Transcript.

Erik: Welcome to the Industrial IoT Spotlight, your number one spot for insight from industrial IoT thought leaders who are transforming businesses today with your host, Erik Walenza.

Welcome back to the Industrial IoT Spotlight podcast. I'm your host, Erik Walenza, CEO of IoT ONE, the consultancy that helps companies create value from data to accelerate growth. Our guest today is Laurent Laporte, CEO of Braincube. Braincube is an IIoT platform suite with business and expert apps that's designed for manufacturing. In this talk, we discuss the uses of advanced algorithms combined with prescriptive insights to optimize production processes at the batch or even at the unit level. And we also explored the purpose of an IIoT platform in addressing the challenges that it and OT leaders are facing, as expectations increase for high quality, low lead times, and predictive visibility.

If you find these conversations valuable, please leave us a comment and a five-star review. And if you'd like to share your company's story or recommend a speaker, please email us at team@IoTone.com.  Finally, if you have an IoT research, strategy or training initiative that you'd like to discuss, you can email me directly at erik.walenza@IoTone.com Thank you. Laurent, thank you so much for taking the time to speak with us today.

Laurent: Hello, Erik, I'm here in France in the cold winter and very happy to talk to you.

Erik: I was looking on your website at the Chateau that your headquarter is in, quite beautiful and really quintessential French tech company. But where are you sitting today?

Laurent: I’m nearby a biggest Chateau, I mean Versailles, where we have the Chateau 14, the King of France. And we have our sales offices for Europe here. So I was working this week with the team here to have to set up our new season.

Erik: Okay. Well, that Chateau is a little bit more expensive, I imagined.

Laurent: Yeah. But our Chateau was a good story. It was an opportunity. When you grow fast, you have to change your offices every two years in every regions. And this was the opportunity to have a really nice place where we can beat around, we have a lot of lands, and it's inside the city we want to live in. So we're pretty happy about this.

Erik: Tell me about the development story here because you set up Braincube in 2007. And as I understand you, and the other two founders are all coming from an operational background, and so that was a pretty ambitious time to be setting up an IoT platform because Cloud was just being adopted at that time. What's the founding story of the company?

Laurent: This story started probably in 1990, because I entered in one of the most renowned French Engineering University named [inaudible 03:21], which has been founded by we 14, again, so it's a longest story. And I have hired the first career in manufacturing, and I've been passionate about production system.

During this career, I made my two buddies, Helen and Sylvan. Sylvan was working in systems and Helen was more on a business operation side. So we had this first 15 years of experience working in the manufacturing business in different countries. I did run some plants in the USA, in Italy, in France, in Finland, so we were all about production. And their early stage of our first idea was two things. The first is internet looks like to be a connector and there was no connection between the plants, the supplier inside the plants. We didn't have any technology, technology-speaking, any connection between the different actors.

And the other thing was we were starting to use more and more data for problem solving with Six Sigma approach. But we were very frustrated about it. Because data collection was manual, long data cleaning was not existing. And the way to use it that we’re pretty old. We were using this statistical study and it's the technology is started in 1936. So the first idea of this before Cloud, before IoT, and you can see connectivity IoT. We had decided and was to leverage the fact that is easy to build a website for developing data analysis technology, coming from Six Sigma mindset to solve faster problem, to be always in condition of analysis.

This was the first idea back in 2005 and we started to build a platform. And we started to build what we were calling websites and we started to push data there from any sources we could get. And we started to do some visualization, data transformation, multifactor analysis, something like this. So the early days were [inaudible 05:31] a numeric transformation of Six Sigma, by an automation of the data collection.

And coming back to the Cloud, the Cloud was not really existing, IoT was not existing, big data was just emerging. The Cloud was for us a trick because we didn't have to develop a very complex software at the beginning, we developed just website. And we could make modification by ourselves very quickly and we could have a version a day if we needed some corrections.

So the idea was to choose internet first because it was easier to develop software, and later, it was because we could share the power computation. And then that's where we entered the Cloud a few months later where we discovered that we could now massively upload data. But to be able to do something with them when they did the giant power computation. And the only way to do this is to do a Cloud scalable infrastructure.

There was some choices at the beginning and choices made with the vision, and there was other things that we discover on the way that was advantages that that showed up for us because we probably had the chance to go on the way of the cloud technology and the cloud become big, and so it become real for us.

Erik: Did you already have the vision for a more integrated AI platform? Or was it more of an individual application that you started with? What was the starting point?

Laurent: Yeah, it was a platform where we were receiving data from file exchange system. We were generating files at the plant level, and they were pushing with a transfer protocol, the files into their dedicated website. We were having three applications. We will call them learn, think, and act. And the idea was, Learn was a visualization module, so you can visualize, centralize information, making some charts and graphs simple; Think was a very specific module, where we were doing only multifactor analysis, well, taking a long time of computation. But it was really where we had the first value of Braincube; and the Act was just a live dashboard, where we were helping the customer to use the defining tickets through visualization or analysis.

Erik: So today, what I see on your approach, “connect, manage, improve, transform”, but probably with a lot more depth than you had in that initial version. Can you give us a high level walkthrough of the different aspects of the product and the value proposition behind them?

Laurent: In really the early days, we were all about problem solving because we're coming from Six Sigma. So the idea was to reduce any type of viability by identifying the factor we should control to kill this viability were the cost most of the time. So we were having typical project like reduction of energy, saving energy, increase of speed or equipment, try to reduce or eliminate some defect or reject, try to reduce the consumption of some of the key elements inside the product to make them less expensive. We were all about reducing variability.

But when we started to sell that, and it has a lot of value, because we were hoping customer to solve prices or to boost their continuous improvement program, when we started to discuss with them, they were having many, many more ideas about how they would use this data that were centralized. They wanted dashboards. They wanted [inaudible 09:39] system. They wanted some prediction. They wanted to be able to feed other systems.

So that's where after a few months of operation or the first years, we discovered that there was an interest is transforming this horizontal approach on solving problem to really larger scope is what can we do to support manufacturing operation with this platform. And that's where we started to really feel this early the stage of IoT. That means we could connect anything and we could use this to have many, many, many use cases.

And what Braincube decided to do is to build a suite of application agent Cloud that will cover the generic needs of the engineers and the production people, so we really helped them to have easy made by click dashboards, data visualization. The first step of problem solving analyses, we have also statistical process control. So we have many apps like this, that are really generic, and most of our customers really enjoy to do them, delivers a lot of value.

And then we build a second side in the platform, which is okay, we cannot know that all the new ideas that people will have, so let's create a framework where we can develop any use cases based on the information we have inside Braincube. And so we started to build kind of local [inaudible 11:09] type of application where the customer can do like your breaks, they can make their own prediction, and put the prediction into visualization and put this into a larger dashboard, and so on and so on.

So today, we have this, you come, you set a platform, you start us with our generic application, and you can build your own transformation journey by developing as many use cases as you want. And Braincube support this because it will run for you all the new application you will build that will be your applications.

Erik: So you remain focused on manufacturing, but you're covering discrete in process. It looks pretty comprehensive, from your website, at least. What defines a good customer for Braincube? What are the requirements of a customer in order to be able to extract real value from the solution?

Laurent: You see, people sometime split the industry between discrete manufacturing and continuous process industry. In our world, it's not exactly this; it’s more automated industry versus manual industry. And we don't go to the manual operations because they don't generate data. But everything that is automated with robots, automation layers, the control systems, and with traceability between operations, that's the typical customer.

So we like mass production. We like automation. We like a little bit of complexity. It means if it's very obvious, we can support the customer by having their performance KPIs live and stuff like this. But when the customers start to struggle, because there is a level of complexity due to the volume to the pressure on costs due to the complexity of controlling the processes, then Braincube will deliver a tremendous value. We can have customers in discrete manufacturing. If they manufacture products that are expensive enough like tires [inaudible 13:20], then it has a lot of value to work with them.

And we go also in process industry where the generates such a beginning of data, and they have interaction of parameters. They don't even understand where data can really show them sometimes why, but most of the time what to do. And what to do is very important manufacturing because you like to understand. But the first thing you need to do is to match your targets. And so the system is really supporting you in this decision making process to be sure that you're wasting less and less opportunity to be at your best.

Erik: So it's not just visualizing the situation, but it's to some extent diagnosing or prescribing actions. But then I suppose there's a lot of intelligence from the operators that needs to be embedded in the machine, how do you then embed that insight into what actions might make sense to do based on a given situation because that's hard to do from a pure horizontal platform?

Laurent: The first way to improve a system is to observe the system. So visualization is still very interesting. And if you don't solve your problem with visualization, you will help to describe the problem better. And then you need to go on a deeper level, try to find the underlying interactions between parameters where it's not to be complex.

What happened in a plant, it runs 24 hours. So what the people are doing what the people are thinking, all their actions, all their reactions are inside their historical information. And we collect data from this machine. And from the raw data, we put this data into a context, which most of the time is what we try to recalculate the exact conditions of the manufacturing off every product. We call it the virtual tween of every product. And this is the level of information or what happened in the past, what decision was made. Any problem that occurred are engraved inside this level of information.

So now we need to use learning technologies to learn what happened, what was good, what was bad, what we should consider of doing again, what we should not consider doing again. And we can sort out from all this action that has been engraved into data. We can sort of what the plan should consider are the future way of running the operations. And so you see, we use the data, but we transform the data into a level of information that is ready to talk and then we use learning technologies to make sure that we will grab maximum insights out of this level of information to answers to the people questions.

Erik: And then you get into this very, very common challenge, which I imagine is even more of a challenge back in 2007 of serving the operations function, but having a solution that's Cloud-based software that requires quite a high level of data analytics competence and so then you have, maybe the IT department or some other, maybe now might be differently defined, but different functions that maybe have that technical competence. I understand that operations is the customer here. But who are going to be the users of this platform given maybe the complexity of the data involved?

Laurent: We have two department using Braincube. The first one is the IT department. Because when you do a project that deals, in fact, you have to participate to a larger project, which is the company data project. We, as Braincube, we focus on the manufacturing side of this data project. But every company today, they have a vision about how they want to deal with data, their business data, their process data, their administration data or whatever data.

So, Braincube is first working with it, and we support their vision and try to improve their vision and build with them a better architecture, at least, to serve our proposal, which will be a manufacturing optimization. And then IT most of the time, they also lead us. Because when we set up this platform, one of the vision is to be able to develop on demand the application and the tools that doesn't exist on the market that has been recognized as critical strategy for their operations.

And it's part of their mission to use data scientists or I would say, suppliers that are dedicated for software development type of mission. And so we work with them because we bring them a lot of tools to support this mission. Then the other big department is the Operation. So in operation, we simplify it as three types of users. You have the managers, you have the technical staff, and you have the production people.

So, Braincube for every either of the three types of users, we help them to save time, and make better decision by leveraging their data. What does it mean? It means when you’re production people, you need to have life information, life prediction, life recommendation, or simply we need to take over some of your decision. It's easy with Braincube to use some of our modules to develop the prediction with auto machine learning type of technology to set up some different [inaudible 19:28].

So we will cover the workshop with a screen that will help the people to have the right information at the right time or to add the right instruction at the right time. For some of the process we go up to taking over the decision making life by making an artificial intelligence system learning continuously and really taking over the most complex decision they have made like multifactor optimization.

The technical staff is mostly working it as a support, so what they need is prepare the future. So they will mostly use more big data application to make learnings that will help them to design the next step. How do we have to improve this equipment? Which equipment we have to change? Which next technology would be valuable for us? They really work on this as a support. Sometimes they do problem solving. But most of the time, they do optimization for the future.

And the managers, we help them to set up the dashboard at a detail level that help them to and  to cover an opportunity for improvement that are not that clear when you just use your P&L, the Profit and Loss accountabilities of results. That means today, we can, for example, give them, if you are a tire company, and you manufacture 15,000 tires a day was multiple models, you have an average cost per type of tires, maybe every day, which is already very good, sometimes every week, most of the time every month.

With Braincube, every tire as is exact on cost of manufacturing. So you will have some models with a lot of variation on the cost. And you need to understand something and you will have some other models with low variability between the cost because the manufacturing more stable for the pole because the way we do it is a better way. So we really have to go to this level of detail and then they can task their staff team to go and use their experience, their knowledge, and today their data to try to make savings in all this potential that are most of the time unseen because we measure them as averages.

This is a game changer. We don't need average anymore. We need to really measure the exact viability of what is happening, because that's where we can identify and very value the project we should consider to make some improvements.

Erik: I was just having a conversation with one of my team members today about the oven process in automotive OEM you paint and then you go through a few ovens to dry the paint. And we were just thinking about, okay, this is apparently consumed something like 60% of the energy in automotive OEM facility, this painting and drying process.

But if you actually could measure the readiness of a vehicle, you might find out that maybe, first of all, it should be lower, but second of all, you might find out that based on the different paint color and paint chemistry, some should be 4 minutes in 20 seconds, instead of 5 minutes, but you're applying that standard.

I think there's a lot of things like this that are simply hidden because you apply a standard, and the standard tends to be at the high end of the distribution because that gives you the safety margin. And that means there's a lot of these efficiencies throughout an organization. And I suppose a solution like Braincube can really start to let you get into these and operate at a much more granular level than you typically would based on the rules of thumb that a lot of operators work by.

Laurent: And I know very well, this example, because Braincube is covering some paint line in automotive for American car supplier and European car suppliers. So we have some experience in this, but it's exactly what you said. The first, you can build the standard operating procedure. And I see them at the road. I mean, this is a road, you have margin, but your first job is to stay in the road when you drive. But it's not the optimal trajectory.

And so what we are doing with Braincube with these two steps, I mean, we support the customer to get back on the road because sometimes they are out of the road. And they sometimes they don't understand that when they push too much this product on the station number 12, then they are a reaction on station number 36 later, and they don't understand this. So we help them to set up the best average condition and we can see as a new process window.

Then when we are in this, we already removed some of the crisis and we get back to the average. But if you want to really to make more money, you need to optimize life. So that's where from this big data analysis that helped us to set up the optimized process we know, then we start to launch station by station or group of station by group of station algorithm running live and that will adjust directly the parameters of we can adjust the one we decided to adjust replacing human decision. And we can do exactly what you said. That means we can pair paint per body, pair according to what happened in the manufacturing, the building, the painting of deals, we can adjust the temperature of the oven, we can adjust the time the car should stay in to eliminate or reduce the risk of having small holes, changing colour, a problem of layering, so stuff like this in the future. And we do it for many, many different processes.

If you take paper industry, you know, you can imagine that you receive chips of woods that are coming from forest. Even if you get the same species of wood, you have the problem of the season, the problem of the weather, you have to problem the storage, and you will process these fibers, and you will have to mix them with most of the time to do recycled fibers coming from newspaper. So your raw material by itself is always something that is you know what you have, but you know the average of what you have.

And so what's coming up into the head box of paper machine, for example, it's a mix that is in average always the same, but it's pretty inconsistent when you look into details. And this small inconsistency is creating a nightmare if you have not the ability to adjust all the whitening agents, the retention aid agents, stuff like this. So yes, we can run by habits. This is a whole time and say it's too complex and I don't improve. You can run by optimizing the average condition and it's already better, you are more stable.

But if you want to be very performance, you need to go to the live adjustment. And I don't know anybody in this planet that can adjust live, the perfect set points for several, I would say dozens and dozens of key inputs and different targets. It means probably 10, 20 quality energy or whatever targets. So the future is and today we start to do it is to have algorithms started to do it for you to finding the best from the learning and creating new experiences that will help to enrich the learning experience.

And that's the magic of the factory of the future. That means we call it the smart factory. And they will be smart because they will be capable of reproducing the best decision in every condition 100% of the time, which human beings cannot guarantee. And we will lead to the human all the rest, reaction to incidents, the preparation of the next step, the new machines, the new technology let the innovative mindset of the people building the future. And what's current, the manufacturing should be something that is becoming pretty obvious where we have to check, load, control, but not to decide and to sell because it's becoming too complex.

Erik: If we dig into this example that you just provided about raw material inputs into paper manufacturing, then in this case you would need data also outside of the walls of your factory as well. So you'd want data ideally from the supplier, maybe from the weather system. So these are all data input  somebody might be processing through Braincube?

Laurent: Yeah, that's the magic of internet. It's easy to today, if the world wants to connect the suppliers. Let's say we're doing some packaging paper, which today is most of the volume in paper industry because all this transportation of goods, we need car boxes. So, you are using recycled fibers. So with different grades and things you are using species of wood with also some durability. But also when you are manufacturing, for example, one of the glue you use is starch. And as you can imagine, starch is not a chemical product, it's a natural product. So you have a lot of variation.

So you can run the try to adapt without knowing exactly the exact starch you are using today. But starch is costing you so much that every percent of starch is really making your profit your benefit at the end of the month. So if you start to deal with your supplier, normally you will know the exact type of search you are using now. But you can't help to specify to this guy that starch will cost less, and will be more efficient for you.

So by teaming up the chain of the suppliers, you make a giant process line, and you open the opportunities for improvement. I have another story out of paper for this.

We were at the French giant helpline manufacturer. But they were starting a new model of plane. And to do this, they're always fighting against the weight and stuff. So they were using some new metal, a metal that has been invented, designed, engineered for this plane. So, as you can imagine, nobody has any experiences in how this metal will be made. Manufacture, will it be easy or not? And so the customer was complaining about quality problem at the end of the matching when the pot was ready to be a symbol.

There was some strange behavior of the material. Okay, so they were not looking inside our plan. They were complaining about the supplier and the supplier were the Forge. They were preparing the metal. They were not the foundry. So what they said they complain at the foundry to their own suppliers. So we had a chain of three companies: one foundry, one forge, and one machining company are the final customer. And so they were fighting for months, it's your fault, no blah, blah, blah. I respect my specification.

So when we were a supplier of one of them, the middle of the company. And so this guy idea, I need to say what if we do something where there's no guilty, but we bring all of the information we get from the beginning of the line to the end of the line, and we crossed them? And we see if we can as one team that we can solve this problem. And what happened, they were right. They were no guilt team. That means everybody could add something in the process. And everybody understood why that we totally, totally master the manufacturing of this new metal and the problem completely disappear.

And what takes the longer time in this was to set up agreement between lawyers to be able to do it, took few weeks to do this, it took few days to solve the problems. So you see the future optimistic will be that company will decide to collaborate because in their best interest, they will make more profits by being optimized all together than trying to secure a position where specification that costs money for everybody.

And we are having more and more chance to link the different actors in manufacturing for many reasons. It can be quality reasons, but it can be also supplying reason. Before you need this famous starch and you consume a lot. But your paper production depends on the grades. So you will use plus or minus starch depending on what paper you're manufacturing. And sometime you have some downtime shutdowns. You can have many things that slow down your production.

So the starch company has to be aligned with you to make sure you will never have any shortages, or you will never have 20 starch coming in the plant and you don't know what to do with all this trucks. So with IoT, and using internet, you can easily set up a dashboard where they know exactly the consumption life of starch at the customers and they know exactly the number of trucks, they will ask to ship the next day to make sure they match the exact demand.

And it changed the relationship. There is no more documented between and waiting for agreements and things. So well, I think this is the future of IoT is connecting everything in the plants and everything between the different actors with the same ambition of being fair being optimized, making the things easier, smoother.

Erik: Your point around the fact that it takes the lawyers three times as long to come to an agreement as it takes to find the solution is an interesting point as well because there's also the potential to develop much stronger bonds between a supplier and a customer. If you have this system set up and you know that your supply chain is optimized for this customer, it's less likely that the customer is going to be shopping around for other suppliers because you've already invested the time to optimize for them and your customer is getting a better performance here. So I think this can also be strategically used to build stronger relationships.

Laurent: I totally agree with it. It's always hard for a human being to change of model. There was a model of predictive models. And today there is a collaborative model. So it's hard for the piece on people to change your mind. But what happened in this project is, first, everybody was fighting, then they try to unify, when they unify, they solve the problem. And we had an award for this in the company, because it was the top management discovered that it's possible to do this.

So I think we are at the early stages of this, because people will realize that manufacture is just a chain. And if we want to be good, we have to make this chain, again, as open as possible. I'm not saying that any one of them share their knowledge, the secret sauce, and whatever, they didn't do this, because everybody could process data. And you can anonymize your calling data, you can run and you can tell them, okay, factor number 12 and factor number 47, output became very seriously. And they will deal with this. And we'll come back to you and say, okay, we can control it on the next level, or there's something we don't really control.

But you don't even have to know what exactly happening. You just need to finger point the thing that it should look at, and it saves so much time to everybody. And if everybody agrees to make the little efforts, the result is so big.

Erik: Well, let's break down the tech here a little bit. I'm curious actually, whether Braincube is a single solution or portfolio.If I'm looking on your website, you have the IoT platform but you have Cloud in Edge broken out separately, and then a Cloud plus Edge, you have digital twin, and then you have advanced apps. Are these all separate products? Or are these basically functions, or architectures for the same product? How do you look at the portfolio?

Laurent: We have a lot of tools and apps inside of Braincube because our ambition is to enter into the architecture of the customer. So we need to adapt depending what they already have, as data collection as cybersecurity storage and things. So we have plenty of opportunity to adapt. We have a portfolio of solutions. Some are end users application. Some are middlewares. Some are tools connectors. And we have servers for the platform.

So what we do is we have a first experience with a customer, which is can we make a proof of concept or proof of value quickly? So we set up a pilot. And we dedicate a team on a first use case, and we tried to make it easy for everybody. And we know to this was Braincube because we can quickly set up things that will compensate the lack of sometimes of what the customer should have or didn't have.

So when we have this value demonstrated, and we have this excitement, where we go, okay, we need to build new use cases, and we need to roll out the first few use cases, then we have this use few days of discussion with IT, where we build the final architecture, and we build the rollout plan and the plan of qualification of the next use cases. So, Braincube was a single monolithic type of software, probably in 2008, 2009 and 2010, then it becomes 2, 3, 4, 5 software solution: one for life purposes, one for big data purposes. And today, it's not anymore. It's more you have to see that as services, servers, tools, connectors, and for sure the final application which can be middleware, or end user application.

So we have this probably very unique opportunity to adapt to any IT architecture situation. And then we can deliver what we know our apps and our frameworks for developing any use cases very, very quickly. That's what the customer enjoyed the most, the speed for delivering value.

Erik: And then this is, I suppose, it's sold as a SaaS offering maybe with some services around training and implementation, is that the underlying business model?

Laurent: So it was kind of subscription and PRO Services. But today, when we discuss with our customer, we understand that everybody has his own pace of transformation. So what they want is more and more a business model that helps to scale with their own expectations.

So today, you have some subscriptions in Braincube, you have subscription and platform separate from a package of apps. But we are having also pair instances type. That means if the guy is using OE solution that has been designed exactly according to what he wants, and he’s using two machines today, 20 machines next month's and 200 machine the months after, so we have this model where we can grow with them. So we have some pair users. We have some pair license. Also, we can have some profit sharing.

Imagine you set up an autonomous decision making on a complex process and you save a million a month on this. For us, it's an experience where we need to go, understand what needs to be done. So we need reductions coming there and saying you have the algorithm of learning. But when with this is achieved, then we deliver tremendous value. And so we can do some profit sharing. That means if the customer is not totally convinced at the beginning that he will work, we can come and go, show them what we can do. And if it's profitable, then they are happy.

So we have more and more offering that becomes scalable to make sure that the customer is always paying for what is really using. And that also was a change. Because before it was like any SaaS offering, we had one solution, one SaaS price according to few parameters like your size, the number of variables storage, whatever. And today, we are capable already adjusting the offer to the exact needs of the customer. And if they grow, then the system will grow with them but the costs are known. And if they back down a little bit because they sell a plant or whatever, their costs go down.

Erik: I mean, one of the most interesting things about IoT platforms is the visibility around the data, gives you the ability to be very creative in terms of the business model. But most companies are really not fully exploring the possibilities right now. They tend to align on one or two models, and stick to that. It's similar to this topic of optimization. So, people apply an average and they say, okay, in general, this model works on average best.

But like you said, you can apply the same principles of optimization to pricing and say, yes, okay, there's an average that works best, but we have 100 different customers, and there might be 100 different solutions for the right pricing model for those customers. It sounds like you're actually able to explore that problem space and be quite creative.

Laurent: I see you understand exactly what we want to achieve. That means we want this flexibility because it makes sense. Manufacturing is all about waste. You need to control your process, make sure you don't waste. You don't waste time of people. You don't waste anything you purchase. You don't waste opportunity of purchasing more blah, blah, blah.

So when you set up a system, and you say you will pay this fixed amount of price, and it’s all included, at the beginning that people are using a percentage of the system. And so they're frustrated about why I am supposed to pay more? We are this life today in this. Is this community today where we like to pay on demand, we like to pay per usage, we like to pay sometimes pair value. I mean, you give me the value, and I share the value.

So I think we understand this more and more. And we have a catalogue of offering where some of our solutions have a different pricing model. And most of them have one pricing model which is intended to be as scalable as possible, so to never waste any pennies of our customers. And some might have multiple pricing models, because it might be super different to be in the place where you have one large production line versus another plant where you 500 a similar equipment running in parallels.

So we understand this more and more. We work with our customer. We answers to a lot of requests for quotation. And yes, our technology as some advance points and people come to us because they want to have our 12-14 years of experience in what we do, yes, we have a team of guys that are coming from manufacturing and that really team up with the customer, they really understand their needs. But also, want to make sure that we grow as a partner inside their organization as the same speed they grow their transformation project.

Erik: Well, Lauren, you've built really a great company. And I can tell that you're still very much actively iterating and improving as you go. I think we've covered I'd say a good bit of the business here. Anything that we haven't touched on yet that's important for the listeners to understand?

Laurent: I think the most important things we learn about these 14 years is do anything really impressive, only thinking about technology. And it's all about taking the time to build a vision where you want to go. And from this vision, there will be different strategy to achieve the vision with many risk or not. And that's where they should start to call us and just say, okay, that's what I have. That's my dream. How do I go there?

When we start to discuss this with customers, the roadmap and the milestone we can set up are giving us really great achievements is cheaper, it's faster, everybody understand what we are doing. And it's also super flexible. During this, if your ambition change, then you are super excited about this: you can decide to go faster, bigger. Or if you think your people have a hard time to accept some of the newest technology that makes sense but there might be socially difficult to accept, you can step down a little bit, make some things that more easier for the people to accept, and try to make sure you won't lose anybody on this transformation journey.

So always start with the vision. And if you don't know how to build this vision, that's okay. We can spend time with you. We sell technology, but at the end, we achieve transformation. And that's we, I think the Braincube mission to transform industry to eliminate waste.

Erik: Laurent, I mean, that's a great closing touch. If there are some folks in our audience who are interested in maybe Braincube and exploring how they can work together, what's the best way for them to get in touch with your team?

Laurent: So we have organization in Europe, in North America and South America. You go on our website, and you have emails, you have telephone numbers, you contact us on LinkedIn, or wherever we are very active. We have a team of business representative that will answers to you. And anybody who wants to directly talk to me, talk to any member of our team, because the things we might be valuable in some thinking they have today, it's so easy with all this platform we have .So never hesitate, there's so many ways to contact us anyway.

Erik: Thank you for your time today.

Laurent: Thank you very much, Erik. It was pleasure.

Erik: Thanks for tuning into another edition of the IoT spotlight podcast. If you find these conversations valuable, please leave us a comment and a five-star review. And if you'd like to share your company's story or recommend the speaker, please email us at team@IoTone.com Finally, if you have an IoT research, strategy or training initiative that you would like to discuss, you can email me directly at erik.walenza@IoTone.com. Thank you.