Cognex Corporation (CGNX) Earnings Call Transcript & Summary

Greetings, and welcome to Cognex Corporation acquisition of RealSense Conference Call. [Operator Instructions]. As a reminder, this conference is being recorded. It is now my pleasure to introduce your host, Greer Aviv, Head of Investor Relations. Thank you. Please go ahead.

Thank you, operator. Good morning, everyone, and thank you for joining us. Earlier today, we announced that we entered into a definitive agreement to acquire RealSense, the current market leader in depth sensing cameras and vision technology for robotic perception and physical AI. With me on today's call are Matt Moschner, our CEO; and Dennis Fher, our CFO. Matt will begin by discussing the strategic rationale for the acquisition, the RealSense technology platform and the growth opportunity we see in robotic perception. Dennis will then review RealSense financial profile and the key terms of the transaction. A presentation accompanying today's call is available on the Investor Relations section of our website. Before we begin, I would like to remind you that certain statements made during this call and in the accompanying presentation are forward-looking statements. Our actual results may differ from our projections due to the risks and uncertainties that are described in our SEC filings, including our most recent Form 10-K. With that, I'll turn the call over to Matt.

Thanks, Greer. Good morning, everyone, and thank you for joining us. Today marks an exciting step forward for Cognex as we announced a definitive agreement to acquire RealSense, the market leader in depth sensing and vision technology for robotic perception. RealSense brings a mature 3D perception platform already at commercial scale and is underpinned by leading proprietary technology and a vibrant developer community. By combining these capabilities with our machine vision expertise and global customer relationships, this acquisition positions Cognex to offer a full stack visual intelligence platform. Through our leading positions in industrial barcode reading, 2D inspection, 3 measurement and now 3D depth perception for robotic navigation, we expect to support customers in an even broader set of high-value industrial vision applications. This acquisition also helps us further diversify the Cognex growth engine by expanding into an attractive new market that is closely aligned with our core capabilities. We are excited about the opportunity to bring these 2 companies together and believe the combination will make both businesses stronger. Turning to Page 3 of the presentation. I'll begin with the 6 investment highlights that underpin our conviction in this transaction. First, RealSense brings technology leadership in 3D robotic perception. The company has built a differentiated and defensible position through a full stack proprietary technology platform, spanning imaging technology, custom silicon, advanced algorithms, embedded software and developer tools. Second, RealSense is the current market leader with the #1 position in the high-growth robotic perception market. Third, it brings proven scale with more than 1 million units shipped and a large active developer ecosystem. This is an established technology platform with meaningful commercial adoption in what we see as a strong foundation for continued growth. Fourth, our technologies and expertise are highly complementary. We enable machines to identify, inspect, measure and guide while RealSense enables them to perceive and navigate their surroundings. Fifth, the acquisition expands our served market through an immediate entry into robotic perception. This is an attractive adjacency to our existing industrial machine vision markets and adds exposure to applications such as vision-guided fixed arm robots, autonomous mobile robots, quadrapads and humanoids. Finally, we expect RealSense to accelerate our growth profile. The business is growing rapidly today and over the long term, we expected to achieve revenue growth of more than 25%. Together, these attributes strengthen our technology portfolio, expand our market opportunity and add an important new growth engine. On the next several slides, I'll provide more detail on these highlights, beginning with how RealSense fits within our broader strategy. Turning to Page 4. As we discussed on our most recent earnings call, we are focused on driving growth through diversification across a broader set of customers, channels and markets and adjacencies. Over time, we expect this strategy to reduce our exposure to the investment cycles of any one customer or industry and create a more balanced set of growth drivers. While these initiatives will take time, we believe they are important to building a more resilient, scalable business with more consistent performance through cycles. RealSense advances this strategy by extending our core machine vision capabilities into robotic perception. We see this high-growth adjacency is creating opportunities to leverage our technology, industry expertise and global reach. Moving to Page 5. RealSense expands our served market into robotic perception, including 6-arm robots, autonomous mobile robots, quadrapads and humanoids. This is a natural extension of what we already do. Today, Cognex Vision systems enable manufacturers to identify and inspect products, take precise measurements and guide fixed arm robots and highly controlled industrial applications. In robotics, our technology helps a robot locate an object, determine its position and perform a specific task with precision. RealSense extends these capabilities into more dynamic and autonomous applications. Its depth perception technology helps robots understand the 3-dimensional world, allowing them to recognize their surroundings, judge distance, avoid obstacles and navigate through changing environments. Together, our capabilities span the progression from identification, inspection and precision guidance to 3D perception and autonomous navigation. This creates a broader visual intelligence platform and should position us well to serve customers as automation evolves from fixed, highly structured applications towards increasingly mobile and intelligent robotic systems. The robotic perception market is expected to grow at more than 25% annually and reach approximately $1.6 billion by 2030. That outlook is based largely on growth in vision-guided fixed arm robots and autonomous mobile robots, where adoption is more established invisible. We view humanoid and quadrapads as longer-term opportunities that are not necessary to support the core growth thesis but could create additional upside as adoption develops. Page 6 highlights RealSense technology leadership built on a proprietary stack spanning imaging, custom ASICs, 3D vision algorithms, embedded software and a deeply integrated software development kit, or SDK, at the core of the platform are differentiated depth sensing technologies in custom ASICs that enable accurate real-time 3D perception with low latency and efficient on-device processing. These capabilities come together in the recently launched D 585, an AI-native vision system that combines step sensing, onboard AI and edge computing to calculate depth directly on the device up to 30 times per second. Equally important is the software ecosystem surrounding the hardware. RealSense STK supports 4 major platforms and has generated more than 1.2 million downloads with approximately 20,000 active developers. These developers help drive design and spec-in wins across OEM robotic platforms, creating opportunities for long-term unit growth. Together, the technology platform and developer ecosystem underpin RealSense's leadership in robotic perception and provide a strong foundation for future innovation and growth. Page 7 provides a closer look at RealSense today, founded within Intel in 2014 and spun out in 2025 and RealSense has grown into the #1 player in robotic perception with more than 1 million units shipped. The company expects to generate revenue of $80 million to $90 million in 2026. RealSense also brings approximately 160 employees and a portfolio of more than 70 patents and patent applications. We are excited to welcome the RealSense team to Cognex and combine what they have built with our machine vision expertise and global reach. With that, I'll turn the call over to Dennis to discuss the financial profile and transaction terms. Dennis?

Thanks, Matt, and good morning, everyone. Turning to Page 8. RealSense meets the core criteria of our disciplined M&A strategy, strong strategic fit, defensible technology, attractive growth and a credible path to our long-term financial framework. RealSense is expected to grow more than 50% in 2026. Over the long term, we expect growth of more than 25%, meaningfully enhancing our through-cycle growth target of 10% to 11%. As the business continues to scale, we see a path toward our adjusted EBITDA margin framework of 25% to 31% and free cash flow conversion of more than 100%, supported by revenue growth operating leverage, greater scale and our global commercial infrastructure. Our near-term priorities are retaining key talent, supporting innovation, maintaining commercial momentum and executing a thoughtful integration. Overall, we expect the transaction to meaningfully enhance our top line growth profile and over time, deliver profitability and cash generation consistent with our long-term financial framework. Turning to Page 9. We will fund the $500 million purchase price entirely with cash on hand, which included $755 million in cash and investments at the end of Q2. We believe we have the balance sheet strength to complete the acquisition while maintaining the flexibility to invest in organic growth and other disciplined capital allocation opportunities, including share buyback. In addition, we expect to provide RealSense employees with a 3-year cash retention program with a target value of approximately $57 million, subject to performance modifiers. We also expect to grant restricted stock units valued at approximately $50 million, vesting over 3 years. Both programs are incremental to the purchase price and are designed to retain key employees and align the team with the long-term objectives of the combined business. Following the close, we expect the related expense to be excluded from adjusted EBITDA and other non-GAAP metrics. The transaction is expected to close in the fourth quarter of 2026, subject to customary closing conditions. Considering the additional alignment of financial period and closing schedules, we do not expect this acquisition to meaningfully impact our 2026 results. We will provide more information on the expected financial impact of the transaction after closing. In summary, we expect the transaction to diversify and meaningfully enhance Cognex's growth profile. And as real fans continues to scale, deliver profitability and cash generation consistent with our long-term financial framework. Now Matt and I are ready for your questions. Operator, please go ahead.

[Operator Instructions]. Today's first question is coming from Tommy Moll of Stephens Inc.

Matt, I wanted to follow up on a couple of comments you had regarding the TAM here. And particularly, you called out the fixed arm robot application is 1 of the key drivers. Can you give us a sense of the end markets underneath that, that are really pushing that growth as you go forward?

Yes, absolutely. Thanks, Tom. Yes, I think what we said was we estimate the robotic perception market today is roughly $600 million with a good growth profile that could reach $1.6 billion by 2030. And we believe that TAM is really rooted in more established and visible robotic applications that are being deployed at scale today. You mentioned a few of them fixed arm. We see that in many of the market verticals that we serve today, automotive, logistics, advanced semiconductor, you'd imagine that these robotic fixed arm application would be deployed across really all 5 of our core verticals today. The other one that's meaningful is mobile robotics, AMRs. And that one does have more of a skew towards warehouse automation, really, you see this quite frequently. And retail distribution, e-commerce, but increasingly, I'd say, in discrete manufacturing as well and fleets of AMR is being deployed at scale. And then we're really talking the next wave of advanced robotics, quadrapads humanoid types is really more upside to that growth profile. And so I'd say the TAM as we're sizing it today is really rooted and again, more established visible robotic applications that we see being deployed today really across our core verticals with potential upside as more advanced robotic types come to market, and we see those get more traction than I would say they have today.

And then as a follow-up, I wanted to ask, how is Cognex the best owner for this asset? And what's the value that you can unlock that potentially others can't or hadn't. And specifically, I'm thinking about channel synergies and R&D optimization, but there may be other elements you want to call out?

Yes. I mean we see just huge potential. I mean let's start with what our goal is, right, and we said it in the script, which is really to be the leading visual intelligence platform of industry. right? And we've been observing for many years now that there is kind of a shift or at least a new set of attractive opportunities. Today, and maybe for the last 30 or 40 years, we've been serving more fixed, very highly structured applications, more discrete manufacturing processes. And we're seeing a new type of applications where machine vision can serve increasingly mobile, increasingly intelligent, even autonomous robotic systems that are doing value-added work next to those more structured machine types. And so again, if our mission is to be the single visual intelligence platform, I think this is a quite natural extension as we see new high-value automation applications arising in applications that benefit from vision. But then you dig into what are the assets of the 2 companies, I think there's quite a bit of complementarity on the technology side. right? We are serving very different applications. But if you sort of unpack the imaging, the embedded processing, the visual reconstruction analysis, all of that is very similar. So the -- you can imagine that there's going to be some technology synergy that arises as we take our IP and share it with the RealSense team and vice versa, I would say. Commercial synergy, I'd say, for sure, very similar logos that we would be serving. Certainly, that would be deploying robotic systems. The RealSense business today, I would describe as more of an OEM business, meaning you do a lot of work to drive adoption and you win -- to get spec-ed into robotic platforms. And so that's a selling style that we know very well and that I think we can amplify with our global reach and commercial scale. And then it comes to deployment, right, deploying and maintaining these systems. And there, again, I think Cognex has a lot to offer the RealSense team when it comes to our global applications engineering, our global services and support and so yes, I mean, we see a great strategic fit in terms of where we're trying to take the company, great technology fit great commercial fit, and I couldn't be more excited to dig into it with the RealSense team in the coming months and years.

Our next question is coming from Joe Giordano of TD Cowen.

Can you talk a little bit about the competitive landscape here? I mean the numbers -- the revenue versus the tab suggests like 15% share at a #1 position. So can you just talk about who the major players are, how that fragmentation has kind of played out?

Yes. Thanks, Joe. Yes. So yes, as we've said, I think in this style of robotic perception applications, we believe RealSense is the market leader. They've been at this for over a decade and built just a tremendous technology stack spanning both hardware, software, and it's not just about technology, I'd say. Equally impressive is the community of developers that they've cultivated in our estimation, it's the most interacted with robotic perception platform in the world, which speaks to not just its power, but also its ease of use and sort of the breadth of what it can do. And so -- that's really important for us is jumping into an adjacent market with a partner that already has commercial scale and really has a mature technology platform. Your question, though, was about competition, right? This is a large growing space and so there are competitors. -- and there's more traditional competitors that might look and feel more like Cognex, right, that are offering traditional vision, 3D measurement that are spanning also robotic perception applications. And then there are more -- there are competitors that I would say, look and feel much more like RealSense is today, with more purpose-built devices geared more specifically to the style of robotic perception applications. But still, you can imagine through our investigation of the market and through our engagement with RealSense, we really were able to build strong conviction that they are the market leader today and have a defensible position that can last many, many years.

And then Dennis, just on the margin profile, breakeven today, it looks like you think you could get into like the Cognex normalized target zone. What's required to do that? Like -- and how hard do you push that? Like what -- when you think about the spend needed, how do you balance those 2 things?

Yes. No, absolutely. I mean first and foremost, I think we really think about profitable growth. And as we outlined in the prepared remarks, we really think that RealSense brings an accretive growth profile to us in that regard, really clearly focused on saying like, hey, we are entering into an adjacent market. which adds additional growth to our general top line, but then also is accretive to our financial framework of the 10% to 11% of organic growth. In that regard, clearly, we will make the required investments to make sure that we are achieving this growth number first and foremost. And second of all, it's also clearly that to achieve our adjusted EBITDA margin framework, it's all about creating scale and creating leverage. The business today is basically from a gross margin profile, somewhere in the in the low 50s, where we see a path towards mid potentially over the longer term towards the high 50s. And as Matt outlined, it's really a business focus on OEM customers. That means it has a much higher sales efficiency in that regard, much less broader kind of focus, less focused -- sorry, less broad-based focus in terms of the go-to-market and that means like you get very nice OpEx efficiency into this business. And with that one, as the business scales, you get also nice fall through to the bottom line. In that regard, at the end, really, our focus is to grow this business over the years to come. And with that, it will achieve that accretive growth profile, but at the same time, it will also lead to achieve our adjusted EBITDA margin framework.

The next question is coming from Jake Levinson of Melius Research.

Matt, you mentioned that there's some overlap in the technology here, but I'm just trying to get a sense of in some of these applications like robotics or AMRs, the advantages of using your traditional Cognex type of product versus perception guidance. And it's not Yes. I'm trying to get a sense of how that's happy to where the pros and cons are.

Yes. Think of it as a similar technology but optimized for different to solve different problems, right? So traditionally, Cognex 3D that you would find on. And 2D, to some extent that you'd find mountain on robots is really about precise guidance and precise measurement, okay? And we would be taking higher resolution, fewer frames to do that, right, because you're optimizing for fundamentally a different problem. The RealSense technology is a bit different, right? It's about higher frame rate, relatively lower resolutions, more streaming type depth sensing because you're having -- the output is really informing real-time motion of a robot, right? So it's often less about engaging with an object or a seen but avoiding determining real-time positioning called localization, avoiding obstacles, avoiding collisions, and so think about the technology itself is similar, but what is being optimized to do is different, and that's why we're calling this out. It's really a separate market and separate application site even if the core technology is is similar. And I'd say that what's really exciting in the past, those data were being sort of streamed off the device with the latest generation of RealSense. Systems that were recently announced. We just mentioned the D 585, a very impressive embedded system gives us the ability to also be running vision analysis software on the device. And so we think that's that's kind of a huge leap for that segment of the market. But hopefully, that's helpful, Jake, to give you a sense, similar core technology optimized for different things in the end.

It is. That's super helpful. And maybe just relatedly, how does this technology kind of integrated into your your stack, software stack, hardware stack and thinking there's got to be a lot of data that can go into one vision, and you've spent a lot of time developing a more common hardware platform. So is there a longer-term integration open.