with it or at one point?
Gern: So Cargill has like a lot of being sourcing. They have like the cocoa facilities and they have chocolate factories.
John: I see.
Gern: And that's what they sell a lot of cocoa products like cocoa powder, cocoa butter to customers, but they also make chocolate. So they also sell chocolate and a lot of liquid chocolate which is used by customers to make their applications and also like solid chocolate in blocks which can then be used to melt them again. And we also have a few, like smaller facilities, which are more in the gourmet business, which really make like the very fine art to some chocolates for high end applications.
John: I see. Now you've been downtown, is that correct? That is kind of a central part of your job. I know you you've visited some of these farms, is that right?
Gern: Yeah. Because, you know, when you and I think about a farm, you know, you see like a big building and green fields and cocoa farm is nothing like that. So I remember one day when I was in the Ivory Coast, we went to see a cocoa farm and we went in a car and drove down the road for an hour or so. And then they stopped in the middle of a forest and there was a small metric. So why are we stopping? Well, this is the cocoa farm. So the cocoa trees are just inside the forest. And so it's mixed with everything. You find coffee plants over there, you find bananas over there. So it's nothing like an agricultural organized type of farm? No, it's a plot of forest that is owned by somebody, and that is where they grow the cocoa trees, and that's where they harvest the Cocoa Puffs and where they do the fermentation. That's basically what to do everything, but it's just the forest.
John: At what point does Cargill get involved in terms of, so the farmers have the beans and then at what point if you were going to start making measurements, you're going to look at what point do you all start making measurements?
Gern: A lot of farmers are organizing corporations and the corporations, they buy the beans from the farmers. You should imagine that your average farmer does something like 400-500 kg, that's it. So that is a relatively small amount and that is where they need to live from for the entire year. So, you know, when they sell their harvest to the corporation, it is a very important day in Ivory Coast and in Ghana because that's when they get money. And that is what they need for, let's say, half a year or three quarters of year until the next harvest. So that is when they buy the books for the children to go to school. You know, so that is the cocoa harvest for a lot of farmers is extremely important because that is what they live off. So typically they sell to corporations and the corporations is where the first quality measurements are done. And typically they look at moisture level, they look at the quality of the beans, whether they're, for example, moldy, what are the beans are damaged. They might measure the fat level. They do some things called the bean counter. They look at the size, the bean counters, and the number of beans in a specific weight. So the higher the bean counters, more debates. So if you want to have big beans that is a better quality. So that is the first time when they actually start measuring that and that is continued throughout the supply chain. So the corporations sell it, for example, to Cargill or to another trader where, you know, the harvest from various corporations is collected and blended into specific plants. They might do cleaning processes. They have to remove like sandstones, things that you find in cocoa beans. They might do extra drying if they wanted to. Typically, they prepare what they call the export lots or batches that they sell to local processors. So that is and so you have a number of these we call them bean cleaning stations, because actually, that's where to collect the beans and where they make those plants. They're controlled by the government. So they do analysis there as well. The government, but also the company does analysis so that everybody has this individual set of data on the basis of that information. They make the lots where they combined from different farmers. They combine the beans and then send them, for example, to the Netherlands or sell it on the market to other companies or process it themselves. But so it's these basic quality parameters which are actually message throughout the supply chain. When it arrives in Amsterdam, they might check that things like moisture might change during transport things. The first free fatty acids, free fatty acids is basically in our flavor. That increases in time. If you store beans for a long time, they put more free fatty acid. So that is also something that is checked because that is important for the process. So there's a lot of quality parameters which are being checked throughout the supply chain.
John: And then then that would be ultimately the type of information that the long term vision here is to have a model where when you....
Gern: Exactly, exactly. Because now plants are basically made on like origin. They might do some separation on the free fatty acids level. They might do for maybe some different types of beans or different. But for example, if you want to have there are some organic beans, for example, or rainforest beans so you have all these different clusters. So that is how currently these plants are made. What at the moment, no one can do is actually make this plant based on an outcome later in the process. So we group things on things that we can measure. But it would be really very interesting if you could make plants based on, say, optimal process ability for this specific factory or if you wanted to achieve a certain flavor in a certain market. And those are things which are not possible at the moment.
John: Right. And where are we humans in terms of a human expert? How well can a human expert I mean, I guess a human expert would look at the beans, but they wouldn't have access to all the I mean, I guess they could in principle have access to the history of the beans, all the measurements they're made on. But presumably, there are some people who are better than others at figuring out how to blend the beans. What is the human capacity if you are going to try to have a machine emulate a human? How do the human experts perform right now to this task?
Gern: I think if you have, like, true cocoa experts, they could look at beans and just maybe smell it or touch it and they could tell you something about the quality. Just as like a brewer would taste a little bit of the malt and he could say whether that is a good malt or bad malt. But it's fairly basic. So they can make like a call about quality when they see based on their experience.
John: Right.
Gern: But, you know, answering any more complicated questions, you know, what is going to happen if I blend these two or what is going to happen if I don't blend these two? But for example, those questions at the moment, I don't think. A human expert could answer. I don't think so.
John: It's very interesting. Yeah, I mean, well, you see superhuman performance all the time for me. So this would be. Yeah, because there's way more information that the machine wouldn't have access to as well.
Gern: Exactly.
John: Yeah, just like a driving car, with self-driving cars, we forget that a self-driving car can monitor everything around the car at all times, right? They have capabilities that humans just don't have in terms of the amount of information coming.
Gern: Exactly.
John: So, yeah, well, Gern, I'd love to keep talking about this, but we are almost on time. I do want to get your advice because I think that you're someone I've always enjoyed talking to. I mean, your background, I think, is in biochemistry?
Gern: Correct. Yeah.
John: But I think you've been through a wide range of experiences. So what would be your advice for someone just entering the field now? You know, maybe they just got their master's Ph.D. in say Food Science or maybe experimental psychology that are coming to the field, what advice would you have that person?
Gern: Well, I'll say what I benefited most from during my career is to be interested in learning things which are not in your field. The ability to be able to talk to people in other disciplines I think that that has been very good for my career, and if I would have stayed within biochemistry, just focusing on like interesting cellular processes, I don't think I would have gotten to the position that I'm currently in. I think specifically because, you know, I did step out of research and went into the industry. And when you go to the industry, you learn about completely different problems and challenges and you get into contact with people who come from very different disciplines. And, you know, I've always enjoyed learning from these people and understanding where they come from, what they're interested in, what specific expertise they have, and then combine that with the knowledge that I have. So, you know, if I could give somebody one advice is just stay curious and be interested in what other people do and try to look outside the four walls of your, let's say, your specialty or your niche area. Don’t become a specialist, become a generalist. Be able to talk to a lot of different disciplines. You don't need to be a specialist in everything. If you can ask the right questions and understand like most of the answers, that is a great help.
John: Yeah, that is a great advice. And how can somebody get in touch with you if they want to apply for a job at Cargill? I mean, I can imagine working for you would be great if somebody could.
Gern: Of course. Yes. I think the easiest way of contacting me is through LinkedIn. And, you know, actually, I think that my first name is unique. I don't think there are many other people in the world with that. So if you just look for Gern on LinkedIn, I think you might be able to find me. But LinkedIn is the easiest way.
John: That's good. And we'll put the link in the show notes. Alright. Anything else that you want to add before we wrap up?
Gern: No. Thank you very much. You know, I always enjoy talking to you very much. I think you've done some fantastic things in the past on sensory research. And not so long ago, I was even reading one of your papers about power of sensory discrimination tests and...
John: Oh, yeah.
Gern: That's quite a long time ago. But I still find it very interesting. There are lots of things that I would like to talk to you about and applying statistics and quality research, but maybe some other time.
John: Maybe over Heineken and some chocolates.
Gern: Exactly.
John: Thanks a lot, Gern for being on the show.
Gern: Good. Thank you very much for having me. Really enjoyed it.
John Ennis is a leading expert in sensory science and consumer research, with extensive experience in statistical analysis and product development methodologies.
