Transforming Manufacturing Reliability Inspections with Mobile Robots

- Thank you for joining today's webinar, transforming manufacturing reliability inspections with mobile robots. My name is Vatche Arabian, application expert for Boston Dynamics. We've got a lot of content today. But before we get started, I just want to run through some very quick housekeeping. We will be sending out a recording of today's presentation following the webinar. So look for that. If you have any questions during the webinar, please use that Q&A button anytime throughout the webinar. We do have folks in chat monitoring and are able to answer questions during. We do want to take some of those questions live. So we'll be pausing in the middle of the presentation to take a couple of those questions and then stick around for the end for a much longer Q&A. Let's start off with some intros. Rich. - Yeah. Good morning. Good afternoon. Good evening to everyone on the webinar. My name is Rich Lyons. I lead our global food and beverage vertical here at Boston Dynamics. I've been with Boston Dynamics for about four years now. Very excited to talk a little bit more about Spot today. It's probably worth noting that while I lead our food and beverage vertical here at Boston Dynamics, this webinar is really focused around integration of mobile robotics for data collections across all industries. So we may have a few food and beverage use cases, as well as some other industrial use cases. But very grateful to be here today. And welcome to the webinar. - Awesome. Thanks, Rich. And I'm Vatche. I've been with Boston Dynamics for nearly four years. And before joining the team here, I spent longer than I'd like to admit working with companies on how they use thermal imaging technology to aid their predictive maintenance programs, which, if you're new to what Spot does, that's no stranger to me. So if you are new, before I hand it over to Rich to really talk deep dive onto the value propositions of Spot, I really want to take a step back and talk about how we got here. So we meet with hundreds of customers. And we've met tons of different use cases. And digesting a lot of what we're hearing from them, typically why they're looking to us for assistance on their journey, we start to see a lot of these themes that you're seeing on the screen right now. Our customers are looking at every way they can increase their reliability and uptime. They've got energy or sustainability goals that they're looking to meet across their organizations. They've got challenges with labor, and they're looking to leverage advances in AI and robotics for more predictable operations. We hear the term manufacturer of the future or factory of the future, or facility of the future a lot from our customers in some form or fashion. And they know to stay relevant. These facilities that have been around a long time know they need to keep up with change. And almost all of you have implemented some type of digitalization or industry 4.0 roadmap. This next figure is really interesting. This is from a new study, 68% of manufacturers are implementing extensive corporate wide sustainability strategies. And a lot of your organizations recognize that this is just good business. On the labor front, we've heard 60% of you cite inability to retain talent. It's more competitive than ever out there, and you don't want your highly skilled employees moving on for less dull and dangerous work. And lastly, kind of touching on that AI front, you've made these investments into advancing to industry 4.0. But you're stuck really getting the full benefit of that intelligent automation because you don't have a cost effective or streamlined way to collect and capture and organize that data. And that's where our solution comes in with Spot and Orbit. So Spot moves around your facility, bringing the sensors to your assets. It's all accessible and you can see all that live data in Orbit. It checks those assets for equipment health so it can use a thermal camera, it can use a camera to look at gauges, it can use an acoustic sensor. So it collects all that data. And when something is running outside of spec, it can create that alert within Orbit. And that orbit's got plugs into your existing work order systems so that your team can more efficiently and effectively have more wrench time and take action. And that's really the short elevator pitch of how Spot fits into this plan. So I'm going to hand it over to Rich to deep dive. - Thanks, Vatche. Just to provide a high level overview of the flow of today's webinar, I'm going to really break things up into two parts. The first part being sort of an introduction to Boston Dynamics and Spot, how Spot is delivering value to our customers today, and also the investment of Boston Dynamics into the research and development of this technology to deliver a highly reliable, consistent, and accurate mobile data collection platform to our customers. We'll take a break after that brief introduction, and then we'll dive into more along the lines of how do we integrate Spot into your environment, how do we manage the data that Spot's collecting and make it easy for you to make business decisions, and we'll also take a look at some real world examples, some real customer use cases, as well as back into how we help our customers develop a return on investment business case so that they can take that to their business leaders and stakeholders to help drive innovation within their facility. We'll conclude with another Q&A towards the end of the webinar. So Vatche shared a quick video here of Spot navigating the environment. Here's another quick one. It's only 30 seconds, but I wanted to just go over this is Spot. This is Spot autonomously navigating your environment. You'll notice the robot has some blinking green lights. Those are visual indicators for us humans to really understand what is this robot doing, is someone driving the robot. You'll notice some key payloads on the robot, which is a thermal-- I'm sorry, Spot cam IR, which includes many different types of sensors, such as thermal, 360 photo, pan tilt zoom, two way communication as well. And you'll also notice a high level of mobility. This robot can avoid obstacles, have this concept of entity detection, whether that be maybe a forklift or a human. The robot can climb stairs as easily as you and I. So making its way to the second floor in this instance. Here, we see the robot sort of at an intersection within this facility. And essentially, the robot's looking both ways. In this case, the robot's been trained to identify a forklift. So if there's a forklift in the field of vision of the robot, it essentially pauses and waits for that item or entity to be removed from its field of vision, and then it continues on with its mission. But where the robot really delivers value is around this automated data collection. So in this case, we see the robot pausing, stopping, maybe doing some gauge reading here, maybe doing some acoustic sensing as well. And there are thousands, if not tens of thousands of these fittings and valves and gauges and rotational assets such as motors, pumps, and gearboxes that the robot really excels at fully automating, going out on various missions, collecting this information, and then automatically returning back to its docking station where essentially the robot recharges, downloads the data that has collected, and returns it into Orbit, our data platform, which we'll talk about in a few minutes. And right now, you can see on your screen, the robot's recharging. It takes about 60 minutes for the robot to recharge. And then it's ready to go back out on its next 90 minute mission. So I thought it would be fun to share some real world images or data collection that the robot has done. Sort of working from the top left, going clockwise, we can see the robot has identified, in the top left, an air leak. So this was inaudible to the human ear, yet the robot was able to pick up this unique frequency of this industrial gas leak. As we move to the middle, and the top, we're starting to take a look at some thermal profiles. And on these thermal profiles, you can actually set upper and lower limits. So if a motor should be running at about 100 degrees, you can say, if that increases by two or three degrees, please trigger an alert. And you can have that real time alert if you wish, or you can have the robot return to the dock and then download that information. On the far right, we see some gauge reading. In this case, the customer has put some red cellophane over the gauge to indicate normal operating conditions. For Spot, what we would do is overlay digital thresholds on this analog gauge, which would be maybe a little bit more reliable than having the red cellophane if that falls off or someone the gauge isn't really perfectly reading. The overlay of the digital thresholds is really nice with Spot. And then on the bottom left, we see some more thermal profiles. In the middle, we're starting to get into some other supplemental data collection such as 5S boards. You see another air leak there. And on the far right, we also have customers that are integrating human health and safety into the data collection path for Spot. So in this case, the robot's taking a picture of an eyewash station as well as a fire extinguisher. And we could integrate other software that says, is the fire extinguisher present or not, or is the eyewash station blocked or not, and trigger an alert based on what the robot sees. So these are real world images, just sharing with you the quality of the data that's collected by the robot with a high level of repeatability and accuracy. So these are the primary economic drivers for Spot today. These are the primary inspection types that are delivering return on investment. So I mentioned a little bit about the visual inspections that the robot is doing. These are those analog gauges, thermal profiles, 360 photo capabilities. The camera on the robot is an extremely powerful camera. It is 30x optical pan tilt zoom camera with thermal and with that 360 photo capability. It also has the thermal analysis. We saw a few images of that in the previous slide as well as acoustic leak detection. So Boston Dynamics works hand in hand with Fluke. This is the Fluke SV600 that's listening for industrial gas leaks in your environment. Things such as compressed air, CO2, ammonia, nitrogen, the robot is able to pick up any of those leaks within your facility, many times inaudible to the human ear. And then finally, on the far right, with that Fluke SV600, we've just integrated new sonic vibration analysis. So think of the robot autonomously navigating your environment and listening to conveyor systems and listening for ball bearings that give off a very unique frequency prior to failure. So you have the robot that's going out fully autonomous, listening, sensing your environment, providing you with predictive analytics that allow you to make business decisions. So let's take a quick look at Spot itself. Here we see the full components of our industrial asset management kit that we deliver to our customers today. On the bottom, we see our docking station. This is where all of the autonomous missions were going to be starting in your environment, as well as terminating. The docking station provides the ability to charge the robot, as well as download the data that the robot's collected while it's been out on its auto walks. You see some of the payloads a little bit more clearly here in this image, you have the Spot Cam on the front of the robot. You have the SV600 by Fluke in the middle of the robot. And then on the rear, there's some additional visualization technology for the autonomous missions. One is a Velodyne LiDAR Puck. There's also 5G LTE modem on that payload on the rear, as well as some additional compute. So think of running real time video image analysis on Spot in real time in your facility and using that extra compute that's on the robot. So several components for the industrial asset management kit. You have the docking station, Spot, the robot itself, the Spot Cam by Boston Dynamics, the acoustic sensor for vibration analysis, as well as industrial gas leaks, and then the visualization component for Spot to navigate your facility autonomously. In the end, on the bottom right where everything really comes together-- and we'll dive a little bit deeper into this. In the second part of this presentation-- is our fleet and data management platform, which is called Orbit. I thought it might be worthwhile just to touch base a little bit on Boston Dynamics and the investment that we make here at Boston Dynamics in ensuring that this technology can be deployed reliably in your facility. So these are some images of our laboratories here in Boston. We have a fleet of over 125 quadrupeds constantly walking around our facility. We do about 6,000 hours of testing within our research and development labs per month here at Boston Dynamics. And we're testing things like motors, hip actuators, new payloads, new software that could be integrated into the robot. We have about 1,500 robots deployed all around the world today in all different types of industries. So food and beverage, pulp and paper, maritime, mining type environments. So the robot really performs well in really challenging environments. Think of even nuclear decommissioning sites. Perfect example where a robot would be deployed to keep a human out of harm's way. Keeping on the reliability note here, our robots have actually traversed the Earth more than three times fully autonomously. So over 78,000 miles traversed by our robots on their pre-programmed autonomous missions. And just last quarter, we had robots that reported back over 400,000 autonomous inspections. This graph is a real solid indicator of the reliability, consistency, and accuracy of Spot. You can see from this graph, over the years, the robot just consistently gets better quarter to quarter in its ability to go out, navigate the real world, avoid obstacles, and collect real business intelligence from your facilities. On the right-hand side, we see a quick snapshot of some of our key customers that are partnering with Boston Dynamics today. All right. So I'd like to pause here. After that brief introduction of Boston Dynamics, our research and development facilities, the components that make up Spot, I just wanted to pause and see if there are any questions on this section of the webinar before we move forward. - Yeah, people did their homework. There's a lot of questions, Rich. And we're definitely going to save time for the end first for a lot of these. But I do want to give you a couple. So one question came in from Paolo on stairs that came in pretty early. So can it climb all types of stairs? If not, what are the stair standards? And Rich, you want to take that one or-- - Yeah, sure. So as we saw in one of the videos, the robot can autonomously navigate stairs. There's no special programming in that in any way. Essentially, when you train the robot to navigate your facility and collect data, if there are stairs involved, the robot can navigate up and down those stairs. There's no sort of special programming that's required for the robot. There are some specifications of those stairs, essentially just normal specifications for a human. Railing widths have to be a certain width for sort of safety for the robot to navigate those environments. Yeah. - Yeah, and you can find those exact specs on the spec sheet. I do note, the max slope is 30 degrees, and then step height, 300 millimeters. But yeah, if you're interested in learning more about those types of stairs, yeah, we've got that in the spec sheet. Is Spot limited by climate conditions snow in minus 30c. So minus 30c is definitely outside of Spot's published specs. So I believe it's negative 20 to 55c. So people are using it for outdoor applications. We've got some solutions there. So we'd love to find out more about your specific application and use case. And we can make some recommendations on how to achieve that. And by the way, it's also IP54. So light rain. But again, outdoor applications aren't out of the question. We have customers using Spot outdoors, so we just love to hear more about that and that use case. There was a question. How is Virginia Tech using Spot and Orbit? And I'm just going to plug that case study. So that's a little off topic from the focus of this webinar, but there is a great case study about how Virginia Tech is using it to research different construction applications. So it's primarily a research application at Virginia Tech. For Spot robustness metrics, you mentioned 125 robots. Is there a number-- Yes, so that number plus the miles traveled. So that's our test fleet only and not including-- and we've shared the bigger number in the past of all our customers robots plus our test fleet put together. So yeah, on that slide, we're sticking to our test fleet for that number. The other question is, how to Spot help increase quality and manufacturing processes. How is it really helping the final product. Rich, do you want to try that one? - Yeah, that's a great question. And Spot provides real business value in many different ways for our customers. Ideally, it's around improving efficiency and accuracy with data collection within these manufacturing processing environments. So if you think about it, the robots taking over a lot of these lower level tasks that humans are doing today. Think of just going out and using a handheld thermal camera or using a handheld acoustic sensor. That's time consuming for your team to go and do that. It also lends itself to a lower level of accuracy and consistency because I may be doing it this week, and Vatche may be doing it next week, and maybe we actually don't do it the third week because other things have come up in our facility. With Spot, the robot is consistently going out on these automated autonomous missions, using its sensors to collect all of this lower level information with a high level of repeatability and accuracy. The data is returned back to our Orbit platform that allows you to make business decisions. So for our customers, this is really important because it allows them to redeploy their most valuable asset, which is their employees, in a way that delivers greater value. So in many cases, the folks that used to go out and take all these handheld measurements and do the data collection, with Spot on the job, it provides them with the opportunity to have more wrench time, which essentially allows them to go out, repair different types of anomalies that the robot found faster. And that impacts the business. It helps reduce downtime, it helps improve efficiencies around energy costs, it helps keep their employees safe in some instances. So there could be some hazardous environments where you may not want to get fully dressed in PPE and put a human in harm's way. You can remotely access the robot and drive the robot into that environment and assess the situation within that environment. So two other ways that I could think of that. We see customers really leveraging the technology-- and I would caveat this as sort of a supplement to the industrial inspections-- is around human health and safety, doing those safety checks around fire extinguishers, eyewash stations, and also supplemental security rounds. So are doors open, are they ajar, things of that nature. So the robot really does span quite a broad range of data collection types that could be important to your business. - That's great. Thanks, Rich. Yeah, so two more questions, and I want to get back to the content. But there was a question on are you using the acoustic payload for both the leaks and vibration. And the answer is yes on that one. It is that one sensor. And this is a good segue to I think where we're going next. How can the data collected by Spot be integrated into outside CMS software? And then there was another question somewhere that I saw about can it be used to pull into SAP. - Yeah, exactly. Great questions. And like you said, Vatche, what a nice lead in to speaking a little bit more about Orbit and ROI. But absolutely, the data that's collected by Spot and returned to Orbit can be integrated into IBM Maximo, SAP, imported into various data lakes that you may want to use this data as well. So yes, you're not just locked into the Orbit platform. You can share that data. - So without further ado, I guess, what's Orbit? - Perfect transition. So let's take a moment and talk a little bit about the data platform, Orbit. In my mind, this is where everything really comes together here at Boston Dynamics. This is your single pane of glass to manage your fleet of autonomous robots that are working for you in your facility. So in Orbit, this is where you're going to be able to identify those real time insights, start to really look at trend analysis over time for your rotational assets or other aspects of your business, and even auto-create work orders. So we do have customers that have the robot connected in real time and are triggering alerts into their engineering and facilities ticketing system, assigning the ticket to a human, who has access to all of the historical data around that asset as well. So what this leads to is a dramatic reduction in mean time to solution. We have some customers that are measuring this today. We've seen customers reduce their mean time to solution from 37 days down to 14 days. So in my mind, that sort of gap from 37 to 14 is really risk and liability for the organization. You have a known anomaly, something that's heating up or leaking or vibrating in a way that could impact your business. It could stop production. It could really economically impact the business. But with Spot and creating those work orders and reducing the mean time to solution, you're really lowering your risk and liability for the organization. So let's take a look at what Orbit looks like. This is the Orbit environment. On the bottom left, those yellow rectangles, those are your fleet of robots. You can see the robots that are operating in your facility. In the middle in the blue, those are the various missions that the robot is going out and performing various types of tasks for your organization. And then in the red, those are the anomalies that are found. And you can click on these anomalies. Click on one of the anomalies. Take a look at, in this case, the gauge reading, go back in time, how did this asset perform last week or a few days ago, what's changed in your environment that has triggered this alert for your team. Orbit is a really powerful data platform that not only allows you to interact with the data that the robot's collecting, it also allows you to make changes to those missions. So in Orbit, you can easily author, modify, monitor missions that the robot is on. You can use the data for real time intelligence. You can export this data, as we mentioned earlier, into other data platforms as well. In my mind, this is the brains of the entire package. You have this amazing mobile IoT platform Spot that's going out and collecting the data. All of the data is returned to Orbit in a way that organizes the information and allows you to identify trend analysis or immediate alerts within your facility. So I might have mentioned earlier a little bit about the pace of innovation. And I wanted to touch a little deeper on this. Spot's been roaming this wild world of ours for about 5 years now. Spot was launched in 2019. And over the last 5 years, Boston Dynamics has just continued to innovate on behalf of our customers. So in this quick timeline, you can see some of the key features and benefits that we've introduced to our customers over the last 5 years. But I wanted to really touch on a few of them on the next few slides. The reason for this is Boston Dynamics works extremely closely with our customers. Most of these features that you're seeing introduced into the Spot timeline are direct feedback features requested by our customers. So things such as the ability to avoid obstacles or entity detection. We're going to take a look at some of those examples in the next few slides. So in this case, the ability to detect a moving object relatively quickly. In these manufacturing environments, they're highly dynamic. You have lots of things moving-- people, humans, forklifts. Here, let's take a look at how quickly the robot can react for something that it wasn't expecting to come into its field of vision. So you'll notice that this case, the robot was on an auto walk. It was blinking green as a normal auto walk would. But as soon as it saw something coming from its peripheral vision, the robot stopped, it started blinking red, and actually gave off an audible signal. So this is really helpful to help avoid collisions or things that could disrupt or interrupt an auto walk mission for the robot. In the next video, we're going to see the robot essentially encountering a human in the hallway. You can see the robot here is automatically navigating around this particular human. And you have control over the variables on how the robot would interact with a human in that environment. An example would be Spot, if you see a human, just stop until there isn't a human in your field of vision. Or maybe Spot, could you just proceed with caution and stop within 1 meter of a human, and then continue on your mission. So these are all popular features that have come along in our timeline over the last 5 years and, again, direct feedback from our customers. This is another great example of feedback from one of our customers, which was-- in this case, this is a food and beverage manufacturer. And that is canola oil on the floor, which is extremely slippery. And the slippery floor was actually impacting the robot's success rate in completing its automated mission. So what Boston Dynamics did was we took this into consideration and actually integrated software that automatically detects when the robot is slipping. So this reinforcement software allows the robot to detect when it's starting to slip, sort of lower its center of gravity, and have three points of contact at all times. And you can see how much better the robot does in navigating this really slippery situation. So it may be a little bit slower. But in the end, it's really about completion of those missions. We don't want any humans to have to invent in the robot falling down or not being able to complete a mission. And this reinforcement learning is a really good example of how Boston Dynamics takes the feedback from our customers and implements the technology for deployment in their facility. Another great example. I think this is the last one we're going to show, Vatche. But this is really around hazard avoidance. About a year, a year and a half ago, the robot would navigate this environment and step on anything that was 30 centimeters or lower. But here, we see wires and carts and things of that nature. Today, the robot understands this semantic context within its environment and it knows, hey, a wire could trip me up. Even though it's under 30 centimeters, I'm going to walk around that wire or walk around that cart so I can continue on my automated mission for data collection. So let's pivot now into some customer reference cases, as well as some return on investment. Finally, you may be wondering, this seems pretty complex. This is complex technology. We have to program the robot to monitor our environment. How would I ever get this successfully implemented into our environment? And I definitely want to highlight that in our Deployment Services at Boston Dynamics in how we wrap our arms around our customers to ensure that they are implementing the robot and this technology in a way that delivers real value to their business. So at a very high level, on this particular graph, we're just looking at what those economic drivers are for Spot today. It's really around downtime prevention, energy savings, of course time savings, and the reduction in breakdowns within your facility. All of these can impact the production of your facility as well. So when you think of improving gross line yield, all of these things can impact your gross line yield and your environment in a food and beverage environment. And the robot's going out and trying to provide you with the predictive analytics so that you can use this data to take equipment or processing lines offline at your convenience versus having something happen unexpected and really impact the production of your environment. So let's take a look at one of our customers, AB InBev. AB InBev has been using Spot for several years now. They're running approximately 3,000 automated inspections per week. This is on the higher end. We see customers running anywhere from 3,000 to 4,000 automated inspections with Spot per week. Again, these are your thermal analysis, your gauge reading, your vibration, your industrial gas leaks. When they first launched Spot at AB InBev, they identified 245 anomalies detected. This included acoustic as well as thermal. I mentioned earlier a little bit about the mean time to repair or mean time to solution and the liability that resides within those two numbers. AB InBev was able to cut down their mean time to solution by 50% And again, in this particular case, they have a confirmed ROI of Spot of under one year. And of course, that changes customer to customer, environment to environment. But you can count on Boston Dynamics to help support that business case to help you build the ROI drivers that will allow you to implement this technology. Another great example is Nestle Purina performing their autonomous inspections. They've deployed over 16 Spots here in North America. Again, the ROI drivers that we keep hitting on today is around energy savings, the ability to improve equipment efficiency. Also think about those repair and replace costs. So in many instances, we speak with customers that are running motors, pumps, or gearboxes to failure and then replacing them. With Spot, you can identify when a motor is starting to heat up or warm up. Maybe it's leaking oil or maybe there's some friction with the bearings. In that case, you can start to make decisions around repair versus replace. And what we find is a typical repair is about 10% of the cost of a full replacement of that asset. Michelin is another great example. I love to share this one because it's really short term return on investment. In this case, Michelin is publicly referenceable. They're doing the thermal and acoustic inspections as we've discussed. But in this case, they've had the robot deployed for only four months. It's only covering 20% of their plant. And not a whole lot of assets being monitored, just about 700 in this case. But they are able to leverage Orbit, identify trend analysis, and they have actually automated their ticket creation and created over 72 Spot work orders. So again, this sort of gets back to taking the humans out of the loop on a lot of those lower level tasks of going out and trying to use those handheld devices. Let the robot go out, manage the automated data collection on those lower level tasks, and redeploy your humans in a way that drives greater value. In this case, it's more wrench time for those texts, which essentially results in repairs happening a lot faster than if it was more manual process. So we're going to pivot now and talk a little bit about return on investment and how we help our customers measure return on investment. Essentially, what we do here at Boston Dynamics is we're going to go on site at your facility, we're going to help you understand where the robot can actually provide value, and we're going to help build that business case that you can take to your key stakeholders to say, Spot can really make a difference in our environment. So this is a bit of an anonymized use case here. In this case, it is a beverage manufacturing company. And we went on site here and actually looked at three areas. It's around the packaging area, the utilities area, and the production facility as well. So typically, before we go on site, we'll ask for some CAD diagrams to get an understanding of your facility, possibly some metrics around the number of assets that you may have operating in that facility. It gives us a good idea, before we go on site, sort of where we want to spend our time and where we think the robot can deliver the greatest value. So in this case, we did a whole facility review. And what we were able to do is when we went on site is identify the types of inspections that we thought would bring value to the customer. We identified the number of inspections that Spot could get to as well as the number of missions that the robot would be completing, and also what the robot would be inspecting. If we go to the next slide, what our engineering team is able to discover during these sorts of visits with our customers is the total number of inspection points for the robot. In this particular facility, it was just under 3,000 inspection points. I mentioned earlier, we have customers that are in the range closer to 4,000 automated inspections per week. In this case, we've identified just over 1,000 thermal inspections, almost 1,200 leak detection opportunities with the robot, 220 vibration inspections, and 230 visual or gauge reading inspections within this facility. So what we'll do with this data is actually put some math behind it. So we'll extrapolate the number of air leaks that we may have found with the robot when we're doing our initial assessment and start to put some data behind that as far as what is the business case look like if this robot was deployed in this environment. The key ROI drivers or economic drivers for this technology is really around-- in this case, you see air leak detection. But that could very well be other industrial gases such as ammonia or nitrogen as well. Labor augmentation, which we've mentioned earlier in the presentation. Obviously, downtime avoidance-- using the analytics to do trend analysis and identify problems prior to something impacting your production or packaging environment. In this case, there were not any additional contractor costs. But what we do is we come up with what we feel is the annual value that the robot can bring to your facility. And the drivers are around breakdown reduction, air and leak detection, downtime avoidance, and then that labor augmentation. In this case, with a full facility integration, we've calculated a return on investment for this technology just under 5 months. Now I want to be clear that this could be different for every type of environment. It's really a linear path that the robot takes with the more inspections that you're able to do with this robot, the greater return that you're going to get for your facility. So in this case, the total number of inspections is pretty high. It's around 2,700. And it's delivering the most value for this entire facility. And typically, in this site assessment or site evaluation, part of the report is going to be us making some recommendations around integration strategies. So in this case, you can take a look at the three options that we've outlined for this customer. If we do something as simple as just three packaging lines, that's only a total of around 300 inspection points, which lengthens the return on investment to about 38 months. When we start to integrate more packaging lines up to 1,000 inspections, we start to see the ROI continue to drop. And then in option three, when we do this entire facility with 10 lines, the utilities, as well as processing environments, that's where we really start to see the value of the technology being integrated for this customer. So we like to break this down for the customer so they see the full spectrum of what full integration could look like with Spot and Boston Dynamics and what a starting point might look like and what the return on that investment would be. Finally, I really want to take some time to go over our deployment services model here at Boston Dynamics. We've really talked about a lot of different things around Spot and Boston Dynamics in the last 45 minutes or so. But I think one of the most important parts is how do I get started with Boston Dynamics. And we have an amazing field application engineering team here at Boston Dynamics that ultimately will partner with you if you decide to move forward with this type of technology. So our field application engineers are assigned to your organization, typically two engineers for a deployment of a Spot. And we're going to work with you for about two months. We're going to come on site, we're going to help map the facility with the robot. So build those maps that ultimately will reside in Orbit. We're going to manage your networking, IT and security preferences for the robot in your unique facility. And we're actually going to document site acceptance criteria. So this will be mutually agreed upon of what does success look like when we're done working with you, our customer, in fully deploying this robot. So typically, this process takes about eight weeks. But it really is an amazing team process. You're going to get to know your field application engineers very well, and they're going to get to know your employees and your environment equally as well. So I just want to really reiterate that you're not alone on this journey. For most customers, this is newer types of technology. They may be thinking sort of digitally forward thinking in other areas, such as smart sensors and smart factory and sustainability and new supply chain integrations. But the concept of autonomous mobile robots might be a little bit different. But that's what we're here at Boston Dynamics to ensure that you're going to be successful with this technology. - All right. Thanks, Rich. So just to put a bow on it and wrap things up, our ecosystem of Spot and Orbit is really the best way to avoid unplanned downtime and improve efficiency of your facility. With Spot, you're going to get that consistent, repeatable quality of data, and Orbit ensures that your team is going to spend a lot more time acting on that data and less time capturing it. And then at the end of the day, you're freeing your high value employees to take more action and then reduce risk and keep them away from the dull, dirty, and dangerous. With that, there's so many questions, probably more questions than we can get to live, Rich, to be perfectly honest. But we're going to get to as many of them as we can. But we will follow up. We see all your questions. We log all your questions. We will follow up with you on any individual questions we don't get to hear in the live stream. And there was a question, is this going to-- link will be shared. The link will be shared so you can watch this material after. So this question kind of ties to a pre-submitted question as well of what is your policy for Orbit system integration, does BD perform that delivery directly, does it work with common IT integrators. And then a question that I think this kind of ties to, what are the strongest barrier of adoption of this technology. We do see, in that early deployment stages, some of the places where the deployment does get slowed down is getting it onto your infrastructure. And we have, as Rich said just previously on the timeline slide, we've done this many, many times over the last three years. We know the hang-ups. We are your partner in this deployment. We will work with you on recommendations. And we can be as involved as you want us or train you on the robot and then it just depends on case by case basis. But Rich, I don't know if you want to add more. - I might just quickly add to that, Vatche. And again, I can't reiterate strongly how important our deployment services are. This is really the white glove service provided by Boston Dynamics to our customers. This is that 8 week professional services engagement where we will work directly with you on the security the IT, the networking, the integration of Orbit, certainly transfer of knowledge to your employees that are going to be managing this system. So really, really important component. As far as other sort of barriers of entry for this type of technology, I think it's really important to have some change management processes in place. And this is also where Boston Dynamics can really support you. When we go on site to do our site evaluation, we're essentially going to build your foundational business case to launch a fleet of autonomous mobile robots in your facility that's data driven. You will have the expected return on investment for this technology. So that's really helpful when you start to think of building these change models within your facility, within your team, because it's mathematically backed up. You have an amazing team here at Boston Dynamics to support you along this sort of digital transformation journey with robots. - And if you're interested in learning a little bit more about that change management, we've got some great resources on our website. We've got a webinar from last year that talks about this in greater detail. We've got, I believe, a white paper download as well. So there's a lot of material. Again, this is something that we come across a lot. So we've equipped that messaging so that you can have those conversations. This is a more practical question on deployment. So do most customers put multiple charging stations throughout manufacturing facilities to maximize mission length rather than utilizing battery. So a new feature that came with for one that we didn't really call out in this webinar, but multi docking for longer missions is now available. That's a new thing. If you look back to some of our content a few months ago, we talk about that in more detail. So that's how people are kind of overcoming that 90 minute runtime. Is Orbit cloud based or deployed on premises? Rich, you want to take that one? - Yeah, either or, actually. Most of our customers today are deploying in the cloud based within an AWS region around the world. But we also have a site hub. So there are instances where customers prefer to retain all data on site, and we can accommodate that as well. - Great. This is a great question as well. For the companies that we use as examples, was our solution replacing their SCADA system or are they using Spot and Orbit as another source of data acquisition. So I think it's different for each person. And I think the standalone testimonials, which each one of those customers, we do have video content and written testimonials for all of them if you're interested in learning more about their use cases. But it kind of depends. I think when we were working with Michelin, they were working with us, with our deployment team to plug-in to their work order system. And at that time-- and there was another question about thresholds and knowing what to set for your robot. They were using the robot, getting that data, and really, their analysts were analyzing the data so that they could set more effective thresholds, avoid false alarms, and know that the data that they were getting was good. And what was really interesting about Michelin's deployment too, they had their training and then kind of went off on their own and experimented and tried and did their initial deployment, then saw that they got to a certain point and said, OK, now we want to engage Boston Dynamics to take it to the next level. And so when we filmed the story with them, it was before they'd hit that point. But everyone's a little different, I think. And I don't know, Rich, if you've got more to add as well. - Yeah. It's not a magic bullet. It's not a one size fits all. Another way to look at this is, I think in one sentence, Spot is your mobile IoT platform. It has this complete package of sensors that can autonomously navigate your environment and collect information. It would be cost prohibitive to, say, put fixed sensors like that all over your facility to allow you to capture this type of data. Of course, you're always going to have some fixed sensors on some of your most valuable assets. But imagine a partner that can freely roam your facility and take 4,000 different types of data throughout. And that's just in one week. So it's complementary in some ways. And also, as Vatche said, it's a little bit different for everyone. But I think in one sense, the easiest way to think is this is your mobile IoT platform. You're certainly going to integrate some of those other technologies to monitor and manage your environment. But this is a technology that allows you to extend in ways that you never thought you could before. - Thanks, Rich. There's a question. Is it possible to plug-in, unplug the arm as an interchangeable payload. So the arm is not an interchangeable payload. And we'd be interested in learning more about your use case there. But yeah, that's not something that can be added in the field today. How does Spot perform in a dusty environment? I've seen Spot in all kinds of environments. And I'm sure, Rich, you could add to that. But personally, I was out with a cement manufacturer and that's probably the dustiest environment I've ever been in. And I know Spot is driving a lot of value at that facility. And we're hoping to share that case study with you pretty soon. But yeah, in an answer, yeah, it can perform. Rich, do you want to add? - Yeah, I think I know where that question was getting. And that's around ATEX rating for the robot. The robot is not ATEX rated, but certainly can perform in these dusty environments. Some of our favorite stories of some of the experiences that we've had in slippery, dusty, oily type environments with the robot. So essentially, as far as maintenance goes on the robot, there's two fans on the bottom of the robot, and there's two filters that you may need to brush off on a weekly basis or a monthly basis. But there's not a whole lot of maintenance that comes with managing the robot. But if it was ATEX directed question, the robot is not ATEX rated, but can perform in dusty IP54 type environments. - Great. Yeah, and there was a question of-- and you answered it-- how often does Spot breakdown. And there's a great webinar that deep dives into our robustness metrics. We've put a lot of focus. We only had one slide in this presentation about that. But what goes into our testing, how we are focused on that-- yeah, reliability and robustness are our largest focus because we know our customers are coming to rely on the data that Spot provides. And we've improved that big time over the years. And that webinar is a great webinar. I just keep referencing all the other great material we have. Let me see. There's so many questions. And we've got only about 5 minutes left in our webinar here. Rich, and we kind of touched on it earlier, but what are some of the strongest barriers to adoption for this kind of technology that you're seeing with some of your customers? - Yeah, I think change management is really important. I always say, getting the key stakeholders that this type of technology impacts in the same room to discuss it is really important because this is really cutting edge technology and can be confusing compared to other types of sensors that you may have in your facility. So I think that's where we really help out, is helping you build that business case for your key stakeholders in your organization to really truly understand the value that this can bring. And we do customize it to your specific plant or environment. This isn't just fully anonymized best case scenarios. We're going to come on site with our field application engineering team and sales engineering team. We're going to really evaluate where this robot could make a difference. And in the event we don't think it makes sense, we'll show you that ROI data. So to me, this is really the foundational report that's going to allow you to make a decision whether or not the robot is correct or not. We also run into some other physical barriers as well, limitations around maybe some doors within a facility. Sometimes we run into that and doors prevent us unlocking other areas. In the food and beverage industry, we also see situations of cross-contamination. So it's a bit of a broad question on where we see some challenges with the robot. But personally, I think really developing that customized business case is the most important part. We see customers start with one robot sometimes. But when they see the potential of a fleet of robots and the ROI around that with the thousands of inspections that could be happening in your facility to keep your humans safe, to keep your production up and running, and to deliver real business value, I think that's probably the most important piece. - Yeah. I think we've got time for a couple more. So this question, it has multiple parts. Can it detect anomalies while walking or does it need to stop to report? And then they also ask, can Spot detect oil or liquids and spills on the ground. Do you want to-- - Yeah, two great questions. So the robot today is not actively listening and learning in its environment as it's navigating. Essentially, what you're going to do is train the robot to go to your inspection points and do those inspections-- thermal analysis, acoustic readings, and so forth. So we like to say that the robot has this amazing athletic capabilities to go anywhere that you and I can go up and down stairs, automatically avoid obstacles, have this entity detection as well. But it doesn't quite have that cognitive capability that you and I might have when we're walking a facility and we might just see something out of the corner of our eye and say, hey, we should move that pallet away from that piece of equipment or something like that. So today, the robot is trained to go out on the inspections that you want it to complete. - Perfect. - The second half of that question is an area that we're continuing to evolve in here at Boston Dynamics, which is this sort of concept of querying the robot on what it may have seen that day or on that mission, such as maybe a puddle or a door that's unlocked or propped or a 5S board that might be missing tools. So that's our future vision of this technology that gets more into the ability to query and ask questions. But today, that's not applicable for the robot. But hopefully, that paints a vision of where we'd like to go in the future. - Perfect way to answer that, Rich. There's a question that I can take. So when will the humanoid be available? So obviously, this webinar has been about Spot, but there's a lot of interest in our humanoid robot Atlas. Right now, we're really focused on primarily automotive parts sequencing, working with our investor Hyundai on that. There'll be a lot more to talk about there later. And that's all that we can really say today. I'm going to take one more-- - The only other thing I might add to that, Vatche, is it's going to be really challenging to go from 0 to humanoid. So the customers that are integrating Spot today in learning how Spot impacts multiple parts of their business, from the types of people they hire to how they deploy their human resources, how they integrate Orbit, our data platform-- those folks are more likely going to be a little bit ahead of the curve when we start to think of humanoid robots coming into play. So particularly the Orbit platform, we are developing that platform in a way that we will be able to, in the future, integrate different types of robots. - Yeah, that's great. So there were a couple of questions on just the general cost of the robot and cost of deployment. And I don't really want to talk about that right now. But I think if you're interested in the business case that Rich talk through, we definitely invite you to reach out. We'd love to have a one on one conversation about what you do, what you're making at your facility, and how Spot can help. And then we can talk to you, talk through that business case and how it looks for your facility. So thank you all for your time. Thank you for joining us today. Thank you, Rich. Amazing presentation. And yeah, we'll see you next time. Thank you. - Thank you, all.