Become a Creator today!Start creating today - Share your story with the world!
Start for free
00:00:00
00:00:01
How dynamic stock control can help poultry companies
20 Plays31 minutes ago
Max Grell, Ph.D., founder and CEO of BlakBear, explains how sensing spoiling gases inside packaging in real time can help companies better organize stock and reduce waste.
Recommended
Transcript
Introduction to the Future of Poultry Podcast
00:00:00
Speaker
Hello and welcome to the Future of Poultry podcast series.
00:00:12
Speaker
I'm Mark Clements and today on the Future of Poultry podcast, we're going to be looking at sensor technology aimed at better stock control through directly measuring spoilage indicators.
BlackBear's Sensor Technology in Stock Control
00:00:24
Speaker
To discuss this, we're welcoming to the podcast Dr. Max Grell, founder and CEO of yeah UK technology company BlackBear, to look at the technology that BlackBear has developed.
00:00:34
Speaker
The company was formed in 2017 and has its technology in place in some 36 sites across the UK and North America, including with Cranswick, one of the UK's leading poultry producers. Max, welcome to the Future of Poultry podcast. Hi, Mark. Great to be here.
00:00:50
Speaker
Thank you for joining us. Max, tell us first about Black Bear, what the company does and how you came to set it up. Black Bear, we put tiny little sensors into packs of food and they tell you how fresh it is.
00:01:02
Speaker
And they do that by measuring gases that are released throughout the spoilage of food. So the gases that we measure, and this is a technology we invented, they measure a wide range of VOCs that are released as food degrades, let's say as bacteria are digesting proteins and releasing gases.
00:01:22
Speaker
And the gases that we measure correlate with two of the industry gold standards for determining quality. One of which is how many bacteria are on there. So like a TVC or an APC measurement.
00:01:34
Speaker
And the gases also correlate really well, of course, with how the food smells, like what's the odor. And then we're able to give that information, bacteria counts, odor scores in real time to customers, and they get it without having to open up the pack again. So it's a new source of data and it's non-destructive.
00:01:52
Speaker
And then you asked me how how we set it up. Yeah. so well the backstory is i I'm sort of ah a scientist and a technology founder by training. I did a physics degree. I founded a software company.
00:02:04
Speaker
And I did a PhD in bioengineering at Imperial College in London. And it was there that we invented this technology that's really good at measuring inside a food package.
The Origin and Inspiration behind BlackBear
00:02:16
Speaker
We invented a lot of other things as well. But this was the one that really stuck. Once we published this, research and this sort of technology, it really got traction in the food industry. you We had big companies chasing after us to sort of test it out. And that's how we knew this was the this was something we needed to devote our devote our careers to. and And so I've spent the last six years going deeper and deeper down the the food industry rabbit hole and applying novel sensing technology into it.
00:02:43
Speaker
Okay, and so where did the name black bear come from? So the only gas sensor that you use every day is your nose. And the animal with the best nose is the black bear.
00:02:57
Speaker
Interesting. Max, tell me food spoilage and downgrades. How big a problem
Addressing Food Waste and Expiry Dates
00:03:03
Speaker
are they? And um are we actually wasting a lot of food that we don't need to because we don't really understand whether it's spoiled or not? Yeah, so firstly, let's talk about shelf life of food and and how we how we handle it at the moment then. So everyone's familiar with, if you say take a pack of chicken in ah in a grocery store, everyone's familiar with seeing ah an expiry date on that.
00:03:25
Speaker
And in the industry, they know that that date has been chosen to cover the worst case scenario. Whatever is going to happen in the supply chain, whatever was the initial level of bacterial load on that piece of chicken, say, that expiry date has been chosen to safely encompass all of that variation.
00:03:45
Speaker
And that means it's great for ensuring food quality and food safety. If you went and bought 100 packs of chicken today, you'd very likely all be completely fine. But what it also means is at that expiry date, when many people at home are making a decision about whether to waste it, at that expiry date, most of that food is still good to eat or still good to cook because the date was chosen to cover a worst case scenario.
00:04:07
Speaker
And so there's this missed opportunity by having a one size fits all static date just applied to everything, spread over everything. And I think another thing that's familiar to everyone in the street, or the ordinary ordinary folk, is when you go to a store towards the end of the day, you start seeing discounted products.
Challenges in Poultry Factory Operations
00:04:28
Speaker
You might even, if you stay long enough, see the products that get pulled off the shelves and and then they get wasted because they can't they're not allowed to sell product that's over the expiry date, especially in your meat, fish and poultry and ah products that can make you sick.
00:04:42
Speaker
And so you know you can start to see on a store level, the problems of of shelf life, and how they start to lead to downgrade or price reduction decisions and and also waste.
00:04:52
Speaker
What people do not see what the ordinary man on the street doesn't see is what happens inside a factory, inside say a poultry factory. Now, you've got this, I'd say it's a sort of a tension where you've got a pretty much a fixed daily kill. You've got a similar number of birds coming into this factory every day.
00:05:13
Speaker
they've They've been grown on farms. It's taken time to get there. And there's ah there's a fairly fixed number coming in. And at the same time, there's the amount going out, which is variable, which is based on customer orders.
00:05:27
Speaker
and they'll be forecast from their customers, big grocery store or restaurant chain is forecasting how much chicken they'll need. But the forecasts change. And ultimately, they will make orders typically to ship the next day. So you know Tesco or Walmart will come to a chicken manufacturer be like, look we want this many chickens tomorrow.
00:05:47
Speaker
And it may agree with their forecast, and it may not. And that's a bit like the supermarket or the grocery store doesn't know how many consumers are going to come into their store every day. They have variable customer orders as well.
00:05:58
Speaker
And so you've got in the chicken factory, you've got a fixed number of birds being killed every day, and you've got variable orders. And they have to circle that square. And they have to do it while hitting almost 100% of their orders, because those will be, they will be held to account. That will be their service level agreement. They have to hit in 99, 100% of their orders.
00:06:16
Speaker
So they've got variable orders, fixed input. And then they've also in the middle of all of that, they've got shelf life. They've got a limited time in which they can sell that food.
00:06:27
Speaker
And so something's got to give because if you have slower orders than expected, suddenly you don't have, you can't sell that product in the way you thought you could for the highest value. And that is where you start getting problems, which can lead to waste and downgrades.
00:06:40
Speaker
If they can't sell what they thought they could sell to the retailer at the price they thought they could sell it, then a you know a poultry company then has to start thinking, right, well, what can we do with it instead? Can we sell it on the spot market? Can we turn it into another product?
00:06:52
Speaker
Is that product at the same value or something lower? Do we have to freeze it, which is extremely costly or or incurs costs and and results in a low value product? So you have this this sort of storm of fixed input, variable output, short life that creates the need to downgrade and and to waste.
00:07:09
Speaker
um Because what they can't afford to do is to run out of stock because they either they don't want to let down the consumer who enters the store or they don't want to let down their retail and restaurant customers who are holding them to account to have high
Innovating Stock Management with Sensor Data
00:07:20
Speaker
fulfillment rates.
00:07:20
Speaker
I think to a degree you've you've touched on this, but perhaps you can expand a little. ah First in first out is standard practice for most perishable goods. But could you tell us more about what for poultry companies, what what are the costs of relying on on on that approach, traditional stock control approaches? Two big buckets I'm going to highlight to answer that question. One of them I grazed on already. you know There's a cost of downgrading. There's a cost of freezing.
00:07:47
Speaker
Maybe 20%, 30% of the product you know has to go through a process like that where it wasn't sold for the value that they thought initially or that it wasn't sold for the maximum possible value, which is typically when it gets sold to retail. um So there's a cost there.
00:08:01
Speaker
But there's another much bigger cost. Because you have this limited time window based on fixed one-size-fits-all shelf life, So what that looks like is a pack in the store has a use-by date on it, but also ah a big palletized container in a factory.
00:08:17
Speaker
could be 500 kilograms, which they might call a combo or a dolaf, depending on which part of the world you're in. That unit also has a fixed life on it, a shelf life on it, based on the kill date of the product or the poultry that's inside it.
00:08:30
Speaker
And because those dates are fixed, they are also chosen to cover a worst-case scenario. And the amount of shelf life you have severely limits the amount of stock you can hold.
00:08:41
Speaker
And the amount of stock you can hold limits the amount of sales you can have. And so what happens is short shelf life, one size fits all risk of a shelf life is suppressing the amount of stock that can be held and therefore the amount of sales that they can achieve. Turns out that more shelf life doesn't just reduce waste, also increases sales.
00:08:56
Speaker
And you know if you've worked in sales, you'll understand this because if your product lasts a bit longer, you'll find that there's more stores that can buy your product because they'll be able to sell through a full case of it before it runs out shelf life.
00:09:07
Speaker
And so the costs of relying on traditional first in first out approaches are the cost of the downgrades. You know, I would typically see about maybe a million dollars per manufacturing site lost to downgrades a year, something like that. And then you've got the missed opportunity of sales that you could have otherwise been getting, which would be double, triple, quadruple that amount, something like that.
00:09:30
Speaker
So, Max, tell us about your way of making better allocation decisions and ensuring that and there are no disappointed customers. Tell us about your dynamic control system.
00:09:41
Speaker
Yeah, so BlackBag technology has three bits to it. One, we mentioned already, it's a little sensor. And a sensor can be applied manually in low volumes, just placed into a pack of food or into a pallet. Or it can be clipped on or embedded into packaging materials at high volumes, so it's on every single unit.
00:09:57
Speaker
So there's the sensor. The sensor measures gases, measures temperature. And it sends out data wirelessly every few seconds. Then there's this sort of middleman, which we call a gateway, that collects all the data wirelessly from the sensors and sends it on to a cloud dashboard.
00:10:12
Speaker
um It can do that through Bluetooth and then through cellular and Wi-Fi networks, for example. And then the third piece is the cloud dashboard, the software, the data, that whole layer. And that is used to make decisions on how stock can be allocated. And and so typically, you would have, you know in the traditional world, you'd have
00:10:35
Speaker
You list it by when was it first in, which one needs to go out first based on FIFO. And we enable you to simply add a layer to that where instead of only having knowledge about FIFO, which one arrived first, you can see which one is actually spoiling fastest and needs to be shipped first.
00:10:51
Speaker
And so we provide an extra layer of data to stock allocation. And we can do that either in our own standalone dashboard or we can just provide an API into the current system that you already use. And how would a company actually integrate your system and use that dashboard? Walk walk us through a day in the life of someone using your system.
00:11:07
Speaker
Typically, you've got, you're managing the stock that you're holding, whether it's on a a shelf in a store or in a distribution center, depot, or inside a factory. and you've got a window at which you have to move that product on. Maybe it's 10 days, say, for example. You've got 10 days to sell it. Then typically you're just all the things that arrived 10 days ago, you're prioritizing them.
00:11:27
Speaker
And then anything that arrived 11 days ago is beyond the shelf life, and then that's wasted. And if it's at sort of nine days, you might be thinking, wow, we have to make a difficult decision here. We might have to freeze it. We might have to downgrade it, which would affect the profitability of that product.
00:11:40
Speaker
So that's not what you want to do. um But these are the these that this is the framework that is used to make decisions right now. But what we enable for those people in a typical day is that instead of just purely looking at what came in first and therefore what do we have to do to that product, they can see real-time life signals picked up by the the true nature of that product.
00:12:00
Speaker
And that means instead of just applying the same decision to everything based on how long it's been there, they can say, oh, you know what, that palette there Well, that pack there, that needs to go first. We need to prioritize that. We need to get that to customers sooner.
00:12:13
Speaker
Or even we need to pull it because it might be might cause a customer complaint. But hopefully, what' usually we'd spot it long before that. And so instead of relying on FIFO, you have real signals. And that means you can spot the outlier packs or the outlier packages that need to be to be addressed.
00:12:29
Speaker
And then for the remainder of them, they have longer time, typically longer time to sell. So you effectively extend shelf life, which reduces the pressure to downgrade, to freeze, to sell on the spot market, et cetera.
Collaboration with Cranswick and Real-World Applications
00:12:38
Speaker
Now, we mentioned that you've been working with Cranswick, which is one of the UK's leading poultry producers and and they' the UK's largest pork producer.
00:12:48
Speaker
And i saw Clive Stevens, head of research with the company, has said, with Black Bear, we gather around a thousand times more data per product, providing us with unique insight and quality and minimising product waste.
00:13:04
Speaker
Tell us about how that relationship came about and um what it actually entails. Yeah, so Cranswick is ah is a really big figure in the in the UK meat and poultry industry. And we have undertaken a ah big program with them to roll out quality testing across 17 of their sites using Black Bear sensors, using our AI-powered software.
00:13:27
Speaker
And what Clive's talking about in that quote, he mentions the amount of data. Typically, if you want to know really what's going on inside your, let's say, pack of chicken fillets, if you want to test that, you have to open it.
00:13:41
Speaker
You have to do a destructive test. You have to send that product often to a third-party lab where they'll look at it in detail. They'll culture the bacteria on it. They'll count them under a microscope. And they may smell, taste, cook the food, that kind of thing.
00:13:54
Speaker
And for all of that, you get one data point in time. You get only the data from the moment at which the product was opened. You don't see the history, you don't see the trend, so you don't know where it's gonna be in the future.
00:14:06
Speaker
The difference with real-time sensing like BlackBare is you're measuring in real-time constantly. So you don't just get a single data point, you get ah you get a line, you get a curve, you get a gradient, which means you get thousands of times more data.
00:14:19
Speaker
It also means you get much faster data because instead of having to wait till the end of shelf life to know if your product is gonna meet shelf life, You can use the trends that you can see in real time in BlackBear to forecast when it's going to hit that shelf life long in advance. so after a few days, you'd have an idea of how it's going to be in 10 days or 20 days, which means you can make faster decisions, which means you can flag things earlier on.
00:14:41
Speaker
And so when Clive from Kranswick talks about gathering a thousand times more data, it's that. That's what it means. So you have a sensor that goes into each individual pack and relays back that spoilage information.
00:14:53
Speaker
That's right. Sensors get placed in individual packs and relay back the spoilage information. And that can be done on a small scale and just used for testing. Maybe you want to improve your products. Maybe you want to change the packaging or the gas mix or extend the shelf life with an ingredient.
00:15:10
Speaker
Or it can be done on a really large scale where you have one in every single unit where you're optimizing shelf life dynamically based on sensor information and using that to allocate stock. And how's that information relayed back?
Consumer Safety and Sensor Economics
00:15:21
Speaker
So it's all relayed wirelessly. Everything is sent wirelessly in real time. So it goes out of the sensor and then is transmitted by Bluetooth or RFID into the gateway.
00:15:34
Speaker
And then the gateway sends it on again through cellular networks, through Wi-Fi or through Ethernet into the cloud where we serve it back to our customers in a dashboard or through an API. But then from the consumer point of view, you you have this...
00:15:46
Speaker
sensor within the package is there a risk to the consumer and and also if there's no risk to the consumer what happens once the package goes for recycling or it goes for landfill so we sell b2b now and our focus is b2b meat and poultry and we're improving operations in b2b supply chains at the moment But when we started the company years ago, we the vision, and it's still vision, was that you know the waste and the inefficiency from fixed printed expiry dates you know really does affect consumers. It goes all the way.
00:16:21
Speaker
And so you know we still want to get there and affect and improve decisions and information that consumers have. And we've we've gone a long way down that path. It's just not where our commercial strategy is focusing right now. So sensors on consumer packages are embedded into the packaging.
00:16:36
Speaker
They're not just resting on the food or something like that. um Are there any risks to the consumer from the sensor? Well, they get more information, so they have but they still have to make a choice at the end of the day. And I think you'll see, and we did ah we did a survey where we asked a thousand or so people, if the food that if your chicken is at the use-by date, do you trust your nose or do you trust the date? Well, let's say it's the day after the use-by date, so it's officially one day over the the life that's been given to it.
00:17:04
Speaker
Would you trust your nose or would you trust the date? And it's about 50-50 split. you know Some people, even if it smells, looks fine, but it's past the date, will waste it. And that's a completely reasonable decision. And you know they're prioritizing their safety. And other people will trust their nose and be like, it's still fine. I'm going to cook it and eat it, which is also a reasonable decision. And they prioritize the waste side of it. So I think even when you have a sensor on a pack, there's still going to be consumers using their own minds and thinking for themselves and making a decision.
00:17:30
Speaker
um you know We're not forcing them to make any decision that goes against their their nature. So I think that the risk profile largely stays the same, you know, largely stays the same or maybe is is improved because there might be a case where we were spotting a problem with the food, yeah a problem that was around the corner because it was going to spoil, you know, in the next few days. And then they could sort of eat that food sooner with that information rather than wait till the date and then risk getting sick or something like that.
00:17:56
Speaker
and But at the end of the day, they're going to keep the power in their own hands about the decisions they make. And we're just providing information. So I think we would be, rather than adding risk, we're reducing risk for consumers. And you can use the sensor across a whole variety of food types.
00:18:09
Speaker
That's right, yeah. So chilled packaged food is a really good fit for the sensor. We're a gas sensing technology, so packages where the gases are contained make the most sense for that. It doesn't have to be a hermetic seal, you know, that is completely gas impermeable, but if it's just, you know, loose apples, open air, not a good fit for that.
00:18:27
Speaker
um But there's other solutions for that. and So we do chilled packaged food, but then our strategic focus is very much on meat, fish, and poultry, and now some ready meals as well.
00:18:39
Speaker
Max, tell me, what what would the cost be of adding these sensors and and also, of course, the technologies behind them to, to let's say, a a pack of two chicken fillets? So a good rule of thumb is the cost is going to be lower than the packaging cost by a fair bit.
00:18:53
Speaker
But the actual cost of the sensors, it it scales massively with volume. It really depends on the number of sensors that are involved. And if we're selling a handful of sensors to a customer because they want to use them for some controlled testing, and that's a very low volume application. and it's a very different sort of scale when we're talking hundreds of millions of packages that are that are needed to cover entire retail product lines.
00:19:18
Speaker
ah But in any case, the scales are less than the cost of the packaging and um far less than the waste reduction that's possible from dynamic shelf life. There are a number of other companies that are working in this field.
Technology Differentiation and Global Expansion
00:19:31
Speaker
How does your technology compare to to that or to competitor sensors? There are a number of other companies in this sort of broader space. I put them into two buckets. There's one bucket where companies have developed a visual indicator that maybe changes color or changes texture.
00:19:47
Speaker
you know Maybe it goes from green to red throughout the shelf life. but you sort of You'd see the red one and you'd be like, all right, this one is spoiling or spoiled. And then there's another bucket where they are digital and they're not purely just changing color and they're digital, but they're typically just measuring the temperature or they're just measuring even just sort of the ID, like the kind of the location maybe in the ID for the product, but they're not measuring the spoilage gas directly. We're the only ones who are digitally measuring spoilage with this technology we invented.
00:20:19
Speaker
So that's sort of the key difference. I think yeah there's there's a big graveyard of companies that have tried color changing. approaches in the food industry. and i think there is a space for it. I think no one's sold it yet. I think there is a space for it, but it's not in the analytical sort B2B world because you know you need line of sight. You need to see the color change to to use it. And the B2B world where we're in, they need numbers. They need data.
00:20:41
Speaker
They need to make decisions in real time. They want to automate things. They need to plug it into systems. So the fact that you need line of sight and and there's just a color change is is doesn't really make sense for that. It's also a fairly low dynamic range on colorimetric sensing.
00:20:56
Speaker
um so there but but But there is a fit for that later on, I think, in consumer in the consumer world. It's just this different place to where we play. And then you know there's there's a lot happening in the food industry in IoT and ah putting RFID tags and BLE tags into food and and just kind of counting your inventory and and saying where things are. There's a lot happening there.
00:21:17
Speaker
I think we're a layer on top of that. and We want to add the shelf life layer to that um on top of the IoT infrastructure that is now being rolled out across some of the the biggest retailers in the United States.
00:21:27
Speaker
So you've you've rolled out the technology successfully in the UK and and indeed in North America, and understand across 36 sites. Where next? do you have plans for further overseas expansion?
00:21:39
Speaker
Yeah, we've had to be quite clinical in the geographies we focus on because we're we're a lean machine. But we are so UK, North America has been our focus. I think the next stop would be continental Europe and then Asia.
00:21:53
Speaker
Looking ahead to further change, you currently work solely with industry, but could you see your technology relaying information directly to consumers? Absolutely, yeah. And we did a big trial with ah a leading food waste charity in the UK called WRAP, where we measured the impact of our data on consumer behavior.
00:22:13
Speaker
um We gave them, we actually gave them sort of Tupperware boxes with our sensors attached, and we measured the waste under their bins, and and we did some surveys and things like that. and so we know We measured that it had a big ah impact on waste reduction, that it had an impact on the consumers were using it to make decisions. So there's 100% a hundred percent um path that we want to follow to bring information to consumers. But at the end of the day, we are trying to challenge something that's been the same since 1987, I think, when M&S brought in the first sell-by date, it was called on food.
00:22:47
Speaker
We're trying to challenge the system of fixed dates. That is a big ask that there's a lot of infrastructure behind it and guidance and regulation and and historical data behind that. So we want to actually bring about digital expiry dates on food. That's what we want to do.
00:23:02
Speaker
And to do that, we need to we need to move an industry. So we're very much embedded in the industry, working with industry at where the source of truth is right now, which is on Monday. ah in poultry manufacturing sites where they're actually putting dates on food.
00:23:16
Speaker
um And that I think is the first pillar for us before we can share that insight with consumers, because at the end of that, needs to be trusted by industry before they'll ah'll kind of bring it out to the consumers. Before we finish up, are there any final thoughts that you'd like to leave our listeners with?
The Future of Food Supply Chains with AI
00:23:30
Speaker
I think the future of food, the future of food supply chains are ones that will feel themselves and improve themselves. yeah We talk about AI and it's so hot right now and AI by itself isn't going to solve food waste. It can't. All it is is the intelligence layer.
00:23:47
Speaker
AI needs to connect to the physical world to actually impact food. And what that means is, there's really three three parts I see. One is it needs information, right? It needs it needs to know what it's what it's making decisions about. It could be information from Black Bear or otherwise.
00:24:04
Speaker
And then the other thing it needs is actuation. It needs robotics and um tools that it can use to make impact. So I think AI is sort of the cog in the middle that makes decisions, but without... Without new data and without actuation, it won't solve food waste. So I think the future for food is really bright and it's closely linked to that of physical AI.
00:24:22
Speaker
And I suppose I'm just'm just excited for that change to happen in an industry that has just so much to so much to benefit and is handling such a precious resource at such high volumes.
00:24:32
Speaker
um that's That's the thought I want to leave. Excellent. Max, thank you. Thank you very much for joining us. Listeners, thank you very much for joining us today. And don't forget to look out for future editions of the Future of Poultry podcast. Goodbye.