TELUS Agriculture: Enabling More Efficient Production with EDA

Jeff Bradshaw: Hi, my name is Jeff Bradshaw. I'm the VP of Global Technology over at Telus Agriculture. And I'm gonna talk to you about enabling more efficient production with event driven architecture.

So firstly, happy Star Wars day.

Just a little bit about me. I am the VP of global technology, Telus agriculture. Formerly I was a CTO at Proagrica. So another good agricultural technology company. Prior to that, I was a founder and CTO, Adaptris, long-standing integration company and actually a partner of Solace. I've worked with integration enterprise integration since 1994. So not as young as I clearly appear. But yeah, that's a bit of bio for me, as you can see from the rugby paraphernalia in the background, I'm a huge rugby fan. So I spent a lot of time over, not playing unfortunate anymore.

So firstly, who's Telus, so obviously the Canadians in the audience should recognize this and be able to answer this very clear. Telus actually, it's the largest telco in Canada by revenue, on the, I think the fastest 5g network in north American market, huge focus there, for us as a business and being a telco, one of the core revenue streams obviously is around, the wide lines and the mobile.

Obviously one of the key things then was how do we diversify? So there's a separate Telus business unit, which focused on the large, kind of B2B type plays. Then lastly they decided to divest and do different things. So rather than rely on the telco revenues, look it up other streams. So they created Telus international, which is the large number of employees. They effectively look after outsource offshore development and support centers.

So one of the big skills with Telus is creating customer service environments and they decided to spin out a separate entity very successfully, but then got into the healthcare market in Canada. So, you know, again, quite regulated environment, very focused on the, patient records and all these kind of things and electronic pharmacies.

And once they got into healthcare, they then decided, well, what are the big things that can influence outcome in health? And the thesis came back that well, let's get into agriculture and health food and help production of food. So we're the newest business unit in the family, rather than rapid growth area. And we actually are an international entity as well. I'm based here in the UK. We have operations in, in, across north America, south America, all across Europe. And down in Asia and also down in Australia. So it is a global entity and also our systems have to cope with that fact as well, which is quite an interesting thing.

So why get involved in agriculture? As I mentioned before, it's an opportunity to diversify. It's also, tell us as a company, a very strong in the social capitalism area. He follow us on LinkedIn. You'll see lots of good stuff to tell us to do within the company. Both in Canada and abroad and also leveraging the experience on health. So around data, lots of knowledge around data and integration and keep us, and ultimately the better agriculture information leads to better food outcomes, which hopefully leads to better health.

So our mission as Telus Agriculture is creating the best producer to consumer outcomes, that no, that's not technology producers consumers, it's actually people like you and me, enjoying great quality produce. Firstly, on the agri side, we have a very large food beverage and CPG operation. So again, you could be working with, battery manufacturers or alcohol manufacturers- for us they're still producers and consumers. So we deliver digital insights and the apparent connects the customers all along the value chain.

And that's the key thing here is all along the value chain. Globally. So quality safety distribution of sustainability are all key parts of our message and what we deliver to our customers, whether they be food or.

So our core strategy really is connecting. So that plays down to the 5g networks and the IOT devices in the field, digitizing that data, ag as an industry is quite behind a lots of others, data is, there are pockets of excellence, but then there's lots of stuff that's not actually automated and still paper receipts flying around. Obviously optimizing what our customers are doing and help them gain insights into their business. And that's a really key one because the supply chain in agriculture is one of the most fragmented of any industry, actually. Manufacturers don't sell direct to consumer. You've got one and often two hops in the world. The people doing the selling largely are agronomists, which they're kind of plant doctors for those of you not ag people. And so the farmer generally gets a recommendation from an agronomist then goes and buys the products, but that's whether he can get it or not, I'll come onto that a little bit later on, and also exchanging that information in a more real-time way, certain processes in ag generally take place at the end of seasons, a very seasonal business. So we need to be able to get that to a more real time, more instant gratification if you will, of the data.

So what are we building right now? Our flagship product that underpins everything we do. He's my baby. It's a product called ADX and the idea being it's the line in the middle. So we work on the left-hand side with the agricultural manufacturers. So the likes of the Bayers, the BSF and Agentas of this world, we work with the ag input retailers. So you're North America nutrients is a popular one. And there are many other ag retail companies, and they're very fragmented retail environments. We work with ranches and farms, so we have quite a large cow calf operation that we support. It's like a cow hotel, if you will, for a period of time, as they're growing and developing, then we work with Aussie manufacturing and distribution. Guys that make the food products, we eat the CPG products we eat at home.

Logistics and transport, obviously it's a key linkage all way along the path is data coming from trucking. And again, I'll talk a little bit about our cold chain offering there and finally, up on the grocery food service. So this goes towards, looking at the behavior. So taking the point of sale data, taking consumer behavior, looking at replenishment of stores, optimizing what's on the shelf in particular areas and monitoring performance of service as well.

So quite a big thing. And this is the very simple slide, but actually the integration for the integration guys in the room, as you can imagine, it's very, very complicated. There's more data in ag than probably any other industry in the world. But, as I always say, it's the dirtiest form of data, it requires a bit more work to actually make it useful and make it more interesting.

So where do we use a vendor in architecture? So I mentioned ADX, which is our kind of backbone across the board. ADX surface data via APIs, and obviously behind the scenes, we're using our event-driven architecture.

Big areas where we're -so first phases of ADX, it's still in its infancy as a product. We're probably eight months in to the delivery of the product or platform, is around collection of data from retail. So point of sale, data, imagery, data, goods, receipt, notes, those sorts of things, which we then can claim information around products from and use those in analytics capable.

IOT is a huge area for us right now. We've just gone live with our cold chain products, which is why I love this one. It's one of my favorite ones, cause it's really tangible. One of the biggest areas of food waste in the world, actually food waste is one of the biggest challenges. So everyone talks in ag about upping production up in yield, focusing on what comes out of the field. But actually we lose something like 35% of all food is produced to waste. And we had a real good use case with our cold chain products. So cold chain includes these little tear off tags that go into the pallets of food and you can then track them for things like temperature, bumping and this kind of stuff. Someone's throwing the car pallets around.

And we actually worked very closely with, one of our customers down in Australia, a big retailer where, and the brief of the project was, they were losing two days of shelf life on strawberries. So quite valued, fruits, and they're trucking them down from Queensland down to the distribution center. And they were wondering what was going on. Why were they losing this two days of shelf life on the produce? So we got involved with our cold chain product, the first release. Got the tags in farm level, as they said, they're loading the berries into the trucks on the farm, trucker drives down to the distribution center and they were unloaded. And now generally again, for those of you that haven't been in this environment, you'll see a guy when they open the truck at the distributions, that's a monitoring temperature at the back of the truck to make sure it's within tolerance.

One of the things we've learned from the current crisis in, in Ukraine is fuel is very expensive. So pretty much the only variable costs that these third party truck drivers have is the fuel. And it's quite a long drive across the state of Queensland in Australia, for those of you who've never been. So they pick the produce up. What we found actually in the first week of the proof of concept phase was they pick the products up, the first thing they do is put the truck into eco mode, then drive 12 hours down to distribution center. An hour outside of this distribution magically the truck goes really cold because they realize they're about to deliver. And when they get to the destination, they're going to be temperature checks. So they turn the engine back, they turn the freezer right the way down, and then the result of this was, again, these refrigerated trucks, the strawberries at one end of the truck were basically frozen at the back of the truck. So at the doors, they were just barely coming to the right temperature.

So yes, they passed a quality check. When the guy comes in and set the temperature. But actually they've damaged the goods by doing that and treating them that way. And this is something we couldn't find out with general standards sensors. We couldn't find these things out until we actually started collecting the data.

Now the tags are emitting a lot of data and we're capturing in real time. As the produce is flowing down the road, we're actually getting all this data. And so again great amount of data, very rich data to play with very rich sets of events. But again, Using our event-driven architecture, we can actually ping those where we see a truck going off-temperature and actually these sorts of events that we can push back to the driver and say, Hey, the trailer's not at the right temperature. What are you doing? So we can actually start to capture those sorts of things that offer real time insights.

So even after the first couple of weeks of this project, the customer actually saved those two days. So it's a really, you know, they saved two days of shelf life of strawberries. Massive waste reduction, hugely successful project. And we're rolling out across the world. Now at the moment, obviously the limitation there being tags. They're pretty cool. I don't have any with me, unfortunately, but they're tiny little tags and you literally just tear it and put it into the pallet with the goods and that can track different things. So again, lots of data and for IOT, for us, it's a huge area.

We do a lot of field data. via our legacy AGI business. So pulling back, what's going on the field, in the farm, there's a lot of talk in the industry about variable rates. So if you imagine, historically you just blanket spray or blanket, fertilize field, fertilize it again. Another issue with the energy cost, fertilizer is incredibly expensive. So what they liked to try and do now is do more variable rates only apply the product where they need it across the field. And we're working with a number of customers where we are the kind of single voice of truth. So did the variable rate give a positive yield? Did it work? So capturing all of that data across a single field, just for context, the tracks are generally in season generates about three hundred and fifty three hundred sixty megabytes of data per day when it's working. So if you multiply the amount of machines and the amount of trailers and the other bits involved are talking about quite large volumes of data. Okay.

The other thing I'll tell you is a little stat. The world's most popular crop is a crop called crop one, because that's generally what comes off when the farmer doesn't bother saying that the crop type in here in the screen in the John Deere tractor rolling. So yeah, a little bit of trivia for you there.

We also obviously weather data, if you look at things that influence what's going on with the field. Whether it's one of the big ones and then we'll also use it for billing so we can see who's using what data, obviously the events are floating around, so we can just consume those events and generate billing off the back of that.

So these are the kind of really early, early projects that we've got in production now. And we've got a whole raft more coming down the line. So it's accelerating, rather rapidly as people are seeing the benefit, because really we're trying to move this seasonal business into a much more real time, business.

So he's a, yeah. Give you a bit of example about some of the things we've connected. The other thing I didn't mention earlier, which I probably should have done is Telus Agriculture actually grew up via a number of acquisitions. So we've got about, I think, 12 different acquisitions that we've made over the last couple of years.

So again, part of the challenge of stitching all those silos of data together, which is again a primary role of our ADX platform. And so looking these things up, whether it's for management data. So FMS is for management software, which we have, I think, four different types within our portfolio today, but around the world, varying qualities of data in there where the partners, we pull that in from partners, we don't have our own, we have weather stations from partners. We don't have our own weather service. So we use third party data for that one, obviously imagery from various sources. You can see NASA there, the European space agency and planet, and then also the equipment manufacturers. And you'll probably recognize those, it's a bit of a who's who of machinery providers in the ag sector. And then for those of you in a more traditional business, you'll recognize some of the, ERP packages and things that we work with. There are some very specific farm ones. But also we work with the big ones as well, because a lot of our customers run those big entities.

So very traditional kind of integration platform here. But again, using that and the EDA backbone and the pubsub model was to share the data around the business and store it and curate it. So main reasons we use, the event-driven architecture in our use cases is really about the customer and empowering them to make that real time decision, the industry generally is a lot slower, a lot more legacy, farm will make about 14 decisions during the season, but it doesn't know the result until you harvest at the end.

So always remember those there's, there's no instant gratification in farming unless you're doing animal, but that's slightly [...] Get sick. So yeah, optimizing rebates is a huge area for us. The other thing is we're stock, you know? So if you think about seeding, for example, it's one of those interesting ones where you don't know what seed you've got to sell before it's been brought in, cleaned , bagged and sent back out again. And you've got a very small window of the year when farmers actually go and buy the seed. Cause being on the growing season, whether it's going to be right to plant, you've got to drill the right thing. When moisture is got to be good. There's lots of conditions about when they grow it.

I urge everybody to watch Clark's farm. Cause it actually, it's quite funny, but it gives a really good insight into life as a farmer and what they get involved in and the decisions they have to make. Also things like, so we're stocking the spotlights of where my seeds. Oh, I need this crop protection, where can I get it from? It's always a big thing. And also things like yield. So obviously with taking that machinery data as they're harvesting it, so pulling back things like moisture content, protein as it's being processed, all these good things and feeding that data back to the farmers to again, help them make better decisions, better plans for their crops.

For the following years, that's all really important. And again, one of the biggest uses actually for me, and anyone who's worked in IOT will know this, there are a lot of events when you're dealing with IOT. It's the, the, just the sensor example I gave earlier running on cold chain, every few seconds, these things are telling me what they're doing, you know, they're very needy. They're always selling you. So actually, the big thing really is filtering out the wheat from the chaff, right? What do I care? What events do I actually care about and be able to filter and subscribe based on those things. So really key use cases actually, and we have many more that we're rolling out.

For us, the ADX platform is our flagship. It's the key platform to Telus Agriculture. And it's the thing we deliver our services on them. All the next gen platforms are all based around this. So it's actually key to what we do.

So thanks for listening. Hopefully that's given you some insights into some of the things we're up to within Telus Agriculture, and also within wider Telus, certainly you want to find out more or you're interested in some of the stuff we're doing? We'll see, I've got a food beverage screenshot here. The other end of the supply, the other end of the supply chain or over to the right. Please feel free to reach out to me. jeff.Bradshaw@telus.com. Yeah, always happy to hear from people who are interested in the industry and finding out what we're up to.

So yeah. Get in touch. Thank you.