Why is LEGO so popular with children? They enjoy it because it’s easy and they see success quickly. This is exactly what AXULUS wants to achieve when implementing IoT projects. In the 36th episode of the IIoT Use Case Podcast, Florian Beil (CEO, AXULUS) and Katharina Schmid (Head of Business Development, AXULUS) use the example of a warehouse to report on how they bring the topic of IoT to life for their customers through a modular building block scheme.
Podcast episode summary
With its smart subscription tool, AXULUS provides a standardized knowledge base enriched with IoT know-how, and the answer to how to get this knowledge on the road – or into a warehouse, as in the specific use case of this podcast episode.
AXULUS is a Reply company and sees itself as an innovative problem solver and idea provider through scalable IoT applications. The one-and-a-half-year-old start-up is active across the entire industrial landscape, supporting mechanical engineering and pharmaceutical companies as well as logistics service providers and electronic manufacturing services. Together with its customers, AXULUS implements a wide variety of IoT use cases by providing them with step-by-step guidance and supporting the entire process.
This podcast’s use case is about using an autonomous drone to inventory a warehouse. The “Thing” of the Internet of Things is the drone, a dynamic sensor that flies around and learns. It collects valuable data by scanning barcodes, taking photos and sending them via a long-range Bluetooth interface to a platform where inventory results are eventually visualized. The path to the solution works according to the building block scheme: I have a problem that I want to solve digitally – I access existing templates at AXULUS and configure them like LEGO building blocks individually according to my needs – the Solution Designers then edit and perfect my use case and initiate the installation workflow digitally.
Florian Beil has been working in the field of Industrial IoT for many years and has always found that “it’s not that easy to integrate a new technology into an existing system”. Exactly this challenge is addressed with AXULUS. The primary goal is to implement IoT technology in a problem- and solution-based manner, scalable and with individual added value for the specific use case – and as tangibly and simply as possible for the user.
Also in this podcast episode: Project work with PowerPoint and Excel adé, the most important KPIs and challenges in warehousing and which other advantages AXULUS offers.
Hello Florian and hello Katharina, welcome to the IIoT Use Case Podcast. I would like to welcome you all. Glad to have you guys here. I would say we start directly with a short round of introductions and then get into the content of the use case. Florian, I’m looking in your direction: Would you like to briefly say a few points about yourself and also about AXULUS, what exactly you do in terms of core business?
Hello everyone, I am very happy to be here. I am Florian Beil. As CEO, I am responsible for a new business area called AXULUS. It is a Reply company. What we do: Primarily, we help companies deploy IoT in a scalable way with a software-based solution, a subscription tool. In this context, we build a lot of dedicated use cases for customers, help them to use them and then actually have fun with IoT. We’ve actually been doing that very successfully with AXULUS for a year and a half now.
To briefly wrap up the round of introductions, Katharina, would you like to join us briefly and say a point or two about yourself and your role at AXULUS as well?
Thank you very much for having me here today. My name is Katharina Schmid. I’ve been working at Reply for six years and I do a lot of IoT for logistics and supply chain companies. In the topic that Florian and I are presenting today, I am responsible for business development. This means I work a lot with customers and partners to successfully equip them with the use cases we define and set up together with AXULUS. That is my main responsibility here.
Florian, you just said that it’s about setting up use cases in a scalable way. What does that mean exactly? Can you tell us more about that? For what purpose were you founded and what exactly is a use case from your perspective?
For me, a use case for IoT is a problem that a customer has, together with a technical solution based on Internet of Things technology. For me, a use case should always solve a problem and thus generate added value. And, of course, it should do it as often and as well as possible. I can always build technical POCs, sample projects, but it’s actually also about using that in a really scalable way and comprehensively. Why were we founded? This is also somewhat related to my personal history and ours. From the background: I have been working on the topic of Industrial IoT for ten years – i.e. use cases in industry, in production, at utilities, etc. I have found that it is not so easy to integrate such a technology into an existing production network. It’s not like smartphones, where you have relatively standardized interfaces and you can use the technology in a “download and play” way, but it’s a bit more complicated. A lot of customers that I’ve seen or met, they firmly believe that that’s the future and I think we all believe that. Otherwise, Madeleine, you probably wouldn’t be doing the IoT Use Case Podcast either. But it is difficult to really introduce such a new technology into an existing production process that is highly complex, that is automated, that is designed to run for 20 years without disruption. And we wanted to address this problem as an exclusive start-up, because I believe the technology has a lot of potential, but to introduce this technology in a use-case-based way, i.e. problem- and solution-based with added value in a scalable way, is another challenge that has not yet been addressed in the market. And that’s really our value focus: we want to help customers deploy IoT, generate added value and then, most importantly, scale that across multiple installations, not just one.
In the case of companies, there are usually different users or stakeholders involved in an existing process. That’s probably also the key, to bring those together and take it all in across the board, right?
Absolutely right. You’ve grasped a very important point there. Classically, of course, the customer is also organized. It has IT departments, production departments or finance departments. The exciting thing, or a bit of a challenge, about the Internet of Things is that various parties actually have to work together very seamlessly to ensure that the technology reaches production and, above all, is accepted by those who have to operate it. But also that it is built securely and scalable, which is then mostly by the IT department, for example. This cross organization and collaboration is of course also an important issue that drives the complexity of such IIoT use cases or projects. And this is of course also an issue we addressed with AXULUS – it brings the different parties together in a coordinated way so that then at the end of the day a stable and valuable use case emerges.
You guys have brought a use case today that we can use to talk about this specifically. Katharina, you said you’ve been in the logistics and supply chain management field for years. That’s your expertise. Can you introduce us here to the topic of what classic problems or challenges in logistics are, so that we can understand your solutions?
In the logistics industry, a bit of a problem is that customers naturally expect some package to arrive at some point. In some cases, they don’t understand the massive effort behind it – labor costs, transportation costs, and so on. For example: I produce somewhere, raw materials are shipped somewhere, then converted into semi-finished products, then further shipped to distribution centers until they arrive exactly on the day, in the time window, where you want them to be finished and you just have to open the door. Behind this are hidden expenses that no one is really aware of. I think that sometimes gets lost a little bit. I want to receive my package on Friday afternoon at exactly 4 p.m. and that’s actually a lot of effort to coordinate with many complex dependencies, with partners, suppliers, customers, between industry and trade to make the balancing act. That’s definitely a really big challenge and also, of course, associated with high costs in terms of personnel, transport and warehousing. That means in large warehouses or large storage halls, where hundreds of employees, a whole team works day and night on it, where trucks drive in and out and of course you get the things that are urgent out again as quickly as possible and put the things that have to go quickly somewhere else than those that perhaps always stay a little longer. Of course, this creates an extreme amount of dependencies, because one person waits for the other, the other wants to get the stuff out, someone has to get the stuff somewhere. There are definitely a lot of people on the road who have to coordinate perfectly. And if someone somewhere is waiting for the other person, that often opens a can of worms. These are challenges, especially in warehousing, and that is of course also what our use case is a little bit about, because a warehouse like this has to work extremely efficiently, of course, and everything involves a great deal of effort.
What are the classic tasks and challenges of these individual employee roles in which your customers are active?
In a classic warehouse, there are employees who do the picking. That means, they are employees who receive so-called pick orders and then process them, assembling materials that then have to be delivered further for a specific customer order. This means that they run around in the warehouse and everything has to be done quickly, one after the other. The things are put into packages, weighed, labels are put on them and everything is prepared as we know it when we receive it. Then, of course, there are the forklift drivers who take down the goods that are at the top of the building, for example, and then make them available at the bottom for the very people I mentioned earlier, who then open up these pallets again and take out the individual parts. And then, of course, there are the people who unload the goods and put them back on the truck. That is, they are those who are at the gate. Of course, they also have to take the pallets from such a truck extremely quickly. Normally, 33 euro pallets fit into such a truck, and they have to be taken out as quickly as possible, quickly put in the right places, and then loaded again so that the driver can drive on again. And of course the whole thing is supervised by logistics managers or shift supervisors who have an overview of the order situation. How far are they? Are they already 50 percent done or are they actually done? And I just need a delivery bill and then I can monitor it that way. That’s about three or four typical roles in a warehouse like this.
How digital are your customers already today? Some of that sounds very manual, what you’re describing. There are probably also different gradations. What is your assessment?
So that’s a mixed bag. There are a lot of people who still work very manually at first, but then reach their limits, because of course there are so many dependencies. There’s everything from Post-its stuck somewhere with inscriptions like “I need another pallet of brush ABC now, the forklift driver has to get it for me” to ultra-modern equipment with a wide variety of technologies, such as voice devices that whisper in your ear where you have to go, and then you confirm it with code numbers. Or you work with RFID devices or RFID tags that do this as automatically as possible or, as in our case, with drones that scan the data and thus make the employees’ work much easier. But there’s really everything – from highly manual to very well automated. There are also automated warehouses where robots automatically move pallets to and from places.
Now you’ve given me the perfect transition: The bottom line is that we are now thinking more in terms of hardware and these individual systems. We always talk about data when talking about IoT, perhaps real-time data, and also generally data provided by hardware, by systems, and by things. What data is of particular interest here in warehousing?
A typical KPI is the lead time. How long has the stock of an item been in the warehouse? When did it arrive, when did it leave? Which items are fast movers and which are slow movers? That is, which items are often bought and are often delivered, and which items actually rather less. A very important KPI is the number of picks per hour. That is, how many individual withdrawals do employees make per hour. A single removal is, for example, one time to take out any brush and then put it on its pallet, which is then assembled for the customer’s order. Of course, a whole lot of data comes together, because if you look at or imagine a warehouse where 100 people work, everything has to run in real time. That’s where the system has to tell the employee in real time, “Now you go to this place and then you take out five pieces of this item.” He then goes there, scans this space. The spaces are then always stored with barcodes, so that this works fast. It has to be tracked immediately so that the system that does the analyses knows, there are five pieces taken out now and I have five less of that item now. Because, of course, there are always orders coming in. And to keep the stock accurate, it automatically generates that you need replenishment. If, for example, there are no more brushes available for the pickers, then there is an order for the forklift driver, who then in turn sees: Okay, I have to go there now and it’s urgent now. He gets the info, “Why don’t you drive over there and get this whole pallet down and take it down to the employees.” A whole lot of data is accumulated there, which can then be analyzed to see where dependencies may exist and which items are particularly important. Which items might also have the best margin? Where could I possibly put a minimum order value, because otherwise it is not profitable? At what point does freight forwarding become profitable? And what better to send with parcel shipping? Of course, there’s also a difference between having a freight forwarder come in and fill up an entire truck and having the typical UPS driver take the packages. So there are various possibilities and a lot of such data is accumulated and in principle the aim is to keep the costs of this warehousing as low as possible and to let people work as efficiently as possible.
Does that also mean these warehouses are in competition with others who place orders there for such logistics services, or how does that work exactly?
Yes, for example, the space is rented by typical logistics service providers. There is also the possibility that you rent not only the place, but also the staff to it. That is, if my company now produces jam, for example, and my biggest customers are large supermarket chains in Germany. But I only have one plant where I produce everything, but everything after that is no longer my core competence. Then I might want to outsource, have the process managed. I can then store this somewhere temporarily and someone takes this service off my hands. So this is either outsourced to typical logistics service providers or the companies do it themselves.
The different areas of responsibility and KPIs you mentioned sound very complex. You also mentioned the use of drones – can you tell us more about what exactly each employee’s job is and then what your solution is for those employees?
That was a use case where we were supporting an inventory scan with a drone, so inventory counting with a drone hardware. Here’s the problem: once a year, they say you have to take inventory and count your stuff so you know what’s in stock. For this purpose, there are of course different possibilities of inventory counts. A very common one is to close the warehouse for three or four days, often at the end of the year between Christmas and New Year’s Eve, which is a lot of work for us then. Then the warehouse is closed and an inventory count is triggered. The employees who are in the warehouse every day and put things away and take them out again are then also responsible for carrying out inventory counts. This means that, ideally, the warehouse is first cleared of orders, they are all processed and then an area in the warehouse is closed and said: “Okay, we are now taking inventory”. This means that the stock may no longer be reserved. It’s just that when you do inventory counts, you often need temporary workers or students to help out. There are forklift drivers who then take out the pallets from above with special forklifts, so-called reach trucks, and then bring them down and every article and every space really has to be counted so that afterwards you can show your auditor that you have done it properly.
Florian, you said at the beginning that it’s about scalable use cases. Now we have learned about the different ways of functioning in a warehouse. How exactly does that work now in this organization with AXULUS?
I think it has already become clear that it is a relatively complex process of logistics in the warehouse. Of course, this also means that if I now have a solution for taking inventory with drones in one warehouse, this does not necessarily mean that it can be done identically in the next warehouse. Because it might be cut differently, the processes might be slightly different, the data formats might be different. What we observed before we started with AXULUS was that many customers approach a topic like IoT roll-out in a very analog or manual way. So typically a project is set up, then workshops are done. What problems do we have that we could address with IoT? Then solutions are built and we then try to carry them across the individual plants with service teams. This is actually a very manual process for a digital problem, I think. If we were to arrange a meeting today, the three of us in private, we wouldn’t start a project; we actually already have a digital communication platform. We use WhatsApp, Facebook or whatever it is for that. That was actually the basic idea. Why don’t they do it similarly, in the whole topic of IoT innovation. That means identifying use cases, then configuring the solutions for them and rolling them out accordingly with digital workflows. If we now take the drone use case as an example, then we have stored a solution template in AXULUS as a use case. This means that when a customer logs in to AXULUS, someone in the warehouse or an operator who needs a specific solution, they first see a starting field and can search for specific use cases templates. Template for us always means problem description and also really quantification of added value. So it actually usually even starts in the value design of the use case, so what is actually the problem that I want to solve and what would it do for me. Accordingly, he can then take the use case and see if it fits. If not, he can modify that with a design tool. I want this target value, this problem, to be solved. These are my asset landscapes, my interfaces, and for that I now need an IoT solution. In the next step, he would pass this description in the system to the solution designer, which are practically the colleagues or companies that can build or deliver IoT solutions. And here, too, it is then very modular. It is a modular engineering tool in which the solution designers can practically configure their solution – i.e. technically assemble the solution from existing modules and integrations, deploy it and, above all, describe how the operator in the warehouse will ultimately get this solution up and running. So the whole thing explained step by step and the entire enablement for the solution also stored. And when that has happened, it is published and the colleague on site in the warehouse can – depending on how automated it is in AXULUS – deploy the solution per One Click so that technically the solution is running. Step by step, the system guides him through the installation and then, at each step, he knows exactly what he needs to do to get the solution up and running. This has, of course, digitized the entire process and made it scalable.
Before I go into this process again right now, here’s a question in advance. We’ve talked about IoT and drone inventory – where exactly is the IoT interface there now? Which devices provide live data here, for example? What is IoT about it?
So the Thing of Internet of Things would be the drone, that’s kind of your sensor. This is a very dynamic sensor that flies around and learns as it goes. It collects the data: it scans the barcodes, takes pictures of the things that are lying around in the warehouse, and that’s the data that this drone then sends to the network via a long-range Bluetooth interface accordingly. And then it goes to a corresponding platform, on which the inventory is then calculated and displayed. This is a simple use case for now. But there’s also a bit of intelligence behind it, because the drones are autonomous, for example, which means they learn how to fly. The data is analyzed accordingly with deviation analyses. That is, where do I have deviations from what I would expect? This would be, for example, an interface to an existing warehouse management system where the target data is located and the actual data then comes from the drone. These are the interfaces that must now be considered for the individual use case. And, of course, that shows that I might have to change some things when I go to the next Warehouse. Because the drone has to learn to fly again because the warehouse is cut differently, maybe they have a different warehouse management system and the data has to come in through a different interface. I then have to configure a slightly different solution in AXULUS logic accordingly and form a variant. And we support that very strongly with the solution designer that we have in AXULUS.
That is, this inventory drone case is one use case of many, so to speak. I have a template in your tool that I use. There I find advice by storing information that I need to implement this use case. The next step is I’m in contact with your solution designers and they put something together and then we go into implementation and we can use the services that are selected to fit that use case. Have I summarized this correctly?
Exactly. The installation is a digital workflow or like a recipe – a step-by-step guide of “Now how do I get this solution into use in my field?” based on a technical solution configured by the Solution Designer. There’s a bit more to it than that, of course. For example, there is also a chat and community functunality. I can also chat directly with the solution designer. But basically, it’s really about digitizing that instruction or those workflows that I need to the point where I can install and use a solution without having to have service teams fly around and install something on site, but really roll it out scalably afterwards. Of course, then you want it faster and faster. So once I have built a warehouse solution for a warehouse in Munich and have the digital workflow, then of course I can use that as a template for another warehouse and probably won’t have to build it completely from scratch, but just adapt it slightly. In this way, you gradually build up your knowledge base and can get these solutions out faster and faster.
That saves me an incredible amount of time. This means that if I now put a new employee at a different location on exactly the same task, he or she can team up with the other and they don’t have to start from scratch. How are customers doing this so far?
It depends a bit on the customer. Of course, there are customers who are still very much at the beginning. For them it is already very helpful that there are templates at all, that I have a collection of ideas. Now there’s a drone inventory use case for the warehouse, there’s a drone inventory or work in progress (WiP) use case where I’m doing takeoffs in production and looking at where are things lying around that shouldn’t be lying around there. You can really get your head around the issues there. All the classic projects I’ve seen so far have been project work. Someone comes in, makes a presentation, PowerPoint or Excel. There are already individual systems that go a bit in this direction. The hyperscalers that are also trying to build certain application templates. But that’s not adequate for the complexity we have out there. This is not a predictive maintenance dashboard that I can slap on everywhere; instead, I have to tailor it very precisely to the use case.
The fact that you already have a wide variety of use cases and that these templates are constantly being developed further also creates a certain intelligence that I can then build on, right?
Right. There are several great advantages that you may only see at second glance. What you described is exactly the case. If a customer or we – depending on who is doing it – use AXULUS and thus look after the whole topic of IoT, we naturally have a standardized information base. What use cases are there? Which solution templates are used? How is this rolled out? And then, of course, I can use it to learn again. You can do things like recommendation engines. For example, making suggestions like: 80 percent of your colleagues, or anyone who is qualified as a press line manager, actually has the following use case going on, take a look at that, don’t you want to install that yourself. This database is already a great advantage or this structured information that you have, which I otherwise only have in Excel or PowerPoint. And of course the next big advantage of AXULUS is that it is a subscription tool. It always runs. That means that once I have that enabled at the customer’s end, they can continuously generate ideas about what other problems there are that I can address with IoT. And doesn’t have to wait for the next IoT project to come along. The third thing is, I can coordinate the roles accordingly, as you said before. This is a complex organization that will deliver specific parts each to the IoT. Of course, I can scale them as much as I want, because in the AXULUS system the roles are stored and also how they work together. This also allows me to roll it out and have it work together in a clean, coordinated and controlled way. I think these are the main advantages.
I recently had a client who wanted to optimize supplier rating management. Maybe also to exert a bit of pressure or to have a bit more control over the whole thing and to be able to say, for example, “The pallets you’re delivering are good and they’re actually bad.” How would that work? At this point, would they contact you guys and say, “Hey, I need a different template, I need the supplier rating management template now,” or how does that work exactly?
Yes, that’s exactly how it would work. That would be a need agent saying I have a problem and I need an assessment tool. The could then log in and could load a template and configure and customize that as he needs. What do I want to achieve with it? What are my parameters that I want to influence and what are the processes that I want to influence there? Or he can also start a completely new template, if it’s a completely new idea, he can of course fill that in accordingly himself. And then that use case would sort of go back into solution design. The solution designer, whoever then makes and configures the solution, could then of course already load the technical solution path accordingly so that the entire IoT use case is in fact realized. But initially you would just open a new use case. But the good thing is, once he’s had the idea and it’s in the system, then everyone else who has access in the system has that idea at their fingertips as well and can think, do I maybe need that with me?
This technical solution that you had just mentioned, which is now more in the area of implementation – how does that work exactly? That is, you guys make a recommendation for certain services, whether it’s an AWS service or a Microsoft service and say, this is the value-add that the implementation would bring now, plus the sensor, plus the hardware, plus the software, or how does that work at that point?
AXULUS is first of all technology-agnostic in principle. This means that the customer actually decides which technology is used. What AXULUS does, it actually builds a kind of LEGO building block kit. Technology today is a bit like LEGO. Theoretically, I have quite a few interfaces and APIs and I can put them almost all together. Not all of them, of course, but I have a lot of choices. And what the Solution Designer in AXULUS gives you is really this LEGO building block kit. That’s where he sort of deposits the LEGO building blocks, which can come from AWS if the customer says I’m on AWS, or from Microsoft Azure, and then also suggests which ones would fit together. So tips like, “This sensor, I’ve hooked it up to Azure before, you could take that again,” or “There’s the following user interface (UI) that we’ve already built for a use case like this, take that again and you can launch right away.” Of course, the system also knows when things are not yet there. Then if the solution designer says this doesn’t fit exactly and modifies it slightly, you can also funnel that out into the development process that this UI is being built. This is then filed as a LEGO building block so you don’t have to build it again next time. We call this feature Modular Design: you break everything down into the individual building blocks and remember which building blocks you have and which ones have already worked together. And you add new building blocks whenever you really need new ones.
You are not only active in the drone and logistics environment. How might this use case be transferred? What other issues do you see? Simply to give listeners the breadth of use cases.
AXULUS really covers these classic industry use cases: from Electronic Manufacturing Lines to New Cleaning Services and much more. We have large machine builders or component suppliers who are building a configurable IoT store for themselves with our solution. That means you can contact them with your problem and they will then configure a solution for you as a customer from their building blocks. Of course, that’s also a great customer experience: I’ve described my problem and then I can see in the system where they’re currently at and when my installation workflow will arrive. There are really all kinds of different areas: Mechanical Engineering, Electronic Manufacturing or Pharma. The topic of innovation that we serve with AXULUS is actually overarching.
I think there are a lot of areas. One is for people who don’t really know how to get started yet. First of all, they are overwhelmed with all the many providers and technologies that exist, which want to record any data from any machines, of which they had to painfully realize how important they are for their production. On the other hand, also for companies that already operate highly automated data analyses. That based on a whole lot of sensors that send data to various databases, then analyze that and recognize patterns, for example, also via AI. There is really everything – from small to large.
If I’m a listener here and have a similar challenge in front of me, want to bring a use case to implementation in a scalable way, with some consulting component in WiP – what’s the best way to reach you guys?
Thank you both so much. These were really exciting solutions with a lot of added value for future roll-outs in the field of IoT.