Rockwell Automation, Inc. (ROK) Earnings Call Transcript & Summary

Hi, everyone, and thank you for joining today's webinar artificial intelligence and the digital thread for tire builders. Before we get started today, we do have a few housekeeping items to keep in mind. The audio for this event will be streaming through your computer speakers. So please make sure your volume is turned up and speakers are turned on. Our webinar platform performs best on Chrome and Firefox browsers. And on the lower left-hand side of the presentation, you will see a Q&A box. We encourage you to enter any questions you have throughout the presentation here, and we will answer them at the conclusion of the presentation. If you are having any trouble connecting to the webinar, please take a moment to refresh your browser and disconnect from your VPN. If you continue to have challenges, please clear your cache. We have instructions in the handout section of the webinar platform on how to do so. All of the panels within the webinar platform are adjustable to resize, simply click the corner to adjust or hit the maximize screen at the top right-hand corner of each panel. Today's event will be recorded and will be available immediately after it's completed. You can access the recording utilizing the same length that you use to access the live event. After the webinar, we will also be sending you an e-mail with the resources from today's event, including the slides, handouts and event recording. And with that, I would like to introduce today's speakers and turn it over to our presenters, Steve and Maddie.

Thanks, Elizabeth. So [indiscernible] is one of the trendiest terms in 2023. You might have been scared of it from Black Mirror on Netflix, that new season. You might be overly impressed with it, with ChatGPT or some of the tools out in the market today. But really, today, we want to talk about what is artificial intelligence and where are the real applications in the entire industry today? So we wanted to spell some of those lists and really get into how can you apply it in your business. The objective of today is to showcase the value of artificial intelligence and what we call the digital thread in the tire industry and really learn a practical way to operationalize it. So we'll start with introductory remarks, go into an overview of digital transformation, talk about artificial intelligence, machine learning, which is a subset of artificial intelligence and optimization case studies, so where we've actually applied this with clients. Talk about our approach and scaling methodology for value for scaling. And then we'll end with a question-and-answer. So as Elizabeth said, if you have any questions, please put into Q&A, and we'll get to it at the end, ideally. So my name is Steven Nguyen. I'm a global digital transformation specialist, specifically focusing on Tire. I'm an AI enthusiast. I think there are so many applications inside and outside of the business world. So I'm excited to talk to you today. But first, I wanted to hand it over to my colleague and co-presenter, Maddie Blanchard.

Thanks, Steve. Hi, everyone. My name is Maddie Blanchard, and I too, am a digital transformation specialist, but I am focused on helping our OEMs globally when it comes to digital transformation strategy, and how digital technologies can apply to their operations and their solutions. So throughout the duration of this, you'll see some poll questions come up throughout, and I believe we have one coming up right now. So how is Rockwell Automation currently serving you today? And I'll give you a few seconds to answer that. Some -- looks like we've got quite a mix of answers. Some see us as a vendor, others as preferred suppliers, and that's great. So I think it's important for -- to start us off by grounding us in our organization. So Rockwell is made up of 3 divisions that are split between intelligent devices, which is the foundation of operations, it's the hardware and the physical device component actually physically installed and then there's software control, which is where you're going to find those software platforms and operational support that enables data collection and visualization and data orchestration. And then lastly, there's life cycle services, which is really the glue kind of tying those pieces together in a way that best serves our organization, and that could be through digital consulting, data science and other supporting mechanisms. Now what makes us truly unique is that we also have an ecosystem of software and hardware partners that are supporting these 3 divisions. And ultimately, what this means for you is that we have the ability to support an open and integrated architecture, meaning that we can work with the investments you've already made into your business and operations installed base, whether that be Rockwell or a third party as well as tie those systems together to really enable that IT/OT integration. Now this shows a snapshot of some of those partners as well as acquisitions that we have made over the years that is fundamentally changing the way that Rockwell serves our clients to do more than talk about digital transformation, but actually deliver on it. To point out a few, we have Clarity who helps serve our clients in cybersecurity and thread detection, there's Emulate3D as a platform for discrete emulation. And most recently, there's Knowledge Lens, who is a group of about 600 plus data science and technologists out of India, really expanding our efforts to further support clients in regions like APAC and EMEA for data science. And lastly, there's Kalypso, who we're going to dive into deeper during our time together today. But Kalypso's, our digital transformation business arm of Rockwell, and it's how we deliver on complex digital projects. They sit within our Lifecycle Services division. And as you can see, the vision continues to grow, and Kalypso itself has been growing in about 3x revenue over the past 3 years. Now what's key here is that we are continuing to make organic and nonorganic growth to really make use of all that data that is out there today and show end-to-end value to you, ultimately positioning us to best serve you in the market. Kalypso as mentioned, acts as the digital transformation delivery arm for Rockwell by providing the digital services, the people, the manpower to discover, create, make and sell better products. And we're a matrix organization that goes to market by 3 industry verticals and across 4 service lines or capabilities. And this operating model really illustrates how we approach our projects. So across each project, we have consultants, data scientists, technologists and services to holistically support a solution and essentially support long-term success. So if we look a little bit deeper within consulting, this is where we're foundationally building the opportunity. We're prioritizing high-impact use cases. What are the impacts of those use cases to the organization? Let's manage the change process and the workflows that naturally come from project delivery and supporting things like sustainability objectives and business strategies as well. We view data science and AI as the biggest unlock to digital transformation. And it's where we're applying the analytics in the advanced computational methods to enable predictive maintenance, autonomous manufacturing and ultimately, the concepts of closed-loop optimization to a system, which we're going to touch on a little bit more today. Now our technologists are key to the process and having the knowledge and the expertise to broadly across various technologies, understand what's out there and what is best to use for delivering on a specific objective or a business outcome. And finally, we find that some of our clients either don't want to or don't have the resources to support projects post delivery. So we have flexible ways of maintaining and supporting through our managed services capability. Now we were established initially in 2004, but we were acquired by Rockwell in about -- in 2020, and we currently have a global footprint with resources located in North America, South America, Europe and Asia. And we have the ability to support clients everywhere. We're continuing to expand those efforts as well as mentioned with that recent acquisition of Knowledge Lens. And we also have innovation centers based out of Europe, Mexico and India. Now one of the core areas that we are seeing clients make investments in is the connected enterprise. We call this the digital thread, but it's the connection and leading of data and capabilities throughout the life cycle of a product and it extends across the entire tire manufacturers value.

Obviously, we have some technical issues.

[ Niraj ], can you still hear me? Okay. I'm going to keep going on then. So again, artificial intelligence can be applied at any of these different stages of the digital thread. So you have discover, create, make and sell. When people typically think of us, they see us in this manufacturing this make area, where we talk about areas like manufacturing operations, but really everything from product ideation influences manufacturing costs and other areas upstream. So when we talk about the digital thread, again, it's artificial intelligence applied throughout. So a question for the audience, and thanks for dealing with some technical difficulties with us. The question is what area of the digital thread do you see the most improvement opportunity for in your business? Discovery, which is really early product development, create product design and development, make, which is manufacturing and operations, or sell after products and services. Okay. Okay. So it looks like we got a mixed bag, and that's excellent. So we're really going to focus on manufacturing operations for the most part today. However, these same concepts that will teach you about artificial intelligence today, think of how that can apply to these other areas. And if you want to continue the conversation, we have case studies that we can share and discover and create as well to get you thinking. So it looks like Maddie's back, I'm going to hand it back over to Maddie to...

Maddie is back.

Thank you. Thank you all. Apologies about that. So it's great things all for those poll answers. So when we look at the connected enterprise, generally, we are looking at connecting every aspect of your business, which we bucket into 3 different areas. So one is the design and development. This is the smart practices to designing new products and how that can impact the factory before it actually reaches it. Now there's the factory, the connected factory. This is where we're optimizing the processes in place today through real-time data, machine learning, architecture integrations and labor optimization. And we're going to be focusing on the factory as we move throughout the duration of today. But I think another key component of this is further in connecting the factory is -- and connecting the workforce. Once those new processes are in place, not only do you want to ensure adoption, but you also want to enable those to be digitally connected through AR and VR for maintenance in first-time right. And improving maintenance procedures as well as that real-time data-driven decision making. I believe now, Steve will be talking to you about some of the high-impact use cases that we see in the tire industry.

Okay. So we have plenty of notable success stories. Again, it's higher, really enabling that connected enterprise that Maddie was just talking about. I want you to take these 5 and understand there are many different applications in your own business. But the first one is mixing process energy management. What we did here is we created dashboards for energy visibility for a client, allowing them to measure and truly optimize their energy usage. They had very aggressive sustainability targets, and this helped them to accomplish that. Again, in the mixing area, we did a process control optimization. So this was a very real use of artificial intelligence and data science in order to optimize the process control of the mixer and standardize rubber quality and also control mooney viscosity. The next one was MES implementation. We implemented for a client across 5 plants along for IoT, analytics and future artificial intelligence capabilities. So that was really a groundwork for artificial intelligence. Again, with another process control optimization, this time in tire building, we utilize machine learning to reduce nonoperational tires coming from bad tire splices. The final area is real-time asset performance monitoring and anomaly detection, also sometimes called predictive maintenance. So we did this on meeting voltage drives in the mixer, some Rockwell, some competitive. And we were able to first unlock asset visibility and then trend this for predictive maintenance of these drives preventing downtime before they happen. Many of these will cover in more detail with some case studies in a little bit. But I want to pause here and have the audience just really quick, think what seems to be the time thread between all of these. The keyword here is something called optimization. The way that we utilize artificial intelligence, the most within the tire industry, within other industries is something called closed-loop optimization. So I will dive into that very briefly but first, I have a question for you all. Very easy yes or no. You have a 50% chance of being right here. Are you currently using artificial intelligence in your operations today? Okay. Okay. So if you look at the results, we have quite a few nos, quite a few yeses. This is totally okay. That's what we're here for today to get you thinking about where you can start, where you can first apply it, right? So let's first talk about closed-loop optimization. And I want to take a little bit more time here because it's going to be a basis for a lot of the case studies that we'll cover. So with closed-loop optimization, if you see these 4 boxes at the bottom, you can almost think of these as stages of analytics, right? So in the observed area, you have what's called you unlock descriptive analytics. And this is a first step where you asked the question, do I have the sensors or measurements in place to be able to track or measure this target variable, whether that's your target splice length or the viscosity of your coming out of the mixer. The next area we move into is the infer area. And this is the first area that we really use artificial intelligence as the unlock and this is unlocking diagnostic analytics. So being able to diagnose an anomaly or an issue before -- or diagnose an issue when it happens. And then predictive is really, can I predict an anomaly or issue before it happens? So the way we train up that machine learning model is by creating a digital twin of the physical system -- and then we train up this machine learning model with the knowledge from our data scientists, your data scientists, your engineers and the historical observations taken from those sensors to train this model up, right? And as time goes, this model increases its fidelity, the more data we have, the cleaner the data, the stronger the model. So coming out of this prediction, you take it to the third step, which is the decision or the optimizer. That's prescriptive analytics, now that we can predict an anomaly or predict an issue before it happens, what are the prescribed parameters that we're going to want to enact to prevent that from happening altogether? So you can think of applications like preventative maintenance or we're going to talk about the mix of the tire building machine in a little bit. The final area of closing the loop is this act or actuator. This is where you can truly unlock autonomous manufacturing. So imagine these prescriptions, these prescribed parameters are automatically put in the control layer, closing that loop and making those adjustments in real time or the other way is to have an operator act upon those prescribed parameters. Either way, you're closing the loop of optimization and those changes that you make are again sensed and plugged back into the model and then the fidelity of that model increases over time, adding more benefits to the business. So this is the core concept, I really want to spend some time here because all these case studies that we talk about today, they all have elements of closed-loop optimization. Looking at where we've applied closed-loop optimization, really the biggest successes we've had are with clients in production scheduling, mixing, extrusion, tire building, curing and final inspection. And again, many of these started with being able to work with the client on. This is the target area we want to approach and being able to workshop and creating a proof of value on these. So now that we've talked so much about what we can do. Let's talk about how we've actually applied this in real-life case studies. Our first case study starts at the very beginning of the tire production process, the mixer. As many of you know, the mixer is one of the most complex areas but also one of the most important as all tire quality is affected with rubber quality. So this client, they were noticing inconsistency of their rubber quality, and they're also realizing that, that was impacting tire quality throughout. So our approach here was to develop a predictive analytics model to correlate mooney viscosity to some of the recipe inputs, the control variables and also ambient conditions. Things are often overlooks like temperature, humidity, et cetera, in order to not only predict viscosity, but also to control viscosity. So if you remember the closed-loop optimization example, not only have that predictive analytics, but prescriptive analytics to the control air, right? So with this, they were able to standardize, control their -- control their quality of rubber and have a better tire throughput. Over to you, Maddie.

So at its core, this is the same methodology from that mix in case study that Steve has shared, but this is applied to tire building machine. And it's really where we've had the most success today. Tire manufacturers, as many of you are probably aware of, are really challenged to meet significant throughput goals while maintaining uptime and producing the least amount of defects in scrap. This particular tire manufacturer wanted to see about a 35% reduction in nonoperational tires and produce about 85 more tires per machine per day. So we took the approach of applying machine learning model and an AI algorithm to do 2 things. One, be able to predict whether a bad splice is going to happen; and two, then be able to prescribe if a defect is going to occur, prescribe the parameter adjustments to proactively correct that bad splice from coming off the machine. This ended up resulting in about a 45% reduction in nonoperational tires and an increase in about 113 tires per machine per day. Once we took this and applied it at scale, we actually saw an overall increase of about 570,000 tires produced annually. Now what I want to go into a little bit more, though, is how the algorithm basically works? And this is like a look at how those things are kind of moving together, right, where we're pulling in the critical production data to start forming -- the critical production data to forming a tire space, which we found to be speed of the rotor, length of the splice, the pressure applied when you are forming those 2 elastomers together. And so we're taking and feeding in those critical parameters, the historical data of them, the real-time production data and simulated data into a machine learning model to firstly identify. One, based on these current pressure parameters am I going to produce a defective splice. Two, if I am about to create a defect, what adjustments do I need to make to be able to stop that from happening. And then finally, you want to have the model be able to confirm whether those adjustments were corrected and stopped the defect from occurring. And this can feed back into the control layer and essentially create that closed-loop optimization process on the machines. It also enables the operator to really worry less about having to guess process changes that they need to make reactively. Now this is something that we have seen a lot of success with in the industry. But we're also finding that at the rate, the tire industry is adopting digital transformation. This same concept can and should apply across the production process in other areas like mixing as you just heard, or extruding to achieve target weight and target dimensional accuracy and surface finish or we can look at it in curing for achieving optimal temperature or pressure for better product quality and performance. And I think the key here is that we see this as one of those highest-impact use cases for our OEMs because the intelligence is embedded on the machine. And this can result in new revenue streams for your business and differentiated equipment and solutions in the market that are addressing these very future trends through digital. So what I'm going to do is I'll let Steve actually dive into a little bit here of what exactly that architecture looks like. Steve?

Thanks, Maddie. And also, one more point on this tire building machine. There are a lot of people that had marked to know you're not using it. This has been -- probably the most successful area to start for many tire manufacturers. There's a large percentage that have already begun applying this in one of their facilities. So like Maddie said, I encourage you all to think about this one and how it could apply. So looking at this overall architecture, there are a couple of things I want to point out here with closed-loop optimization, and this applies for other areas as well, not just the tire building machine. One point is if we look at this predict model area, we built this model using a Logix AI module that plugs into the chassis of the controller. So it all existed right there at the edge with the controller. Not to overcomplicate things or confuse, but there's actually 2 models that are going on with this closed-loop optimization. So 2 areas of artificial intelligence. The first is your prediction and then the next is actually with prescriptive analytics, you actually use another model for creating those prescribed parameters. So again, that was done in Logix AI. At the very bottom, you have a supervisory layer. And what that does is that just visualizes what was going on in the tire building machine, for example, giving things like what are those prescribed parameters, giving you charts of improvements, even things like how much reduced nonoperational tires have you been able to achieve now with supervisory. So again, take that concept, apply it to any of these other closed-loop optimization examples we have. And this architecture can be flexible. It's really -- it's really a conversation that you have back-and-forth with whoever your digital partner is here. It's a question for the audience. Where are you currently deploying your operational analytics, if you are. Cloud, the Data warehouse, Edge or not applicable? Great. Wow. Okay. This is excellent, actually. So if we look at the results, we're seeing a mixed bag, right? We have some people saying Cloud, Data warehouse, Edge, not applicable. I would say there's really no wrong answer here on where to apply your analytics. If we look at the overall architecture, there's really 3 core areas of analytics that we already mentioned and gave into -- in the polling question. But the first is in the enterprise layer, so Cloud analytics, that's really to look at enterprise-wide data. So maybe you're doing analytics comparing plant to plant OEE, right? The next area, a layer down in the Data warehouse, maybe you're comparing line to line in your actual facility. Now in the Edge, this is where we have learned, you need to have your closed-loop optimization and your true artificial intelligence analytics because for these real-time decisions that are made with closed-loop optimization where you need to make these prescribed changes quickly before those measured conditions change then needs to be done at the Edge, so you have reduced latency, and you can make decisions quickly. So that is one of the key suggestions that we have for any company looking to do some sort of closed-loop optimization. Analytics in the Cloud is good, Data warehouse as well, but if you're going to make these real-time decisions on the asset, on the machine, in the control air, do it at the Edge. And then, Maddie, I know you had a comment here, too.

Yes. If I could just add, one of the things that we find is that it's very common to talk about what about IT and OT convergence. And we find that a lot of our clients want to do it, but don't necessarily know what that looks like. And so -- as you can see here, this architecture kind of displays like a high-level look of what those parts all look like, where the blue represents the OT side and the red representing the IT side. Back to you, Steve.

Thanks Maddie. In the tire industry, it's not surprising to anybody on the call, companies are aggressively pursuing sustainability goals and sustainability targets. One, very, I guess, simple way to approach simple in theory is to just minimize energy usage, right? But we can't just minimize energy usage if that's going to affect tire throughput or tire quality. So one way we've actually approached this is with that same tire building machine example, we built upon that machine learning model. So we realized, okay, with tire building machine, we already did a closed-loop optimization and created a machine learning model to optimize performance, right? Now can we build upon that model and add an energy usage as another function. So because we can already reduce nonoperational tires, how can we keep that going but also minimize the energy usage as we go. So what we did for this client is we created that predictive analytics model, and we help them to minimize their energy usage and allow them to apply the same learning across their facility and even globally. So this was a way that they were able to chip away at that very aggressive sustainability goal that they had for 2030. The final area of the tire production process that I want to talk about is final inspection and specifically computer vision and final inspection. However, if you have computer vision and other parts of tire production, same thing applies here. So at this client, they were using 2D and 3D sensors, so a great start. They had that descriptive analytics. But they were taking that information and they're manually looking at the images to try to figure out the defects and classify these defects. What we did for this client is we built an artificial model or machine learning model to actually detect these defects and then take a step further and classify them. Defects that might not even be noticed by the human per se. So miss cures, speed damages, cuts, excess rubber, poor appearance, all these different areas of defects were able to classify them. So this is changing the operator in final inspection from having to do this manually and guessing defects to giving them the right data at the right time and allowing them to be more informed in how they approach these defects. So again, an example of artificial intelligence augmenting the workforce here. Our final case study today. This is preventing medium voltage drive failures with machine learning, and this was specifically in the mixer area. Looking at medium-voltage drives that were vendor agnostic. So we had ABB escala and again, Rockwell PowerFlex drives. The client had critical drive failures at multiple plants, and this was costing them millions of dollars of first downtime and also lost production. Our approach was first evaluating where these issues are happening. But the issue was the client didn't have visibility into their assets and where this was happening. So going back to that closed-loop optimization example, the first unlock was descriptive analytics. We unlocked condition-based monitoring to see what is the visibility into the asset health, what is actually happening when a downtime event occurs. Once that data was unlocked, we then move to predictive analytics, where we could trend all that information we are now pulling from condition-based monitoring, and we could trend that to be able to predict these failures before they happen. So that's just one example of going from condition-based monitoring to predictive maintenance and unlocking that. In a future state, maybe we go prescriptive, maybe we go autonomous even, right? But these are just all different stages you can approach. The last and maybe the most important when it comes to projects like these was knowledge transfer. We worked with the team on site, the maintenance staff, the client in order to transfer this knowledge and give them full ownership of this project as they asked, right? So you have to think about the life cycle. Final, I think this is the final. Maybe we have one more after this. But what use case resonated the most with you here, today. We have mixing optimization, tire supplies, energy management, machine vision on final inspection, predictive maintenance on drive failures or other. Okay. Thank you all for participating. Again, we have a spread out application. And I think this is really important to point out because it depends on what area you may be focused on or if you overlook in the entire plan, an entire enterprise and maybe really client-to-client basis. Everyone has a different baseline on where improvements need to be made. So as you're going on, we're going to talk about really how you can get started with artificial intelligence and really figuring out if all of these seem interesting to you, what's the right place to start? So with that, we're going to cover our approach and scaling methodology, and I'm going to hand it off to Maddie.

Thanks, Steve. So yes, as we just kind of talked about areas that we've delivered, things that have been successful to organizations. And now I want to talk about how were we able to bring value to those organizations. So many companies have different philosophies on really how to approach digital transformation initiatives. And one of the pitfalls some companies fall in is really believing that you have to establish this digital foundation before really starting projects and having to select all of this technology and software and have everything in place and then only at that point where you start actually developing applications in the field. And the problem with this approach is that not only are you creating potential islands of technology, but you're also investing a lot of capital before seeing really any real payback. And so our philosophy and approach to delivering value is through value first -- our value for scaling methodology. This is where we're identifying, not proof of concepts, but proof of values. We call them minimal viable products or MVPs. And an MVP is a narrow and specific scope to an identified business problem with the intention to scale it in the future. And the goal here is for our clients is really to have them see the value within 8 to 12 weeks. Now to get there, we have to start with use case discovery, where we're really looking for and identifying the highest value use cases prioritizing those to your organization. And then we'll move into doing more of a technical assessment of the architecture, making sure the requirements for those priority use cases are there. I want to make sure that the right infrastructure is in place and start forming and confirming our hypothesis of the value. Now once we do confirm some of that foundational infrastructure knowledge, we then want to identify the size of the value through exploratory data analysis where we're really creating a strong and model, to start mapping out how we would start to approach a machine learning model. And then from there, we move into the developing of the MPP and testing that value hypothesis. And then finally, it's hardening the solution for full implementation at scale. Again, the value is to bring -- the goal is to bring value to your organization within 8 to 12 weeks. Now this is a look at how that use case exploration and prioritization really looks like, and it's the output of an activity we typically propose embarking on early in the digital transformation discussion, where we're looking at your digital strategy and collaboratively talking across parts of your organization around key focus areas and diving into barriers to eventually kind of map out the specific use cases and identify places that we think we can deliver on those or work together with you to lead. They're categorized as ease of implementation, what's the impact to the organization and what's the cost to implement. And then together, once those are identified, we want to work together to focus on the use cases of the highest benefit, least effort in the highest return to the organization. So Again, as you can see here, this was an output for a different client. But it really helps us get a good idea of where to start. And just note that this does differ depending on each plant's needs. And now I will hand it over to Steve to talk about what the road maps typically look like.

So we take that use case prioritization as a great first start, right? As Maddie said, we can first evaluate what data you have available, your architecture, do a use case prioritization. The next step that a lot of companies ask for is an actual road map, okay? We help to prioritize and rank these use cases. What does that look like over 1 year? What does that look like over 3, 5 years? That is something that working with a digital partner that has consulting but also implementation experience can help to provide. So right here, you have some examples, things that we cover today, predictive maintenance, machine vision, tire building machine optimization, one that we didn't cover, HVAC management. We have different layers of these projects. So maybe we spent some time looking at prerequisites, then we go into discovery, we go into developing the actual solution and then scale and support where we expanded across an enterprise, and we provide support or we give complete knowledge transfer and ownership to the company, right? Everything below the dotted line, maybe those are use cases that we see are very important top 8 of all use cases. However, in this phase, we're not going to cover it. So we'll table those for next year. We'll table those for 2 years, but we can still build out that road map. Okay. So I want to take us back to our final slide of the presentation about how we enable closed-loop optimization across the tire production process. As I mentioned earlier on this presentation, we really focused in on production in that make area today, but we've applied artificial intelligence and you can apply artificial intelligence in other areas of your business operations. So hopefully, this webinar helped you connect this term of artificial intelligence, this trendy, trendy word. It's a tool used to augment your workforce, it's a tool used to optimize your business operations. And although it's undeniably powerful, it still needs human intervention in order to determine where, how, where and how to use it and how to train it. So with that, I'm going to move to the final question of would you like us to follow up to discuss artificial applications in your business? If you don't want to answer right now, that's totally fine. We have a follow-up survey, but we will also provide our contacts, Maddie and mine. You are free to reach out, ask any questions. Again, I'm an artificial intelligence enthusiast, that sounds super nerdy, but I really do get excited about this technology and what it can do. Okay. So here are our contacts, you can take a screenshot and thank you so much for your time. I think we're moving to question and answer now Elizabeth. Is that correct?

Excellent. Yes. Thank you, Steve and Maddie. That was a great presentation. And we do have a few questions that have come in to the Q&A box. And for anyone that hasn't gotten their questions in just yet, you're able to submit them now and we can see if we have some time to get to them. So I'll go ahead and kick it off. The first question that came in is, can you speak to whether the calendaring optimization utilize motor or current power output in the process analysis?

Eric, so that was a great question. What I can say is -- and I'm not sure of all the specifics -- but typically, motor current empower is utilized in the initial correlation of variables and training up the model. So for example, in the mixer area, motor current and power output is actually a key input to the machine learning model. So I'm assuming for calendaring, knowing how energy intensive it is that would be an input as well. So I hope that answers your question, but that's also part of what we do prior to engaging in a proof of value is looking at when Maddie mentioned that Stroman model, that's what variables correlate with our target variable. Does motor current empower output correlate with this target variable? And can we build a machine learning model to correlate those? It's a great question, Eric.

Okay. Next question that came in is, what is the first step in enabling closed-loop optimization or autonomous manufacturing?

Let's see, Elizabeth, can you repeat that question?

Yes. What is the first step in enabling closed-loop optimization or autonomous manufacturing?

So if I'm hearing this correct, what's the first step really for closed-loop optimization. This is similar to -- if we think about the predictive maintenance use case, -- the first step is getting the data. So descriptive analytics and unlocking that. So again, that question of, do I have the right sensors in place in order to measure or identify these target variables? So unlocking the script of analytics, something like condition-based monitoring and looking at what are you trying to trend? What are you trying to track? What are you trying to analyze? Thanks for the question.

Next question we have is, will artificial intelligence take jobs?

So I'll take this, Steve. So actually -- and Steve had mentioned this, but AI is in manufacturing certainly going to augment the workforce, but it's here to help make people more efficient in their roles. So in the tire building example, we're providing the operators with the prescribed parameter changes. We're giving them visibility into the data. And really, the operators are the ones closing the loop. So they're still involved but we're really just making that machine more efficient from a throughputs point of view and reducing the amount of defects. Thanks, Elizabeth.

Makes sense. Yes, we have a couple more that have come in. So where do you see the most manufacturers starting to use AI?

I -- it really depends. So I would say we have done a lot of work with the tire building machine and controlling that target splice profile. That seems to be a very sustainable and very quick project that we've been able to deliver in 8 to 12 weeks typically. So that's a great starting place. However, it really -- again, it depends on the baseline of the client and -- we've also seen predictive maintenance being a very important and quick fix. So I would say those 2 are across the board pretty widely applied. But again, I'd encourage some sort of use case prioritization even if it's internal to figure out where are you going to have the highest return on investment and highest return on value for your company.

Yes. I'd add to that too. One, we're able to help you identify also those high-impact areas of where AI might make sense in the operations. But aside from tire building where we've seen a lot of uptick. When we're looking at this from the OEM perspective, we certainly see there being value working with our equipment builders on the mixing side of the coin or within calendaring or within extruding, curing is another great example. So really almost every point across the production process from an OEM standpoint is somewhere that we can help you. And that could be in the closed-loop optimization or that could just be getting started, right? Looking at the sensors that are going to go on there. And then what's that long-term goal? Is it predictive maintenance? Or is it fully autonomous operations?

Perfect. Okay. Another question that came in. We have an internal IT and data science team, what would be the benefit of working with you?

Yes, that's great. I'll take this one. So we've got decades of experience delivering digital into tire and it's really accelerated by a lot of our work with many of those top clients in the industry. So we're bringing in our industry experience to help augment the workforce and so by working with you, we could either be an extension of your team to start tackling some of those more difficult AI challenges that you haven't been able to get to yet, kind of expanding that capacity or we can work together with you on problems to work together with you to start solving challenges together. So if there's things on the docket that you're looking to go after we're here to help. But we don't want to target any projects that don't have a clear return on investment. So that's our priority focus.

Excellent. Sounds good. Well, with that, we are going to go ahead and close out our webinar for today. Thank you, Steve. Thank you, Maddie. This is a great presentation with a lot of good questions come in after the fact to you. And thank you, everyone, for attending our webinar today. There will be a post survey that pops up immediately at the end of this webinar. We do ask that if you have a chance that you could please take that for us. We're really working to keeping -- to improving our webinar topics and ensuring that we're providing value to you. So please participate in that survey if you have the chance to. And if you would like more information, if you would like to reach out to Steve or Maddie, you be contacted by Rockwell and Kalypso, you can also indicate that within that post survey as well. So thanks again for attending this webinar, and we look forward to seeing you at our next event.

Thank you all.

Thank you.