Andrei Danescu: 0:03

You have a daily snapshot, you want to go back in time at any point. You see exactly what the state of the operation was. So when you, when you look at optimizing for the future, if there was something that was seasonal, and you only run it for, I don't know, let's call it the winter season, and you can go back in that period and see exactly how did my warehouse look like? And where can I drive efficiencies for the next winter season in the next year?

Tom Raftery: 0:27

Good morning, good afternoon, or good evening, wherever you are in the world. This is the Sustainable Supply Chain Podcast, the number one podcast focusing on sustainability and supply chains, and I'm your host, Tom Raftery. Hi everyone. And welcome to episode 25 of the Sustainable Supply Chain podcast. My name is Tom Raftery, and I'm excited to be here with you today sharing the latest insights and trends in supply chain sustainability. Before we kick off today's show. I want to take a moment to express my gratitude to all of this podcast's amazing supporters. Your support has been instrumental in keeping the podcast going and I'm really grateful for each and every one of you. If you're not already a supporter. I'd like to encourage you to consider joining our community of like-minded individuals who are passionate about sustainability and supply chains. Supporting the podcast is easy and affordable with options starting as low as just three euros or dollars a month. That's less than the cost of a cup of coffee and your support will make a huge difference in keeping the show going strong. To become a supporter, simply click on the support link in the show notes of this, or any episode or visit tiny url.com/s s c pod. In today's episode, I'm going to be talking to Andrei from Dexory and we'll be talking about robots. And in upcoming episodes, I've got some great ones lined up. I'll be talking to Elizabeth from A E Global. We'll be talking about packaging, sustainable packaging and Lenny Marano from Lectra. Dario from Propel software. And John Sicard from Kinaxis. So some excellent episodes on the way. If you're not already following the podcast, be sure to click follow so you don't miss those episodes. But now, as I mentioned, I'm talking to Andrei. Andrei welcome to the podcast. Would you like to introduce yourself?

Andrei Danescu: 2:19

Absolutely. It's great to be here. Thanks a lot for having me, Tom. So I'm Andrei. I'm Andrei Danescu, one of the founders of Dexory and I am the CEO. My background is in this technology space, robotics, autonomous systems. So, that's how Dexory came to be from a passion of using autonomous robots to collect real time data and using this data, using this information to really drive business and industry transformation for the logistics industry.

Tom Raftery: 2:48

Okay. And tell me a little bit more about that. How did you get into robotics and real time data?

Andrei Danescu: 2:57

So that's a that's an interesting question. I'll try to be brief because it kind of goes back to, it pretty much goes back to my childhood. I've I've always been very passionate about technology. I've always been fascinated about robots. I was I was obviously a geek growing up very, very much into tech and throughout my Basically, throughout my life, throughout my career, I've always been excited about creating new products, creating new technology. So, as, as I, as I grew up and I started going through my studies, I I've focused and specialized in robotic systems, electronics control systems. Then as I went through my career, before Dexory, I did motorsport. So in motorsport, I did data engineering. I did control systems. I was I used to work in in Formula 1 and that's really triggered my, my imagination of what you could do if you have this ability of collecting real time data from any space, and you have the ability to transform this into insights. So really, can it drive business value? And what kind of what created Dexory was this intersection of autonomous robotics, as well as the ability to use information. And feed it feed it forward, turn it into insights and drive value for businesses. So that that's the combination that really got me into robotics. And then that's the combination that got robotics into Dexory and into the product that we have today.

Tom Raftery: 4:27

Okay. So tell us a little bit about the product that we have today. So far, I've figured out that it's a robot that does something with real time data. Now, I obviously know a lot more because I went through the website, we had the prep call, et cetera, et cetera, but for anyone else listening, that's where they are right now.

Andrei Danescu: 4:42

So that's that's. In a nutshell, what we have, we have some robots and we have some real time information and then where things get really interesting is when we start looking at the product that we offer as a company. So this product is very much around digital twins. Now, it's a digital twin platform for logistics specialized specialty is warehousing and warehousing operations. What this means is the way we describe it, we describe it as a global visibility platform for the logistics industry. So it gives any warehouse operators or multi warehouse operators, the ability to really control. Every element of that operation to really have visibility on how can they run the business more efficiently? How can they maximize the operational efficiency, the sustainability, the quality of every single location. What's really unique about it is it's powered by the world's tallest autonomous robots. So, our technology consists of this this very tall robots that collect the data, collect the information from warehouses. They then transform it into insights and push that informational insights into Dexoryview So Dexoryview is the product that we're talking about here. This is the digital twin platform, the global visibility platform. And this is the single source of truth for warehouses where they can go in, access any area of the warehouse, any product, search for anything from any location in the world.

Tom Raftery: 6:10

Okay. So just to break it down, you've got robots that are tall. And the reason the tallness is important is because in warehouses, warehouses themselves, the pallets can be quite, quite tall as well. So your tall robots can see up to the top of those pallets, can see what's there. And they're using computer vision to make a map of the warehouse, which goes into the digital twin. And therefore people can see what's in any part of any warehouse at any time. Is that more or less it?

Andrei Danescu: 6:50

I think you've described it very well. So, look, look, look, looking at it in more detail. Basically, we, we were very excited about the opportunity of helping the industry become, become more efficient, helping the industry become more sustainable. And a lot of this actually came from the, the challenges that we faced during the, during the pandemic. And I don't know if you, if you remember, but there

Tom Raftery: 7:13

I remember.

Andrei Danescu: 7:15

massive issues with electronic parts with pretty much everything. And you couldn't really couldn't really predict any of this because you had some components today and then tomorrow, other components were missing or other things were short or in short supply. So that's kind of how we had the idea. We had the idea of what's the way of bringing or closing that visibility gap? We, this is what we call it. We call it a visibility gap. So when we looked at what a warehouse consists of, it has these very tall racking systems. And exactly like you said, we have this very tall robots, which are the same height as the racking system, which allows us to collect an incredible amount of information. So we can scan over 10, 000 pallets in an hour, which means that on a daily basis, we can scan the entire warehouse. We can scan the entire space and really bring that power of real time data insights to customers.

Tom Raftery: 8:10

Okay. I'm assuming that the alternative to this is someone going along with a clipboard and scanning the shelves and making notes, or maybe an iPad, scanning the shelf and making notes. Or I know more recently I've talked to companies that are doing things like they're running drones inside warehouses, up and down shelving systems, is, is that kind of it in terms of state of the art right now? Or is there something else I'm missing?

Andrei Danescu: 8:42

I think that's that's a good description. That's, that's, that's pretty much it. I mean, Realistically, what you described first, I think that's pretty much the the status quo in, in, in the way people are doing this. I mean, you have teams that are going around warehouses with clipboards. And that is a very slow and very inefficient process. Some people have tried drones. There's, there's various there's various issues with the speed of scanning and the ability to actually return the same amount of data, the same amount of information. And that's 1 of the reasons why we decided to go with a ground based robot that allows us to really scan everything very quickly. Because we spoke to a lot of we spoke to a lot of players in the industry and spoke to a lot of people in the market, which have a lot of experience in running warehouses. And we said, what would you need in order to get that a hundred percent efficiency to, to access that next level of sustainable, efficient operations. And a lot of the time, the answer is really we, we don't have an ability to collect information on a continuous basis. We don't have an ability to to become a data driven industry. So, therefore, that's the opportunity that we saw. We said, okay, how can we create something that gives. In the large industry players and small industry players, pretty much anyone that operates this, this location, this environment, how can we create something that gives them the unique advantage that real time data gives you in in other industries and in other spaces in the world. So, when you look at the people running around and collecting this information with clipboards, like you're saying, the problem there is that it's very slow and also, it's very hard to get the data richness that we can, we can provide. So, when you look at the digital twin, you have a complete 3d reconstruction of the environment. You have computer vision, like you mentioned, which allows you to extract a lot of information from pictures. You have pictures in every location, which gives you a continuous view of the space. So, if you want to go back in time. A month ago, or 2 months ago, you have an exact snapshot of how your location, how your warehouse was operating. And that's very, very hard to do with other technologies, because on one hand, you need a lot of people. You need a lot of of or a large team doing this, but on the other hand, you can't afford to keep closing operations to be able to scan using other type of technologies that are there may be out there.

Tom Raftery: 11:06

The other one, of course, I forgot to mention was the kind of smart shelving units, but those are horrifically expensive, I guess.

Andrei Danescu: 11:13

That's a, that's a very good point. Cause I think when you look at the 100,000 pallet warehouse, even a 10,000 pallet warehouse, 15,000 pallet warehouse there's challenges fitting these cameras and then there's challenges consuming really all the information, all the data because it doesn't just magically go from 1 camera to the server or to the location where you need to use it. So I think that's that's not necessarily horrifically expensive. It is extremely expensive when you work out the practicalities of deploying such a system, but it's also very impractical because if anything breaks. If you have a couple of cameras that get knocked by a pallet or by a forklift or whatever might happen, and then your data becomes incomplete and you start losing that, that richness and that continuous continuous value they can bring.

Tom Raftery: 12:01

Okay. And your autonomous robots, how often would they scan the shelves? Is it once a day, five times a day, a hundred times a day? Is it entirely up to the warehouse owner? Where does that lie there?

Andrei Danescu: 12:17

So I would say we scan at least once a day. The current the current offering and the current state of technology allows us to scan a million square feet or over a hundred thousand pallets on a daily basis. So if you have warehouses that are smaller than that it means you can scan them multiple times a day. This is done with 1 single robot in 1 single location.

Tom Raftery: 13:12

Okay.

Andrei Danescu: 13:13

And of course, if you have warehouses that are bigger, you can then use or deploy multiple robots to scan even even faster and even more often. Usually, what we find is that working very closely with additional warehouse systems allows us to know where there's more activity in this, in this environment, and then we can actually prioritize those areas for more often scanning if it's needed.

Tom Raftery: 13:36

Okay. And seeing as you bring it up. What is the connectivity between DextreView and other warehouse management systems like, do they talk to each other very straightforward. Is it through an API or do you need to do some kind of integration or, you know, where, where are you there?

Andrei Danescu: 13:56

So it's, it's relatively straightforward. I would say. It, we have a very simple and very easy way of integrating with other systems. You can do data exchange in various in various ways. You can work on an API basis. So if there are existing APIs that we want to be either integrating in integrating with pulling data from pushing data into or if we want to go into a deeper integration, we also have the, the opportunity to do that. So Dexoryview is a a system that we've designed as a, as a complete product. So, obviously, it offers multiple ways of ingesting data, exporting data, but also integrating with other systems

Tom Raftery: 14:34

And it's cloud delivered or is it on prem?

Andrei Danescu: 14:37

We have options for both. Primarily, it's cloud cloud delivered as it's a lot simpler to maintain and it avoids any kind of complication for for the customer. That's that's probably one of the key points for us. We want to make this technology. If you provide the technology that is autonomous. It needs to be really autonomous. The customer doesn't have to do doesn't have to do anything about it. So the easier it is to deploy and the faster it is to deploy, the better it is for the customers. And that's what we're providing.

Tom Raftery: 15:08

Okay. And when you say Dexoryview is the product, so the robots are, and you said one typically per warehouse, do customers buy the robots? Do they lease the robots? Do they get the robots free with a subscription to Dexoryview? You know, where, where, what's the business model there?

Andrei Danescu: 15:32

So, the robots are a part of of the product that we offer. And the reason why I say Dexoryview is the product, because you cannot separate the two. You have a holistic package that allows you to get this real time visibility on operations. So there's a subscription model that is very similar to the vast majority of software subscription models, and this basically consists of a bundle that includes the autonomous scanning robots in order to supply the, obviously, the real time data on a daily basis.

Tom Raftery: 16:06

Okay, cool. Good. And do you have any success stories, any customer success stories you can talk to?

Andrei Danescu: 16:13

Yeah, I mean, we have a bunch of case studies, of course, on our, on our website and online that we can, we can talk about. But also there's there's a fair number of large corporate customers that we're working with. We also have some, some smaller customers that have seen a very rapid ROI, a very, very quick return on the investment on the technology. And also more importantly, they have seen a very big impact at, at the business level. And that's 1 of the most exciting things for us is bringing a product into the market and seeing the value it brings to our customers.

Tom Raftery: 16:49

Okay, and in terms of sustainability, do you have any metrics you can speak to? You know, X customer deployed Dexoryview and as a result their warehouses are now Y percent more efficient or anything like that?

Andrei Danescu: 17:02

Yeah, I mean, I wouldn't look at necessarily generalizing, but what we're seeing with a fairly large sample of our customer base is that in the ideal world, you want to operate at 100 percent efficiency, right? There's never going to be 100%. You might be at 99. 999 or 999. 597. But what we're seeing is that customers usually start in the region of 90, 85, 92, 93, even some of the better ones, you know, 90 to 93 percent. When you're looking at the massive warehouse was with tens of thousands of pallet locations and a very large footprint, you know, 5 percent inefficiency does really end up costing a fair amount of money. So what we've seen is that deploying our technology and more importantly, also working to fix these issues, because of course, the technology gives you the actionable insights, but it does rely on people taking the actions to fix it. So it's a huge enhancement for the existing workforce. We've managed to take warehouses that are over 80,000 80, 000 pallet locations to less than 10, 15 errors on a daily basis. So we've pretty much taken them to the 100 percent efficiency, especially if you round it. But we've, we've literally eliminated from, you know, a thousand errors down to 10, 15, 20 errors. So a huge improvement in the way they can run the operations. And then obviously what that translates into is better use of resources because you don't have to spend so much on, for example, the lighting in the warehouse for people to find various things or lost goods. Better use of the existing infrastructure, forklifts that need to be charged and so on. And obviously, a happier workforce, which at the end of the day, does go towards the sustainability concept, because we are not going to have a very robust and agile supply chain. If we don't have people that are very excited to work in that space as well.

Tom Raftery: 19:13

So that the workers don't enjoy running around the warehouse looking for lost parts.

Andrei Danescu: 19:20

I mean, I can't remember last time I enjoyed looking trying to find something I've lost. It's not exactly a past time activity, right? So, no, they don't really like that. And, and also the other bit that comes on top of it is when you have to when you have to ensure that certain cargo or certain goods make the next mode of transport, the pressure becomes really, really high. And if you think about it, when you, especially when you look at the wider impact on sustainability if you're operating in a a cargo warehouse, and the cargo that's supposed to go on a certain plane misses that plane, then you have to go back to the queue. You end up having this snowball effect that impacts the entire operations across multiple, multiple different flights. If you send the wrong thing to the wrong customer, and then you have to rebook it, you have to put it back on another plane, bring him back to source. So there's a lot of there's a lot of let's call them hidden efficiencies because they actually stem out of the warehouse operations. But they have a much wider ripple effect across the entire supply chain.

Tom Raftery: 20:26

Okay. And I mean, it seems on the face of it to be a very simple solution. You've got a robot going along the warehouse, quite tall with covered cameras, taking pictures the whole time, sending all that data into your Dexoryview product, so that everyone knows where everything is in the warehouse at all times. What's next? Because, okay, that's, that's that issue solved. Are you looking at other problems? Are you looking at ways to improve the current solution? Or, you know, what's, what's next on your, your playbook?

Andrei Danescu: 21:11

So we're we're continuously looking at becoming better, right? We're always looking at being able to scan faster, being able to do more with the existing technology, more in the same space. I think it, it goes back to the question that you asked me previously is how many times a day can you scan? At the moment if we can scan a hundred thousand pallets in, in a day, if we scan twice as much, then obviously not necessarily that we would need to, but we can scan a lot faster, which means we can bring a much better detailed picture of what's going on with the operation. We're, of course, always looking to go bigger and higher. So we are continuously exploring new ways of accessing new heights for our technology. To put it in a more poetic way.

Tom Raftery: 21:55

Pun intended,

Andrei Danescu: 21:56

And yes, exactly. And I think one of the big areas of development for us is being able to bring more and newer technologies into into the industry. So we are doing a lot of work with artificial intelligence to be able to drive further optimizations on how the operations are being run. To be able to extract the last pretty much the last 0. 01 percent of value from the existing warehousing system things, the existing operations, because when you look at the warehouse again, we go a little bit more in the sustainable sustainable direction. If you have to heat up the space, you have to use electricity to light up the space and so on. And you have to do this operations regardless. You could run the whole warehouse at 1% efficiency, 10%, 50%, or 100% efficiency. You end up having to, to use these resources. And then if you have an inefficient utilization of that environment of that space, then obviously it's not very good for anyone and nobody wants that. So using artificial intelligence to drive further optimizations to store your goods better to be able to put more pallets, to put more goods through the same space, through the same warehouse. That's where we see a huge opportunity. And we also see a huge opportunity for enhancing further the existing workforce, being able to work hand in hand with people to make their jobs easier, to make their day to day a lot more pleasant.

Tom Raftery: 23:30

Okay, okay, cool. You are going to be, or your robots are going to be scanning lots and lots and lots of pallets in lots and lots and lots of warehouses. That's a huge amount of data you'll be scanning and storing. Are you going to be looking across that data for any trends that you can then, issue reports on, or say to people, other warehouses in your industry do this kind of thing, you know, any, look for best practices that you can share with your customers, any of those kinds of things.

Andrei Danescu: 24:10

So that's definitely an interesting idea. I think we should we should absolutely explore that further. But but this is exactly what I was telling you about the artificial intelligence systems we have at the moment is being able to analyze the trends inside the warehouse. And being able to proactively come up with solutions and ways of making that operation more efficient. So we haven't really expanded today at the cross industry level or cross market level. I think for us, the big, big focus is to make sure that we can give everyone the ability to run their operation at a hundred percent efficiency. What's really interesting is whilst we have this incredible ability to collect the information, it's also very, very important how do you process the information into insights? Because just to give you an example, I think we can collect anything between three, 400 gigs an hour every hour. So a lot of our technology revolves around processing that data in real time and not storing things that are not useful for for anyone really. Because going back to the connectivity question you asked me, if you're trying to push 400 gigs an hour, 500 gigs an hour, every hour, then you start having this massive gap between the point of collecting the information and the point of actually returning the insights. So there are technologies capable of processing all that in real time. Pretty much the robots are scanning in the warehouse and you see the results. You see the information and you can act on it straight away. So that gives us the ability to analyze the trends in a lot more detail at a much, much higher speed with a much better resolution and other things would would enable us to do.

Tom Raftery: 25:53

Okay. And in terms of just. Working with the people who are working in the warehouse at the same time as well, it's an autonomous robot, does it scan around looking for people to make sure it's not going to bump into anyone? Does it have Uh, Warning sounds, you know, how does it, how does it make sure that the people in that environment are safe?

Andrei Danescu: 26:17

So it's designed as a collaborative robot. It's designed as a collaborative technology. That's very important point, because especially as you're scanning these big spaces, you need to be able to operate with 0 impact on the day to day operations with 0 impact on the existing process, the existing flow. We have a large number of safety systems built into into the product. So it will be able to avoid obstacles. You would be able to navigate dynamic environments. It has the standard warehouse warning lights, the blue lights projected on the floor, the buzzers. So the audio signals and audio warnings. Very, very important point. Because these are environments that are real, you know. This is not about creating a technology in the lab. This is about creating a technology that is robust and it's ready to be deployed anywhere in the world. So that's that's always been a front of mind for us. And from the beginning, that's exactly what we looked at doing was make a technology that can operate in any environment with any number of people around it. Forklifts, pallet trucks, anything you might imagine in a warehouse, of course.

Tom Raftery: 27:28

Okay. Fair enough. Cool. We're coming towards the end of the podcast now, Andrei. Is there any question I did not ask that you wish I did or any aspect of this we haven't touched on that you think it's important for people to think about?

Andrei Danescu: 27:44

I think we obviously covered quite a lot. One thing I would say is probably looking a little bit closer at the speed of information and the speed of data gathering, because I think that's one of the fundamental parts of why this technology is so impactful. If you look at if you look at, okay, fine, people might do a stock take once a year or once a quarter, and then you can do it every, every week or every other week. That's fine. Okay. You've created some efficiencies there, but I think the value that you really, that exponential value that you bring to the industry is really being able to collect this data on at least a daily basis, sometimes multiple times a day. That's really when you start seeing the compounding benefits. That's really when you start seeing that value that you can derive from being able to optimize your operation to that last finest level of detail. I think that's that's an interesting one we haven't explored so much around why we are actually collecting over 10,000 pallets an hour, but I can I have already mentioned why, because I think the speed at which you can collect the information, the frequency at which you collect the information is one of the most important points is it's also what's giving you the ability to have a continuous view of the operation, so you have a daily snapshot, you want to go back in time at any point. You see exactly what the state of the operation was. So when you, when you look at optimizing for the future, if there was something that was seasonal, and you only run it for, I don't know, let's call it the winter, the winter season, and you can go back in that period and see exactly how did my warehouse look like? And where can I drive efficiencies for the next winter season in the next year? I think that's one of the that's one of the key points that I am always keen to highlight as well.

Tom Raftery: 29:35

Okay. One thing I forgot to ask earlier, so I'll ask it now, when your robot is scanning the shelves in the warehouse, can it actually identify what's on the shelves? Can you train it to say this is this particular product or this particular SKU or whatever?

Andrei Danescu: 29:53

Yes, the short answer to that one is yes. So I'm using using the computer vision techniques. We can detect obviously barcodes, QR codes and other types of labels and decode them and understand them, but you can also do, analysis on the type of goods you have on this palette. So the robots scan very, very closely to the shelf. They're about 40 centimeters to half a meter away from the shelf. So there's no when the robots are scanning, there's nothing that's going to get in the way or in between the robots and the actual shelves. So the pictures that you're taking are the actual goods that are on on pallets stored inside the warehouse. From this information, you can extract quite a lot of value. You can understand how much a product is available for picking at any point in time. What are some of the unique identifiers the labels? You can look at the pictures and say, okay, this is my my box of specific, specific type of clothing or a specific brand I have in the warehouse, so it has a lot of flexibility built into it because warehouses are live environments. And you might not always have a barcode. You might not always have a unique identifier. So you use this rich data sets to extract as much value from it as you can.

Tom Raftery: 31:05

And I assume as well it's a learning system so that if you tell it once this particular shape is this particular product, it never has to be told again or it shouldn't have to be told again.

Andrei Danescu: 31:16

It's designing where we're continuously working on that to to get better and improve over time to have a deeper and deeper understanding of how things are usually displayed in a warehouse, what things are in a warehouse and how to, to best recognize them in the future.

Tom Raftery: 31:32

Okay. Super. Great. Andrei, that's been really interesting. If people would like to know more about yourself or any of the things we discussed on the podcast today, where would you have me direct them?

Andrei Danescu: 31:44

I think the, the easiest way would be to go to our website, Dexory.Com. There's a lot of information there, but also reach out to us on LinkedIn, reach out to myself on LinkedIn. I'd be always happy to have a chat and dive into more detail.

Tom Raftery: 31:58

And where does the name Dexory come from?

Andrei Danescu: 32:01

Ah, that's a very interesting question. So we wanted to have a a coin term that brings together digital twin technology, extra sensorial capabilities, which is how we, we look at our robots are autonomous robots that are capable of scanning. And the kind of, how do we, how do we make this, how do we make this into something that also sounds high tech? So that's how we looked at extra sensorial. We wanted to have a couple of X's, a couple of Y's in there. So we, we did quite a big, quite a big brainstorming around it. And we looked at XORI and then we wanted to bring the digital twin side of things. So Dexory is, is what we, what we landed on.

Tom Raftery: 32:43

Fantastic. Great. Andrei, that's been really interesting. Thanks a million for coming on the podcast today.

Andrei Danescu: 32:49

An absolute pleasure. Thank you for having me.

Tom Raftery: 32:52

Okay. Thank you all for tuning into this episode of the Sustainable Supply Chain Podcast with me, Tom Raftery. Each week, thousands of supply chain professionals listen to this show. If you or your organization want to connect with this dedicated audience, consider becoming a sponsor. You can opt for exclusive episode branding where you choose the guests or a personalized 30 second ad roll. It's a unique opportunity to reach industry experts and influencers. For more details, hit me up on Twitter or LinkedIn, or drop me an email to tomraftery at outlook. com. Together, let's shape the future of sustainable supply chains. Thanks. Catch you all next time.