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

Hello, and thank you for joining today's webinar. Before we get started, we have a few housekeeping items. [Operator Instructions]. Today's event will be recorded and will be available immediately after it is completed. You can access the recording utilizing the same link 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. Additional information regarding today's topic can be found in the handout panel of the webinar platform. With that, I would like to introduce today's speakers.

All right. Thank you, Shannon. So let's kick it off. So good morning, afternoon and evening to all of you from whatever you are joining us on the globe. So welcome to this live webinar and the discussion on the topic of how to leverage artificial intelligence for sustainability in the semiconductor and high-tech industry to win in this new ESG economy. So my name is Anuj Mahendru, and I'm responsible for our global semi and high-tech business at Rockwell. Additionally, I do represent Rockwell semi's smart manufacturing chapters in North America and also had the opportunity to coach here some of the key industry events in the past such as the Global Smart Manufacturing Conference, as well as some smart manufacturing pavilion activities at SEMICON West. We are also a member of the Fab Owners Alliance. And today, on this call, I have the pleasure to facilitate this webinar and host a couple of my distinguished colleagues from Rockwell and Kalypso. So just for the clarity of our listeners, I would like to highlight that Kalypso is a digital consulting arm of Rockwell we acquired a few years back. With me on the call are Andrea Ruotolo, our Global Head of Sustainability Business; and Chris Calabro, who is the Senior Manager and semiconductor practice peer from Kalypso. So let's quickly go over the agenda of today's meeting, I'll -- as I said I'll kick it off and share the industry point of view and also talk about the problem statement. As I go around, talking to many of my peers in the industry and many of the trade forums. And next, Andrea will share our own sustainability goals and our own Rockwell's journey in the performance to the ESG initiatives. She'll also, during her course of the discussion, share some industry references and key use cases. And finally, Chris will drive us home and share some specific use cases where AI can be leveraged to solve sustainability-related challenges in the industry. To make this webinar interactive for the audience out there, you have the opportunity to participate and provide your answers to a few polling questions during the course of the webinar. And we'll have a live Q&A towards the end of the discussions today. You may also chat in your questions, as Shannon mentioned earlier. So with that, let's dive into the webinar here. So let's start off with the very first polling questions. So does your organization have a formal or informal commitment to the sustainability goal. So if you can take a few moments to really kind of put in your answers, we would really appreciate that. So it looks like 100% of the audience. It's still now. It's still coming in. It looks like we are getting almost 83% of the audience. Let's give it another 10 seconds. I think we have a decent spread of 80% to 20%. So 80% of you have responded that yes, we have some formal commitments and the 20% still are talking about informal. So thank you very much for your responses. So let's get into the my point of view portion of that. All right. So as we look at the industry, and again, folks, this is not new for many of you, we're in the midst of once-in-a-generational massive, stunning cycle. Here, obviously, in the U.S. with all the CHIPS Act and around the globe. There's almost 100-plus new fabs, which are being either already in flight or they have been announced all over the globe. And in addition to the major investment just in the semiconductor space, for Rockwell, I cover both the semiconductor and high tech. There is a bunch of hyperscalers when you talk about the data centers, and there's a big boom in that industry as well, and that falls into our high-tech segment of our business. And then you couple what's happening with regards to national security, the self reliance, the governments around the globe are putting in billions of dollars in terms of incentives which has been lately the catalyst behind all these mega projects. And when we start looking at what are the key drivers, many of you know that this industry is known for its cyclicality over the many years. But then as we look at the key drivers, which are really kind of fueling the growth, the sustainable growth in this industry, many of those drivers are secular. And you talk about some of those secular drivers things like 5G, you're looking at 6G, IoT, electric vehicles, digital transformation. And then we start really honing in on the topic of today's discussion, which is all around machine learning and AI. That's something which is going to stay with us for many, many years to come. So again, just to kind of level set the audience in terms of where we see the growth in the industry in the future. With that -- sorry, I'm trying to advance my slides here. Let's get into the problem statement area. So I talked about these investments on the other slide. And this is kind of a representation of a GHG flight map for all the electronic -- key electronic manufacturing, I would say, hubs in the country, here in North America. And which is kind of overlaid with the WRI's, World Resource Institute's, water stress map. It kind of shows you the intensity of all what you have going on. In terms of where these investments are happening in those water stressed areas and the drain of the resources. Just to give you a little more perspective, when you talk about a semiconductor fab and the amount of power -- and again, just to many of you, this is not new news, but the amount of power it takes to generate -- the generation of the power, which is needed to actually fabricate the semiconductors. You are looking at a massive, massive amount there. I was actually in Taiwan a few years back, and we were talking to some of the folks at TSMC, they almost draw close to 10% of the island's overall electricity. So they're looking at ways how they can kind of get 1 or 2 points of energy efficiency, so they can get really approvals from the government to build the next fab on the island. So -- and the water usage, you look at Arizona and you start thinking about all the new fabs, which are going into Arizona and even in Texas. Again, there is a massive, massive strain on the resources as we look at what's going on in the industry and the resources they need. As we talk more about when we look at from a driver's perspective and the challenges this industry has. Obviously, there is so much happening in the industry in terms of the new chips which are getting produced. The size of the chips are getting smaller and smaller. You hear about TSMC's latest 2-nanometer node, which they are producing, obviously not at scale. But the amount of again resources, resources -- natural resources, which goes into to just create and manufacture and process the chips is that overall energy intensity on the chips is just monumentally rising. And that's a major concern when you start looking at the amount of greenhouse gases, which those are producing. And when we start looking at the pressures which the industry is based off, whether it is the internal pressures or the external pressures from the stakeholders, from the investor communities, from the financers of these big mega projects. Each one of those stakeholders are looking at what are these companies actually going to do in terms of sustainability and as you pick up the reports -- the annual reports of any of the big companies, each and every one of them has lofty, lofty sustainability goals to get to certain net-zero content by 2030, 2025. To some, it goes all the way up to 2050. But again, the focus around this topic, this is no longer just a topic around yes, it's going to be a check box. It is becoming a more and more of a business imperative for the industry. And it is important to look at how they can solve this problem. Yes, there are ways they are doing -- they're taking steps to date, the more conventional ways, the way they are taking steps today to address these issues but it has to really go above those normal conventional ways and look at what are the disruptive technologies, which can really help solve the problem. So with that, I'll just cite a couple of examples of some of the studies from -- again, from the internet and it talks about the overall impact of how artificial intelligence and the impact of artificial intelligence is going to be on the overall footprint of the energy usage and the draw -- and look at from a standpoint of how much actually in terms of value generation, when we talk about real dollar value generation, what does that mean when AI is leveraged in the industry for solving the sustainability challenge. Lastly, I would talk about where are the issues really in the fabs when we talk about sustainability and we talk about the scopes -- the different scopes in the sustainability. The Scope 1, the Scope 2, the Scope 3, as you may heard of. Majority of the real problem, which comes into play is in the Scope 1 and the Scope 2 areas. When you look at the Scope 2 areas, we focus around where it is -- where the issues are, obviously, there is -- there's a whole lot of process tools in that clean room, right? They are creating a lot, they need a lot of energy to operate, and there is a lot of emissions which are coming out because there's a lot of gases -- processed gases which are being used on those tools as well. And then you start looking at the facilities. Again, you look at the big assets, whether it's the chillers, the compressors, in some cases, the boilers. All of those are big energy assets -- energy hogs and that all of them are producing emissions. And again, the challenge remains to the industry is, "Okay, what do we do to solve some of those problems" because they are worried about how -- if we try to apply some of these disruptive technologies, whether it's AI or ML, what is it really going to do to my process? What is it going to do to the quality of what I'm preparing or manufacturing or processing. And those concerns, as we talk to many of our peers in the industries like should we apply -- should we be applying AI into some of those areas because in their minds, maybe we should take an area which is not going to impact the quality, the ultimate yield on those semiconductor wafers, maybe we should look at some of those areas in the past, whether it is a subfab area, where there is a little less of an intrusion before we actually go into the clean room. So a lot of focus again in terms of how to decarbonize stuff. So today's discussion is really going to focus from an AI perspective and a use-case perspective, more around Scope 2 portion of the sustainability. So with that, I'm going to call my colleague, Andrea on the call here. And before we actually get into that discussion, I'd like to maybe address -- maybe elephant in the room, some of the -- some of you guys maybe in the [ audience ] may be thinking, we're talking about AI, we are talking about sustainability. It doesn't people feel like an oxymoron. You have to -- when you were looking at AI -- to deploy AI, you need -- model it, you need a lot of resources. You need data centers, you need supercomputers like more higher computing power. All of that is [ a bane ] in the industry. So Andrea, maybe for the clarity of our listeners. Maybe you want to talk to you -- I'm sure you have been able to discuss this question before, how would you kind of share your thoughts and your point of view that AI and sustainability of acting as an oxymoron. Andrea?

Thank you, Anuj. Yes, that's a very good question. And I'd say we have more questions than answers at this stage on how are we going to address that. One thing I can confidently say is that if we utilize AI without accounting for ESG, environmental, social and governance metrics, we run the risk of shaping an unsustainable future. So for example, from an environmental impact, AI technologies can have a significant carbon footprint due to the computational resources that they require. Neglecting environmental aspect of the use of AI within an ESG context can lead to increased carbon emission. So if we are using AI for sustainability, then we need to take into consideration actually the metrics that account for the carbon footprint of that utilization of AI for sustainability and really look at the overall net of carbon emissions. Is AI contributing more than it's actually -- it's contributing more in reducing the current emissions than it's creating those carbon emissions. Well, there's just one consideration under the environmental pillar. Now let's look at social equity, for example. If AI systems are developed without considering social factors, they might reinforce biases that inequality is present in society. For instance, biased AI algorithm can lead to discriminatory outcomes. So the call here is for everyone in this room and leaders when we think about using AI for sustainability that we actually look at the use of responsible AI for sustainability. So this question has been in my mind for a while, and that's why we have recently launched the first leadership working group in the world of looking at integrating responsible AI leadership -- responsible AI performance metrics within the ESG framework. We believe that, that might be the way to most effectively address this concern that Anuj brought up. So if you're interested in knowing more about this leadership group let us know in the chat, we know you can connect with me on LinkedIn, and I will be happy to provide more information. But clearly, because it is the first leadership group in the world addressing this question we still don't have the answers. But at least we have posed the question, and we will be working through these challenges. Now moving on to where -- as a company where Rockwell is at, we have a commitment to enabling sustainability. And this is from the cover of our 2022 ESG company report which was released in February of this year. The report is available on our website, and it shows more in detail the progress initiatives and goals in these 3 areas. Today, I will be walking through some of the key highlights. This is a summary of our own carbon footprint we enhance our sustainability analysis capabilities in 2022. We expanded our sustainability analysis and reporting capabilities with the implementation of a new technology platform that combines sustainability targets with data analytics for clear use of today and the future. We reevaluated our environmental emissions reporting to that reflect today's regulatory concerns, customer expectations and also shift from a U.S.-centric to a global approach. And then on Scope 3 emissions, we have an SBT, sales-based target, goal setting, which is selling our product environmental impact and that is, I think the journey most of you might be on. This need of connecting with your providers and understanding what is the reparability, reuse, recycle, product longevity, energy conservation, sustainability packaging and sustainable packaging and end-of-life management of the product that you are buying to run your operations. We have a long standing history as a company, focused on the pillars of ESG and here are some of our awards and commitments. I wanted to highlight that last year, we raised our rating to gold for EcoVadis, which is provider of business sustainability ratings with a global network of more than 800,000 rated companies, which places us in the top 5% of all the companies reviewed. So that was a brief overview of what are we doing as a company. Now I wanted to share with you what are we doing externally, how are we helping our customers. And we are helping our customers to achieve sustainability in 3 areas by expanding insights, expanding impact and expanding innovation. The insights pillar is focused on sustainable manufacturing, in which we help manufacturers, utilities providers and producers to reduce their carbon footprint by operating more sustainably and productively by using automation technologies to provide the data. This is all data-driven. To provide -- you need those real-time operational data points to get the insights to find opportunities to optimize in the areas of energy, of water and waste. Then for expanding impact, these are solutions and partnerships that we have in place to help scale the impact across value chains. By providing solutions that bring more visibility around product life cycle management. And then the third pillar, innovation, we are expanding innovation by helping accelerate the energy transition and circular economy by helping OEMs and climate tech startups scale new green production and green innovation. I will bring up some of the key studies and success stories here in a few more slides. So, this is about our official partnership with the World Economic Forum. We are collaborating with industry peers, academia and the civil society to shape the agenda to accelerate the decarbonization of manufacturing supply chain. Also, WEF has recently launched the Industry Net Zero Accelerator initiative in collaboration with other few partners like Rockwell. And I think that's a fantastic initiative because it's looking to establish a cross-industry space for executives in the manufacturing ecosystem to share knowledge and best practices and support the adoption of innovation at an accelerated pace so that we can reach net-zero targets. And here, you will see there are 2 white papers that were released, where you can find what have been some of the technologies and approaches implemented -- initiatives implemented to help corporations achieve their natural goals. So this is a great way to learn from others and probably into the future. If you have any questions about the case studies or about these reports just please reach out, and we'll be happy to connect you with the right people. This is basically -- we are highlighting here some solutions that are already operational. We have solutions around emissions, around energy, around water and around waste. And just very high level. We also mentioned the solution from Kalypso, but Chris will walk you through those solutions in more detail later on. But on energy, we have -- we are launching, actually, if you join us at Automation Fair, you will see our real-time energy information system to provide visibility into energy usage, storage and operations from machine level to enterprise level in order to better understand, manage and report on -- and benchmark the performance of energy consuming processes. I wanted to highlight that one in particular, because now corporations are being regulated and asked to report on their Scope 1 and 2 emissions. So I think this is a solution that if you have not implemented already a system, it would be worth exploring. We also have solutions, as I mentioned, on water and waste. Just on water, I wanted to highlight that we are leveraging AI to ensure water quality, reduced energy and chemical usage, and I will be presenting one of those case studies just very briefly in the next few slides. So diving deeper into our smart energy management, this refers to an approach to managing energy consumption, which involves collecting usage data from industrial automation systems, then monitoring, creating production contacts and analyzing the data for insights and using those insights to drive energy efficiency, optimization and emissions reductions and support reporting needs. Doing this efficiently at scale requires digital solutions that automate and centralize these steps into a cohesive, connected and real-time process. So just highlighting some of the approaches we have for smart energy management, including, for example, smart objects, which is a technology which is available for use in Rockwell Automation's control systems, our control -- Logix controllers. Today that's already available and the technology models data into standard information models and formats and create contacts between energy and production data from disparate sources connected to their Rockwell controller. This data is then made available for ingestion into edge gateway of databases and other IT application servers. And so this is one of the examples where we are bringing in data from the edge and being able to transfer that data up to the cloud, where then you can aggregate, analyze it bringing the insights and also report on your sustainability progress. This is a smart water overview. Our smart water management refers to an approach to managing water ecosystems in industrial operation. This involves gaining visibility into water data and water-related energy data from industrial automation systems, monitoring and analyzing that data for insights and you've seen those insights to optimize water-related production processes and support, again, reporting needs. I will talk about some of these case study where we use artificial intelligence very soon. This slide is highlighting the strong relationship across the water, wastewater industry, but also with a water conservation effort. We have a few different partnerships where our technologies are used to explore new innovations that can address water sourcing issue. For example, with West Virginia University, they are using robot technologies in a research lab that is taking acid mine drainage for the -- for recovering rare-earth elements and critical minerals cleaning and restoring the water. Then with Severn Trent in the U.K., they are leveraging Rockwell's technical expertise here to leverage AI for autonomy in waste catchments to reduce flooding and pollution. Cisco is Rockwell's tech partner, where we are codeveloping a digital water solution. And for companies to be able to reduce energy and chemical costs associated with pumping and treating water. If you join us at Automation Fair, you will be able to see the great application that we are codeveloping with Cisco. So this is a case study -- so these oil and gas producer used model predictive control on 35 pumps and 3 pool transfer units in past to reduce the instability in water injection operation. The increase in water injection exceeded all expectations, reaching 2.4%, which is equivalent to an increase in injection of close to 36,000 barrels of water per day. So this water injection meant an additional about 550 barrels of oil per day, which could be expected from this site. So in addition to this, the company achieved greater stability and operational continuity and was able to eliminate losses associated with the no pool levels and no suction pressure in the past. They were also able to reduce energy consumption and lower CO2 emissions. This is a great example of optimizing production and sustainability. You will notice, if you go to our sustainability report that we talk about not only increasing sustainability but increasing productivity and sustainability at the same time, which is the approach that will help you achieve your return on investment expectations, but also your return on sustainability goals. This is an example with the California water and wastewater treatment plant, which used an artificial intelligence and machine learning loop tuning software from Rockwell to create control -- closed-loop control processes on top of their existing control infrastructure to fine-tune these processes' parameters, which resulted in significant energy savings, CO2 emissions reductions and also the reduction of chemical use in their process. Multiple sustainability wins with the implementation of one software application within their existing operation. And then finally, how are we helping to support the circular economy. In this case, we have partnered with the Royal Mint to build a facility to safely recover valuable metals from U.K. sourced waste electronics. In this case, the waste electrical equipment, along with electronics and collectively known as e-waste is one of the fastest-growing waste streams in the U.K. and the world with less than 20% actually currently being recycling globally. Actually, in the U.K., they are throwing away over 300,000 tons of electrical items each year. In this case, the Royal Mint is using Rockwell's technology to recover over 99% of gold and other valuable materials from electronic waste. And so those were some of the examples where we are helping to reduce energy, water and waste. And with that, we will go to Anuj, if you would like to walk us through the next polling question, please.

Thank you. Thank you, Andrea. Great details in terms of those -- some of those use cases and really when we talked about energy in the fab, I talked about the resources in the fab, as I earlier touched upon, where they are kind of. So maybe the question for the audience, where are some of your -- where are you actually -- your organization is really focus in, in each of your organizations, especially when it comes to semiconductor fabs, in the subfabs, or in the clean rooms or on the both, right? So just trying to understand, from your perspective, where your focus has been. So these [indiscernible] responses. Okay. I see a mixed response there. We'll give another 10 seconds. All right, I think we have the responses, again, it's a mixed bag. So we see that some of you are focusing more primarily around -- it's kind of an even mix between the subfab and clean room operations. And again, I see a lot many who are saying that we do not -- maybe it's because of you don't have a fab where you're working in. So we understand the responses there. So I think this is a great segue to the next and the final portion of the presentation because we talked about subfabing facilities and clean room operations. So Chris will kind of take you through a couple of those use cases specifically, again, leveraging AI in those two areas. So Chris, do you want to take it on?

Thank you, Anuj. Okay. Screen seems to be frozen here. Well, while this is figuring itself out, we're going to walk through a few examples around sustainability in manufacturing. Manufacturing is right for sustainability improvements, whether it be through traditional energy efficiency, focusing on wages, whether it's implementing renewable energy -- All right. Implementing renewable energy or reducing yield loss and yield improvement, in particular, is an area that is very well suited for artificial intelligence and machine learning as these tools can analyze vast amounts of data to identify trends, highlight key parameters that affect yield, predict equipment failure, optimize processes, et cetera, et cetera. So with that in mind, let's take a look at some real-world examples here. So the first one is about chilled water. So for those that may not be familiar with water chillers, in it's simplest form, a water chiller uses water to remove heat from the system and is typically the preferred type of chiller for that. The heated water is then returned to the chiller to be cooled before being circulated back through the system. So the goal of this project was to determine if model predictive control could be used to optimize the chiller load while maintaining the proper water cooling in order to minimize the overall energy consumption required. So the chilled water system includes many, many components to the chiller, the chilled water pumps, cooling water pumps, cooling tower, heat recovery and other components. And there can be upwards of 10,000 parameters that can affect energy consumption. So the first step in the process was to build a model using training data containing lots of different variables, things like ambient temperature, humidity level, the age of the chiller, the overall health of the chiller, many, many others. And then these parameters were analyzed to identify the key parameters related to the energy efficiency of the system. And then once these parameters were identified, the model needed to be trained on how modifying them had an impact on efficiency. So the model recommends changes to be made and then actual results were compared to predicted results. And so many, many, many training runs were conducted with the results of each being fed back into the model to further refine results. So over time, as the model continues to improve. Now you can look at not just a single chiller, but now you have the ability to look at multiple chillers and you can start to compare one chiller's performance to another and begin to determine maintenance schedules, for example, or other life cycle activities that may need to occur. So you can see on the screen here, the outcomes of this project, and I think this is a great example of how AI and ML can be used to improve overall efficiency and positively impact sustainability. So our next example, so if you haven't worked in a wafer fab, it's a very complex process involving over 1,000 steps to manufacture a wafer. And depending on which step in the process, your lot of wafers is at, the wafers may be processed individually or as a batch. And so obviously, the further a wafer progresses through its process, the more valuable it becomes and the cost of scrap and rework increase. And given the current market dynamics where demand exceeds supply, it's more important than ever to produce as many Tier 1 wafers as possible and reduce the amount of rework and scrap requirements. And so one of the main things that drives quality and throughput is machine uptime. So the longer I can have a machine producing quality products, the more wafers I can produce and the better the utilization of resources like electricity and water. Conversely, if my machine starts producing out-of-spec products and I have to perform rework or scrap the wafers or take the machine off-line to clean or repair it, not only has throughput gone down, but I'm still not as efficiently using my resources. So in this example, we're talking about diffusion. So diffusion is a batch process, so all the wafers are process together, where material is deposited to uniform thickness on the surface of all the wafers in the batch. And this process typically takes between 3 and 8 hours to run per batch. Now you can almost never achieve your theoretical throughput due to many factors, right? Maybe not all of the zones on your machine can be used because they're getting dirty or they need to be cleaned. There's differences in chemistries from batch to batch. And so you have to clean the tool between recipes to prevent unwanted chemical reactions. And there's many other reasons why theoretical efficiencies are never achieved. So any time a cleaning has to be done or a process changeover is required, the tool must be requalified. And so requalification is done by running qualification lots. And these qualification lots take the same amount of time to run as the production line. And it typically takes around three qualification lots to get the tool back online and production ready. So this can mean that you could have a tool that is not producing production material for an entire 24-hour day in some cases. So the goal of this project was to develop a model to describe the relationship between different recipes that are going to be run on the same tool. So if one could understand the effects of recipe A on recipe B or vice versa, then you could pre-adjust the tool settings to allow bringing things into spec more quickly, therefore, allowing for shorter qualification cycles. And so by bringing the tool back to production status is quicker, you have reduced the downtime and increased the throughput while also using less resources to produce quality wafers. And these resources include everything, electricity, raw materials and even to [ cooling capital ]. So the result of this work was that for the recipe study, the productive time of the machine was increased by about 16 hours per cleaning, which is obviously significant. So this meant that you could go from recipe to recipe potentially without needing to do a cleaning cycle and reduce the qual runs from around 3 to 5 down to 1 or 2. So a fairly significant improvement in throughput. And then there's future applications for this as well. You could not -- you could look at not just the recipes, but you could you could gain a deeper understanding into the chemistry of the recipes and then which would allow you to perhaps restrict fewer zones in your tool and use more of the tool's capacity even as it nears its preventive cleaning cycle. Right. And our last example here. So the previous example, we discussed how to optimize planned downtime for things like changeovers and cleaning. And as Mike Tyson once famously said, "Everyone has a plan until they get punched in the face." Now no one is literally getting punched in the face in the fab. But the point here is that unexpected events happen and things don't always go according to plan. So maybe there was an issue with machine A during the shift and you couldn't get to do preventive maintenance on machine B. And now it's over into the next shift to deal with. So decisions have to be made. How long can I run the machine B before I have to do that maintenance? What are the risks associated with this? So if I'm having to move planned maintenance activities, which we know are critical in ensuring that we produce quality wafers, if I can develop models to predict when a machine will fail based on more than just first-order effect. Now I have a lot more information to help me make my decisions. So I'll be armed with the information based on data to identify the level of risk, allowing me to know when to postpone maintenance and when not to postpone maintenance. And knowing this also allows modifications to the overall production schedule to further optimize throughput. So again, by using data analytics and machine learning, overall efficiency can be improved in this case by predicting machine failures and minimizing unplanned downtime. And these -- again, these games are across the board. Gas, water, electricity, raw materials, human capital. So with that, we will open it up for questions.

Before we go there, Chris, there is one last polling question and thank you, Chris, for walking us trough some of those specific industry examples and use cases. And hopefully -- to the audience between Andrea's use cases and the references and between Chris' discussion, hopefully, you guys found some interesting nuggets to think about and would love the opportunity as Rockwell around the globe that we interact with you guys in different fabs or fields and facilities of yours to look at where your initiatives and your goals are. And certainly, try to align and figure out ways of how we can provide some of the solutions to really drive some business outcomes in the sustainability space. So let's look at the responses for the last question, only question there. All right, so it still looks like people are putting in their responses. Well, so here's the responses. So some of the people that -- most of the people on the call are not exploring the -- currently the use cases. But again, this has hopefully given you some proof of thought to look at why you should be looking at AI and how it can help you in the future. And certainly, we look forward to interacting with you on those. With that, let's get to the Q&A portion of it.

[Operator Instructions] First question, what represents the most significant obstacle in utilizing AI to drive progresses in sustainability? Andrea, do you want to take that?

Yes. I will be happy to take that. To start with my -- with the answer here. I view -- there are multiple obstacles today. One of them is that data is in silos and the quality of that data is questionable. And so one of the key barriers is to have to break those silos to bring in, especially when you want to drive sustainability, you are not looking just at one metric. It has to be the change of that metric within the overall ESG, environmental, social and governance framework. And so with the creation of the sustainability roles, now you would have leaders that -- looking to bringing ESG data points to then do a comprehensive reporting on the state of your corporation sustainability. So part of this reporting process will help to remove some of those barriers, will help to break down some of those silos. But today, we are at the early stages of removing the data silos. And then another big challenge for the use of AI for sustainability is that we don't have many professionals in the world that know about both worlds. We have professionals that know about sustainability, we have people that know about AI, but very few actually understand how the 2 connect together. And so what I brought up earlier today, this leadership group is also looking at cross-pollinating, to have leaders from sustainability, to have leaders on AI and helping people to upskill and to be able to understand the other side and to connect the dots. I don't know if Anuj or Chris would like to add your thoughts here.

Absolutely, Andrea. I think I'll add a couple of thoughts. You are spot on your commentary there. But as you look at the industry and you mentioned about the silos and you talked about the difference in different experts, right? So it's all kind of bringing from a semiconductor perspective. It is an intensive process. When we're talking about that many steps, which takes months to create a chip on a wafer, it's the subject-matter expertise of that process, understanding how the interaction of some of these chemicals and gases and the further process elements can really impact the full yield and without -- quality of the wafers. And what can be bought either up or down in that space and to actually address the sustainability directives and the goals of the company. That's a mix, which really people have to think about when they bring these things together, yes. So that is an [ up stake ] and that's why I said earlier, people need to think about where they can start applying sustainability use cases or the use cases in certain areas of the fab and look at those as Kalypso is saying -- I'm seeing a little bit of Chris's thunder around MVP, the minimal viable projects, right? And where are those projects, whether they actually show the return on investment and then try to scale it into other parts of the enterprise. Chris you have any comments?

Yes, I would second Andrea's point about data, right? I mean the key to AI and ML is the data. And like Andrea said, it tends to be siloed. And so getting your arms around that data is foundational to the success of an AI or ML or MPC program. So I would second her point about figuring out how you get all that data into a data lake or whatever your architecture looks like to be able to be used by these AI and ML tools.

Wonderful. We got time for a few more questions. What strategies could be used to spend -- to speed up the progress towards enhancing the sustainability of the semiconductor industry?

I can kick it off and maybe Chris, you can take it from there. I think we talked a little bit already about that from a strategy standpoint, looking at specific areas, again, in the fab, where there is lesser interaction to the actual process. Looking at the facilities aspect of it, looking at just a filing management aspect of it because these fabs have gigantic campuses and what could be done to optimize the usage of certain elements and certain energy-intensive assets of how they are operating in the buildings in those facilities that could be an area for -- to start up. So those would be the areas where I would say we start looking at putting in a strategy because you can see some immediate returns in some of those areas. And then you -- as you build your confidence. And as the earlier speakers talked about the silos and the experts in those silos that are trying to come together to solve the bigger problem, whether it's in the clean room and the usage of some of those expensive chemicals and gases and how those expensive gases actually gets released out into the atmosphere, the scrubbing of those and the overall emissions aspect of it, applying some of those technologies of AI into those areas would be the next step. Chris?

Yes, I would -- I agree with what you said there, Anuj. And I would say one strategy that we like is -- this can be overwhelming, right? Like you said, there's hundreds and thousands of processes and steps and inputs into making a wafer. And so it can be overwhelming. Where do I start? So one approach that we like is don't try and boil the ocean, right? Let's start small, pick an area whether it's a particular pain point or a particular area that does have good data and start small, right? Go and get that MVP like a new set, focus on that minimum viable product, let's get some quick wins and let's show the value of making this investment into AI through quick wins, right? And that is a very good strategy to help get the sponsorship, get the funding needed from executives to move initiatives like this forward.

Yes. And I'd like to add to that I totally agree that it's a complex and multifaceted challenge. And one approach that corporations initially took was to create this estimation and use equation to actually predict and then report under sustainability progress. And that would just not work in the short and medium term because both investors and regulators would be expecting the corporations actually report on real operational data that can be traced back to the source. That would be actually auditable, sustainable sustainability data. And corporations that are taking that approach from the get-go are what is called digital -- twin transformers, that they are bringing together sustainability and digital transformation at the same time. So corporations that actually look at bringing in real operational data into their reporting are these twin transformers that are showing to be outperforming their peers by 3x. They will give you market event -- market competitive advantage, but it will also give you the positioning in front of your regulators and investors that you can trace back those statements to this source of that data, to the source of that operational data. And I think that that's really the way to go.

I think I would also add, Andrea to your point, lastly, you touched earlier on my question around AI and sustainability. So I'll almost flip it just say sustainable AI, right? And the companies were kind of starting in that direction and showing the sustainable AI aspect of AI, right? That would be something which would be kind of used as a poster child for the others to even follow and look at those very specific use cases, which really drive some meaningful value back to their goals and initiatives.

Yes.

Great. Time for one more question. To what extent is sustainability a critical factor within the semiconductor industry?

Great question again, Shannon. I think I have touched basis on some of the reasonings earlier in the discussion. One more point which comes to mind, this is more tied to North America and especially tied to these mega projects, which are going in as a result of the CHIPS Act, which is currently going on in terms of their application process. So if you look at what Department of Commerce have even stated in some of their tenets and recommendations for some of these mega projects for the customers and for the clients who are actually trying to go after the incentives under the CHIPS Act. One of the key tenets is around sustainable operations, right? How do you ensure to the -- how can you prove in your application process that you are going to be a more of a sustainable enterprise as you put these mega plants on U.S. soil. So again, thinking upfront for the companies to -- whether these are the EPC companies, they are working with to develop these [indiscernible] project and keeping again, sustainability front and center in their mind because sometimes the sustainability projects, they tend to be -- people think then I have an existing plan, I need to modernize a certain asset because I need to hit a certain goal on that certain critical asset, which is drawing a certain amount of energy as an example. But as you look at these greenfields and you look at the backbone and how you are developing the overall infrastructure, I think this is something which is crucial, again, factor for industries who a, highlight some of those policies or some of the methodologies when it comes to implementing some of the strategies which we talked about on the call, and the infrastructure which is going on is able to take on those strategies and able to leverage those strategies to actually drive, again, that core sustainable enterprise piece. I think this will stand out on their applications even as the Department of Commerce look at, some of the things and the recommendations they have laid out. So that's one area I wanted to add on, on top of the other things, which I highlighted, why it is crucial, obviously, because of the intensity of what it takes for -- as we build these fabs in the different areas in the country and around the globe.

And I'd like to add that sustainability is not just a nice to have in the semiconductor industry. It's a necessity for the long-term viability of the industry and the well-being of the planet, it also -- so for example, it will have to be done from a risk-management perspective for example, because unsustainable practices can lead to supply chain disruptions, resource shortages and increase operational risk. So actually adopting sustainable practices is a way to manage these risks and build resilience into operation. That's another benefit of actually integrating sustainability into your decision-making, operationalizing sustainability and creating this change. But, the time is now. There's regulatory compliance and we would be happy to help and guide you through that process.

Very well said, Andrea.

Wonderful. And with that, we will end thank you to all of our presenters for a wonderful presentation. If you asked a question and we did not get to it or if you have any other questions, we can follow up after the webinar. Please stick around to fill out a survey that we'll pass around after this webinar. And if you would like to be contacted or like any additional information, please click that contact me check so that somebody can follow up with you. And with that, we'll end for today. Thank you all again so much, and have a wonderful afternoon.

Thank you.