Robotics in Facilities Management

Thank you all again for being here today for our October webinar. Some quick housekeeping before we get started. We are recording today's presentation, and it will be available and sent to you following the webinar. And also, if you have any questions during the webinar, we're going to have a couple points when we take some audience Q&A. So make sure you use your Q&A button to submit those questions, not the chat. Use the Q&A. And with that, I'll hand it over to our presenters for today. Jess, do you want to get us started with intros? Hello, everybody. Welcome to the webinar. I'm Jess Park. I'm a Product Manager on the Spot team. Hi, everyone. I'm Dan Zuba. I'm a Key Account Manager here at Boston Dynamics. Awesome. Well, first of all, we appreciate you guys all making time today, or tonight, wherever you are in the world. But whether this is your first webinar or your 50th webinar, we feel that you're going to be able to see some valuable stuff that we're going to share today. So first of all, I wanted to at least start off with just highlighting some of the tools that Spot is using today to provide value for our customers. Even after commercially selling this robot for five years, we still have customers-- not customers-- we have prospects ask, what do you use the robot for? And so here are a couple examples of what our customers are using today. Whether it's visual data, collecting that different data to get insights, or using the thermal camera to take the temperatures of pumps or motors or different types of assets, or listening for the different types of air leaks, gas leaks, as well as the more recent mechanical inspections that we launched earlier this year, Spot continues to provide a significant amount of value for our customers. I did want to camp on the visual side for a second. Earlier this year, we did release Orbit AI, which has been really exciting. And you saw a little bit of that video that we just shared. And a lot of people were surprised, that oh, you're a robotics company getting into the AI space. But actually, we've been in this space for many, many years. We go to the next slide. So I wanted to highlight a few of those examples. Now, the AI space is massive. So there's many areas that we can dive into that. But some of the ones I did want to highlight were around reinforcement learning and computer vision. Whether it's something that you see on the far left with the Agility, using different types of RL policies to be able to get a robot with wheels do a flip like that, which is helping some of our research teams for a variety of different areas. It could be using the policies that we see here in the video in the middle for the autonomy of the robot. Anyone that's been along this journey from the beginning has seen the continued increase in reliability on our robot's navigation. And you're seeing it here navigate a pretty challenging environment, whether it's walking over cables or around ladders, which in our customers dynamic environments. This happens, quite a bit. All the way down to manipulation, which is a very exciting world that we're continuing to evolve into, with some potential future robots as you start to see here. But we've been doing this for many, many years. I did want to show an example of this. So when you think about reinforcement learning, on the left here, you're going to see Spot navigating, if we'll call it navigating, but struggling as it's walking through the basement of one of our customer sites. It was very, very slippery environment. And so instead of just having the robot be programmed in terms of how to navigate this, we introduce reinforcement learning as a way to help the robot solve these and learn and then be able to adapt to that. So instead of just running a couple different scenarios, we're able to run thousands and thousands of different scenarios around what the robot can expect in that environment. And here you're seeing the after effect of being able to introduce different types of AI. And it starts to get really exciting. As we've mentioned previously, Orbit is the new tool that we launched back in May of this year. We're really now taking the same types of technology and bringing it to our facilities. You're going to see a little video here on the bottom left. And this was the foundation layer where we released a tool that's called Site View. And we are tapping into the cameras on the robot to provide our customers with essentially a 360 view of their environment. The ability to look into the past and to see what has happened. And this has been really valuable. There might have been a safety incident or some type of scenario where we need to go back in time and see what had actually happened there, or what were the events that led up to whatever was the situation. But this creates that framework and that foundation layer for us to then add the technology around a video language model to be able to add different types of inspections very quickly to the robot. So if we look at the next slide, what's really exciting is in the last call it four or five months, we've had over half a million inspections that were completed using our tool, Orbit AI. And we're really just getting started. A lot of customers are working with us to say, hey, here are some of the areas that are working, or these are the ones that we're starting to test on. And we're continuing to see very positive results. So just a couple of examples. And these are some of the visual inspections you want to be thinking about when you're in your facility. And some of these might resonate. So whether it's safety inspections, looking to see if the fire extinguishers are in the right place or if a ladder is present. Areas where they're supposed to be kept clean, and maybe there's debris that's building up, or doors that have been propped open that are supposed to remain closed all the time. Could be things like OCR reading, using the robot to capture an image and ultimately convert that to a digital reading to be able to read numbers and safety incidents like spills or corrosion build up or change over time. This tool is really giving us the foundation for really where we're going to go. We released the first version back in May, and we are going to continue to invest in this side to ultimately continue to refine the model, and potentially release more components of this to ultimately create more value for our customers. Great With that, we'll take a couple minutes and just take a few of the questions that were pre-submitted and coming in the chat. And you can keep submitting those questions throughout the presentation today, and we'll take a bunch more at the end. But the first one, I think this is a good segue for Jess. Can you talk about if Spot can recognize open or damaged doors, maybe using those AI features that Dan just talked about? Yes, absolutely. So doors are one of the first things that we were able to use AI on. Because there are very discrete locations, it's easy to teach the robot where to look at, what to look for. So open doors, closed doors, and whether there are any damages, those are some of the things that AI can solve. Great. Dan, I think this is a good one for you. How can Spot help in remote monitoring of pumps and drives? Yeah, I mean, it's a great question. A lot of our customers are using those inspections with Spot in areas where they either don't-- it's a farther location. And so maybe they're tele-upping it from a far away place where they'll have the robot traverse the environment autonomously. And when an anomaly is triggered, it alerts someone somewhere in the world to be able to take control of the robot and get a better view of the situation and ultimately decide if it makes sense to send a technician. It could be in different parts of the world. It could be in different parts of the facility, but the ability to not necessarily have to staff up every area and use the robot to be the eyes and the ears on the ground. Great. Paul asked a related question. What's the advantage of Spot doing these rounds instead of having a person walking around the facility? Do you want to take that one, Dan? I mean, there's a lot of benefits around that. First of all, most of our customers tell us that they don't have the workforce to do the rounds and readings that need to be done. Many times I've asked the customer, when was the last time you did a thermal inspection of these assets. And they say, well, we try to do it quarterly if we have the resources, but there are other priorities that are constantly getting in the way. So whether it's the reduced workforce that customers have or just the work that your employees ultimately want to do, do they want to go around and look at the things? Do they want to go and read gauges? Do they want to go take temperature readings, or do they want to do the more important work and actually fix or find the solutions for the problems that they have? And so it is definitely a better solution to have a robot do this than to have the individual go out and do it. Gotcha. I want to ask a question about Spot's navigation and perception. Is it possible that the robot can detect behaviors from people, or if the site is changing and there's something in its path, can Spot get around that too? Dan, or just either of you. So Spot is very much aware of the surroundings as it's walking. So when it encounters humans, Spot gives the right of way to the humans and waits for the humans to cross. And we also have features like crosswalk features where there's high traffic. So when before Spot is crossing the crosswalk, it looks left and right, just like a toddler would. And only when it's safe to cross, Spot crosses that crosswalk. In the future, we want Spot to be a lot more dynamically aware of its surroundings using, of course, AI and the reinforcement learning. So all of those are also on the roadmap. OK. Maybe we'll take one or two more here. Jess, here's another AI inspection one for you. Can Spot determine when a pallet of raw material or product is getting empty? Yes, with some caveat today. And that's another area that we are developing the future even more. If the pallet is at the same location consistently, Spot can do a really good job as to determining whether pallet is empty or full. If the pallet is moving around and the location is not predetermined, Spot is not going to be able to use AI to determine that question. But we are working toward a future where AI can be applied to dynamic environments so that you don't have to explicitly teach Spot where to look at things to find the answer. If I could add one thing to that too, Mikayla, I think what's interesting is-- and we're not highlighting this today-- but when we think about AI, it can be very complicated or seem very where do I get started. But with the tools that we've released as a part of Orbit AI, an example around that, you could quickly take a picture with your phone, with the robot and quickly run tests on the queries and determine, does this work, does this not work? Which really enables our customers to move quite a bit quicker, testing out the different applications around what artificial intelligence can bring. Great. I think last one for now. We had a couple questions pre-submitted and live right now about Spot's navigation and tight spaces, hard-to-reach areas, dust. Dan, do you want to maybe talk a little bit about some of the environments that you've seen Spot work in and how it handles some of those trickier areas? Yeah, absolutely. I mean, Spot spots all over the world and a variety of different environments-- dull and dirty, clean rooms, oil and gas facilities, food and beverage sites, breweries, where there's water everywhere. There's a couple of things that you want to think about. There are areas where the robot's going to perform better than others. But as we continue to use things like reinforcement learning to enable the robot's autonomy, say, in a brewery, or if we're focusing on dynamic environments where there's a lot of traffic, or these dull and dirty places where it's a lot of dust in the air. There are certain precautions we need to do, whether it's a little bit more, I would say call it routine maintenance to make sure the filters and cameras are kept clean, or whether it's these tight areas. And we want to think about, how do we create this map in these missions that will allow the robot to get to those areas successfully. And that's part of our process, where we help our customers understand, hey, here's where the robot is going to be able to be deployed very successfully. Here are some challenges areas. And then we might need to make some tweaks to ultimately get the robot to where we want it to go. All right. With that, I'm going to hand it back to Jess to keep our presentation going. Awesome. Thank you for all these great questions. As Dan mentioned, Spot is a great tool to digitize your facility using its multiple sensors as well as its capability to use AI, just PD's expertise in AI. We're building toward a future where Spot makes sense of its environment as it's-- is walking its site on its own, collecting visual, thermal, and acoustic data without you having to teach it explicitly where to go ago and what to look for. As these data get collected in Orbit, Orbit AI can detect changes in your site over time. And as far as detecting anomalies, Orbit AI is detecting changes. It can inform the right people at the right time so that your team can take on action on a problem before it becomes something that is unmanageable. Furthermore, all this data that Spot collects can be piped into a system of your choice so that you can leverage your own AI models on top of the Spot data. So historically, Spot has been marketed as a maintenance tool to serve the reliability engineers, as well as maintenance technicians. But what we are learning is that the same Spot can serve multiple personas in your organization. For example, line owners, say reduce unplanned downtime by doing this predictive maintenance. Plant managers see that their employees can upskill and also their employee morale get boosted. Security directors can see threat, detect threats early on, and safety and compliance officers can keep people out of harm's way. And data scientists can use the data that Spot is collecting for their own AI development, all the way up to executives realizing their own business goals using Spot. As you can see, there are many teams and many personas who can benefit from Spot. So it's very important to engage all these different stakeholders early on so that they also become champions of your robotics program. So now I'm going to dig into some of the specific value props for some of these personas. So for the reliability team, the Spot value prop is to predict and prevent maintenance issues. As I mentioned earlier, Spot can do the routine equipment inspections autonomously. And when it's finding any anomalies, and it can alert the right people at the right time. This ends up freeing your team for higher-value tasks that Dan mentioned during the Q&A session so that they can focus on fixing issues and building out a solution for the problems that you see over and over. Orbit can also integrate with your CMMS so that when the alert gets generated through Orbit, it can create a work order so that your team can respond to it right away. This all helps avoid unplanned downtime and help you reduce energy cost. The next one I'm going to talk about is environmental health and safety. Using Spot, you can be audit-ready and avoid fines because Spot can continuously monitor your site, looking for debris and puddles on the floor, detect-see burns in the hazardous areas, and also identify any missing emergency and protective equipment so that you can be ready when the audit happens. One of our customers, Keggel, is actually using Spot for this particular use case. So I'd like to showcase that in the video. [AUDIO PLAYBACK] One of the things we've been recently trialing out is AI visual inspections. So we're using the robot to walk around the plant, check things out, and then assess, via the AI natural language, questions what it finds. So you locate what you're trying to capture. You take a picture, and then at the end, you basically ask a question. And this question could be a numerical one, a yes or no, or even an open-ended one. Side View is a 360 plan view. What we're currently using it for is to support our food and safety quality team by basically having another set of eyes in the plant. Spot walks this route multiple times a day, and then you can identify leaks of seeds and oil early. And you can also historically look back. [END PLAYBACK] Awesome. Spot can also help the security team. Security team can teleoperate Spot to assess dangerous situations. And also they can dispatch Spot on the auto-walk security patrol on a periodic basis and then look at the feed that Spot is sending from the comfort of their control center. And when something strange is detected in that feed, security team can take control of the robot at that point, start teleoperating to assess the situation even more. We also have some partners who built security solution on top of Spot, such as Aztlan, who is serving customers in terms of premier. Security. Now, you've seen a handful of personas and the areas of where Spot is providing value. And so, I think it's really important that our customers at least start in the areas where we know the robot provides value. Some of those examples I mentioned earlier, look at the acoustic and the vibration and that whole space. But some of our more advanced customers are really starting to lean into this AI space. And this is where it gets really exciting from a value play. So as Spot continues to provide continuous streams of structured data, that plus the ability in the continued enhancements of our Orbit AI model, that is helping customers exponentially expand on the different types of inspections. So whether they are using tools that we're providing or they are using that data that is being collected and creating their own models, the amount of visual inspections that are out there that can help our customers move from just describing what's happening to taking that, and to put it into a data lake and move to predicting and prescribing what's happening is actually pretty exciting. And I wanted to highlight one of our customers that we work really closely with, Intel Foundry. Maybe you guys have seen this case study. If you have not, I highly recommend you reading it. You can scan that little QR code at the top, or I believe it's going to be dropped into the chat. But we've been working very closely with the team at Intel and RealSense on this case study, but specifically with Intel for the better part of two years now. And they started very similar to most of our customers in the journey of where they're at. What can Spot be used for? Where can it ultimately be deployed? And ultimately, we landed on the rounds and readings and very critical areas of their fab. So could be ultra-water processing facilities, could be sub fabs, could be the cooling distribution areas. And the impact for Intel has been quite huge. So as you can see here, just a couple of stats, whether it's the saving of the workforce and having them focus on much more important tasks than to go around and do these rounds and readings, catching issues of critical assets before they go down, extending the life of a functional parts, ultimately moving from a world of preventative maintenance to conditional-based monitoring. And then that middle bullet is really important. That continuous stream of structured data that they are leveraging to feed into their AI-powered tools, that ultimately creates better insights. This is enabling customers like Intel, and again, you can read the case study, to ultimately start small with that one robot. And now they have a goal here by the end of the year to ultimately deploy 30 of these all over the world. We are really excited about what's coming in the pipeline in the Spot and Orbit roadmap. Since Spot commercialized in 2019, we've continued to improve on Spot's autonomy, as well as inspection capabilities, along with the PD's own AI capability. Now we want to wrap all those and give you the AI capability into your hands so you can make more sense of and insights out of your facility. Imagine Spot walking around the site, making sense of its environment, identifying objects and assets of interest, and labeling them, and continue to monitor to alert you when an anomaly happens, when changes happen so that your team can action on them right away. We're also building a lot of features for various personas in addition to the maintenance team. So the traditional maintenance team use cases, as well as EHS and security teams, are one of few of the personas that we are targeting to build features for. And lastly, what is really important is that using all the integration that Spot and Orbit offers, you can plug-in the solution into your existing workflow without disrupting your current process. So we want the Spot and Orbit solution to be part of your team, rather than being something that you need to adapt by disrupting your processes. So with all those exciting things that you are curious about, I'm going to hand it over to Dan so that he can share how you can get started with this Spot project. Yeah, thank you, Jess. And this is the questions we get a lot. Like, this is really exciting. This is cool technology. I don't even know how to evaluate it. How do I build a business case? So whether you're watching this webinar live or you're going to watch a recording and this is your first experience in trying to understand, can I use this technology, or you're a customer out there who says, wow, I do have a robot, but it's only in one part of our facility. How do I continue to see the value so that I can expand this across our enterprise? We have a very detailed and defined process to ultimately help that out. So I'm going to quickly go through just so you can see what this looks like for an example customer. So one of the things that we start with is we use our engineers to ultimately come on site and do what we call a technical evaluation. Customers have to understand, where can this robot be deployed? What are some of the areas that need to be concerned about? What's going to be the highest value? And so for this example customer in the food and beverage space, there are three areas that we know Spot performs very well, provides a valuable solution. And typically we see payback in a very quick time. And those areas are the packaging areas, utility areas, or production area. So we actually will send an engineer on site and work very closely with your team and start to really map out what the facility looks like. We want to identify what are the number of inspection points, how many visual inspections, how many AI inspections, how many thermal or leak or mechanical, all of these different inspection types that are out there and map out, ultimately, the path around what the robot is going to do and how we're going to set this up for success. We then try to quantify that with you to understand what is the value of finding an air leak, or what is the value of keeping the line running versus the unplanned downtime. And we put this all together and put together a business case that you can actually use. And so if you look at this next slide here, again, whether we are focusing on our first robot or we're focusing on deploying in just a small area of our facility, or maybe it's all of the packaging lines, or maybe it's the packaging, the processing, and the utility space, or you already are an established customer who has seen incredible value and you want to understand, how can I be like the other customers you guys mentioned that have gone from 1 to 10 to 20 to 30, even 50 robots, so that we can see this value being realized across our enterprise? And we put this together and hand this off to you that you're able to use to help sell internally. And then when you're ready to move forward, we have an entire integration team that's going to come on site and lead you through this journey from start to finish, all the way moving to site acceptance testing and handing it off to our support team. So again, whether this is robot number one, whether we're looking to expand to small areas or across your enterprise, we have a very detailed process that is ultimately going to continue to help create value for your organization, whether that's avoiding unplanned downtime, whether that's saving energy, whether it's keeping people out of unsafe conditions, or whether it's the value of what this data is going to bring for future applications, Spot is the tool that can help and Boston Dynamics can be the one to help you bring that to life. Amazing. Thank you both for your presentation. That brings us to the end of our content today. But we have quite a few questions to get through. So we're going to dive right into those for a while. We've had a number of questions come through about the deployment process, just beginning to end, what does it look like to get started with Spot on site? Dan, you just touched on it with what you were just presenting, but do you want to walk us through some of those key milestones and give us just a little bit more detail? Yeah, I mean, I think each different, call it milestone of the process is very well-defined. And I would highly recommend reaching out to our sales team where we can walk you through it. But it starts all the way in the beginning around that technical evaluation that's going to help you, one, build the business case, but two, most importantly, make sure that the environment and the assets have been identified. We then hand that off to an integration team that moves to a project to where we actually work very closely with your local teams to ultimately get the site ready. There's some change management pieces that we need to focus on. There's IT, there's health and safety and security and getting everybody on board to understand what this new team member is ultimately going to do. And we then move through our project phase and send our teams on site to ultimately get the robot up and running. And then we meet our site acceptance testing that we've already agreed upon. And once that is done, we hand it off to the facility for operations. Once we do that, we are continuing to value track. We want to make sure that what we've deployed matches back to the hypothesis of the business case that we put together. And then we use that as the framework of how we can now take this and go from location 1 to location number 2, and then scale throughout your enterprise. About how long does that entire process take, on average? Yeah, that's a great question. It varies. But I would say, on average, it's around six to eight weeks. There's a lot of factors that are involved in that. And typically, as you can imagine, this process has started early in the process before a PO is even issued. And then once that PO is issued, there is work that needs to be done before the robot is actually going to be delivered on site. So it works in parallel as the robot is being ready to be shipped. And then once it comes on site, we have some prep, some final prep work, the team comes on site and then we're off to the races. Great. What about the-- Jess or Dan, you guys can tag-team this. What about the training process for Spot? You drop it on site, how do you show it what you want it to inspect? And maybe we can cover a little bit more detail around how the AI inspections work specifically, too. Yeah, I think there's a couple of ways you can look at that. So training is a very exciting spot-- no pun intended. For us right now, you might have seen our recent announcement around Boston Dynamics University. And we continue to create content to help our customers see the value and continue to improve their knowledge around what the robot can do. So there are as detailed as advanced training that we offer, whether it's a two-day, fully immersive, autonomous training, autonomous robot training class that can be done either in Boston, at our offices, or on the customer site, all the way down to API training and manipulation type training with the robot. So have a whole fleet of classes, if you will. But then we also spend time with our customers when the team comes on site to deploy the robot. We want them to understand and see, well, how do I use this? How do I record an inspection point? How do I test out these Orbit AI inspections? So whether it's the resources that we have online or some of our on site that we'll bring as part of our integration or the advanced classes that are available. We have a whole suite that we can help our customers be successful. And even from the product perspective, we are making Spot to be super easy to operate. So initially, we had engineers come to the site for like two weeks to teach the robot what the map is, what assets to inspect. But as we develop the product features, we are making it easier so that all customers need to do is record a safe and healthy map, and all the inspections can be added later based on the data that Spot has already collected by walking through the site. And we will continue to make investment in that space so anybody can fully leverage Spot without extensive training. Awesome. Jess, I think this is a good one for you. We've had a lot of questions around integrations. How can I connect Orbit with my preventative maintenance tools or other third-party systems? Can you talk about our options for data integrations? Yeah, Orbit has a public documentation when it comes to APIs. We can post that on chat. And your IT team can fully leverage all those APIs to pipe the data into your system. One use case we see over and over is having the alert trigger work order in your system. So we also have a feature that is like a low-code work order integration, where you just type in like URL and some basic mapping of the data between Orbit as well as your work order management system. So when alert gets generated in Orbit, it pipes into work order system. So the work order gets created there automatically. Great. Can you see a live view from Spot's eyes when it's walking around the plant? Either one of you want to talk about Teleop and Orbit and everything that you can do inside of Orbit? Yes, you can. I see this as a really good feature for security team, for example, thinking beyond the maintenance team. So imagine a security team, from the comfort of their control center, looking at the live feed that Spot is sending, and only acting on dispatching of real security guard or calling 911 if Spot detects something, without having to send people to do the continuous patrol. So that's a very real example of how this live feed can be used. But even from the maintenance and EHS's perspective, so you can see the spotlight feed and just monitor the site continuously. And I see that a lot with emergency response teams. If there might be a situation where it's unsafe to send a human into an environment and Spot can be a tool that can get there to collect data to ultimately determine if it's safe before we ultimately send a human in there. We have one question about maintenance for Spot. Either one of you want to talk about some of the basics of what it takes to keep spot up to speed, and how you can work with our support team? I think the most important thing is making sure that Spot can see clearly, because Spot needs to see to be able to navigate its environment. So depending on the site, it's anywhere between a monthly lens wipes or a quarterly lens wipes. There are some sites that are really dusty, in which case it might require more frequent wiping of the lens. So that's one. And we also have a health check built into the Spot system so it continues to calibrate its joints to make sure that everything is performing properly. So that we also recommend like every three to six months. And when you purchase Spot, we also send you an eye examination tool that you can run every once in a while so that the spot perception is also calibrated. Go ahead, Dan. On our support website, we have a pretty detailed monthly checklist that highlights some of the stuff that Jess said there, but the amount of time it should take from a customer is very low. Cool. We had a few questions about how many Spots are actually out in the world working today? Dan, do you want to take that one. Yeah, there are thousands of Spots out in the world today, all over the place, all across this globe. They are in many different countries, many different verticals. We don't disclose exact numbers, but we're not-- we're measuring them in the thousands right now. Not in the 10 thousands just yet. But we'll get there. Yeah. Are there any customers in Europe, Germany, other regions outside of North America? Absolutely. Europe is a very exciting area of the world for us. We actually have our second office that was opened in Frankfurt, Germany, a little over a year ago, and that's for service and support to ultimately, help our customers that are in Europe. You can look at a couple of our case studies that we have on our website and there from customers in that space. But the European market continues to be one of our most, I would say, important markets. Great. We had a few more questions come through about change management and how to introduce Spot to your employees to your site. You guys want to take that one together? Sure, I can start. I'll say the change management piece is very important. When the team members on site understand what the robot is going to be doing, they start to get really excited about it-- once they realize it's going to be doing some of the tasks that they're not interested in. And so we have a lot of materials and videos that help educate some of the members around what is coming and what to expect. And then we recommend things like naming contests. You everyone can name what they want their Spot to ultimately be called. And that starts to bring excitement into this. That is not just a robot that's coming. This is actually going to be one of the team members. It's going to be doing a lot of the tasks that we ultimately don't want to be doing anyway. Yeah. One of my favorite names of Spot that our customers gave us Nacho. They are tortilla manufacturers and their employees gave the Spot Nacho, and it works perfectly with the color of the dog as well. Connor asked a question, does Orbit or the Spot software have the ability to manage a fleet of robots? The example is robot one is halfway through a task, needs to recharge, and robot two takes over the task. So maybe we can start by talking about how that scenario would work. If the robot needed to recharge mid-mission, we can also talk about some of the enterprise capabilities in Orbit. That's a good question. So Spot knows to return to the docking station when it's getting low on battery, so that it doesn't collapse in the middle of the mission. So Spot has a built-in capability to not die because it ran low on the battery. In terms of multi-robot fleet, Orbit is built for that. That was the initial purpose of the Orbit software. So Orbit manages all the Spots within the customer site, even all across multiple sites. So one enterprise might have multiple locations where Spot is deployed, and their instance of Orbit can manage spots across the site. And our future vision is that all the PD robots will be managed through Orbit so that you or your team learns one experience, one software experience, and that can be applied to all the robot management. Great. I think we have time for just a couple more questions here. Dan, this is a good one for you. We had one question come through about the ROI, the big ROI question. How do you typically justify a Spot purchase? Want to give us a refresh on some of those big themes to highlight? Yeah, absolutely. I would say for the customers that this is new, it's hard to evaluate new technology like this. And so that's where our defined process will help you calculate that for you. So the big buckets that we see the value coming in is energy savings, preventing downtime, preventing equipment failure, re-augmentation of workforce. And then ultimately the data. What is the value of what this data could bring? And so we put that all together. And as part of that process that I outlined a little bit ago, we'll help you start to calculate that and figure out what it's going to look like. On average, many of our customers are seeing payback in 18 to 24 months. We do have some customers that have been public that, on average, it's closer to 12 months. Each different environment is going to be a little bit different. Each different driver of the use case is going to be slightly unique. So I would highly recommend anybody that is interested in, where do we even start and how do I evaluate, is to reach out to our sales team. We'll be able to walk you through what that process looks like. We'll be able to coordinate when it makes sense to send an engineer on site and actually help you evaluate this so that you have all the information you need to know, one, what's the value you're going to provide to the business. Two, what is the areas of our facility where we're ultimately going to deployed. And three, once we move forward with this, what does this actually look like to implement this. And all of that is part of our process. OK. I'm going to ask one final question. But if we didn't get to yours today live, don't worry. We'll be following up to answer your questions afterwards. So stay tuned for that. But to bring us to a close today, Jess, this is for you. How does Boston Dynamics and vision Spot evolving over the next few years? What's some of the directions that we're headed with 51 and beyond? Yeah, we are working on a lot of different applications. So Spot has proven to be a great tool to collect data. It collects data consistently without complaints, and you get the same quality data over and over. So immediately following that capability is giving you more capability to gain more insights using that data, using our AI capabilities as well as what I see as our differentiator when it comes to collecting data using multiple sensors together on the same asset. So that's one. And the other one is like looking at all these other adjacent markets where we are seeing Spot delivering value using the same set of skills it has already. So security market may be another one that we expand on our application. And the other one that I'm really excited about is because Boston Dynamics has multiple robots, robot platforms. So not only Spot but we also have Stretch. We also have humanoid robot called Atlas. We're envisioning a world where all these robots can come together and deliver true value to you, whether it's in the industry or even in the service sector. So that's like a long-term vision that we have. But we want the robots to be truly helpful for the humans and be either a team member or somebody at home taking care of the chores and whatnot. So we have an immediate future that we're looking at delivering value to our existing customer and adjacent market, but also accelerating the development so that we can build out the future where robots can just truly be the help that humans need. I think that's a great note to end on. Thank you, Jess and Dan. Like I said, if we didn't get to your questions live today, we'll be following up with you. But in the meantime, you can use the QR codes here on the screen to check out some more info and get in touch with our sales team. So thank you everyone for joining our presentation today. The recording will be sent out following the presentation, and we'll hope to see you at the next one. Thank you.