Autonomous Patrols, Real-Time Intelligence

[MUSIC PLAYING] Thank you all for joining us. I'm Vatche Arabian. Today, we're going to be talking about Spot for security. If you've got questions during the webinar, we've got moderators waiting in the chat for your questions. And then we're going to try to save like 5, 10 minutes towards the end to take some of your questions live. In this first segment, I'm joined by two amazing product managers to talk about Spot in security applications. And then we're going to take a quick break. And in the second segment we're joined by our partner Asylon, who uses Spot in their managed security offering. To kick things off, let's do a quick round of intros. Jess, you want to start us off? Hello, everybody. I'm Jess Park. I'm a product manager on the Spot team, leading enterprise applications. What that means is I make sure that our industrial customers in the maintenance space, as well as reliability space, their pain points are solved, as well as security application where we keep tenants, guests, and visitors as well as security teams safe. I'm Matt Montecalvo. I work with Jess on the product team too. My focus is on the government applications, we call them. So that's basically public safety, defense, nuclear, all those types of applications. It's keeping people out of harm's way essentially. So if you're new to Spot for any application, I think it's important for us to walk through the basics. So Jess, what is Spot? Yeah, Spot is a walking robot dog. It has a bunch of sensors on top of it-- visual, thermal and acoustic. And it walks around the facility and collects the data to give insights to the users. So we have the first product market fit in the industry inspection application, where Spot is looking for and listening for any anomalies that the machines might have so that customers can detect machine failure early before it becomes a problem. And that's become something that we can extend into the security market so that Spock can look and see things for the security team to figure out where there might be security breaches and security threats so that security teams, as well as the community can be safe. So thinking of security immediately brings to mind safety. And I guess if you haven't seen some of our other content on Spot for keeping people out of harm's way, Matt, you want to cover how it's being used to do that today? Yeah. I look at it as Spot's the platform. And to Jess's point, we built the industrial inspection, I'll call it configuration of the robot. And then we also have another configuration which is like a manipulator arm and a variety of different sensor payloads or packages that get put on the back. So that's basically then used in different applications for different threatening or hazardous scenarios, whether it's bomb disposal like CBRNE, like chemical, biological, radiological threats and trying to just distance people from those threats so that they don't potentially get harmed, essentially. And that same ethos sort of applies to security. Security is a situation where people might be in harm's way. And the robot's going to keep them away from that. To dig into that leap or parallel, I guess, an overlap from industrial inspection, how are ways people would use some of the things we've been doing in inspecting assets? How would that translate over to security? Absolutely. So for industrial inspection, what Spot was doing is alerting to the customers, hey, this motor is overheating. There's a puddle on the facility floor that needs to be cleaned up for safety. That same principle applies to security application as well. Spot is walking around the facility looking for suspicious backpacks, suspicious people, and also look for any breaches that might be happening on the facility. So imagine broken windows, opendoor that should be closed. So all of those data is informing the security team to see whether there are any security threats that need to be mitigated immediately, or look at these changes over time to see whether any risk mitigation needs to be implemented on site. Furthermore, it also keeps people safe. Not only the guests and tenants and visitors, as well as employees on the premises by closing those security threats, but also the security teams. As Matt mentioned, robot is becoming the first defense, first set of eyes so that humans don't have to go in there first, exposing themselves to the threat without knowing the context in which they are entering into. Talking about change detection and spot detecting, how does that work? What software is driving it? Yeah, using AI. Everybody is talking about. So we recently announced our partnership with Google DeepMind, and that really unlocked a lot of different models that we can deploy to many different applications. And the speed at which we are delivering new models is a lot faster than we were able to do in the past. So in '25 2, we are releasing new models in addition to what we had before-- backpack detection, human detection, as well as portal detection. And this is very exciting because it's a concept called encounter. Before you had to teach the robot, hey, this is where you are going to look for any anomalies. But with encounters, robot doesn't have to be taught where to look for those anomalies. It's walking around in its security path and see whether there is anything that shouldn't be at these places. Instead of authoring discrete inspections, you are setting like site view inspections so that robot is walking around just looking for the object types. So it can be backpacks, it can be humans, it can be puddles. Or in the ongoing basis new models get developed. So like anything new that might happen. So one thing that I'm really excited about that we are researching on is fence breach detection-- along the fence whether there is any holes in the fence, did anybody cut the fence or did anybody dig a hole under the fence to enter the premise. So that's really exciting that our customers can leverage the AI capability that we have to get more insights around their facility. I think one thing that I like to think about, too, is that enables you to remove the need for someone's eyes on the screen at all times. Because Jessica, you said, if it's discreet, you're really just telling the robot to look at these things. And for everything in between, someone has to be paying attention where you're now removing that need so you can have one person watching multiple robots or just really doing something else, and the robots are alerting them on an as needed basis versus one or two people paying attention to only one or two robots at that time, too. Yeah, that's an absolutely good point. Like somebody from the Security Command Center can manage multiple sites instead of just looking at one site and looking at the data feed. And this data feed is all going into Orbit. That's correct. Yeah, so the other piece of this too, that I find really interesting is for our existing customers who are using this today, this is really unlocking more utility for a different team. I think in the industrial side, we're seeing the big adoption by reliability engineers, maintenance teams, Spots' ownership has been in that world thanks to these advances in AI and our product development. We're now unlocking even more utility for the facility managers and then the folks that are worried about security, compliance, safety, all that. Absolutely. And these teams have their existing platforms that they're using already. So you mentioned orbit. Orbit can have a face, or orbit can be used purely for the APIs and integration capability it has. So security teams for instance they have physical security platform that they already use. So imagine in a movie that you've seen in the Security Command Center, operator is looking at multiple screens with the data feeds coming from multiple CCTVs and things of that nature. Orbit data can be fed into one of those screens using the integration that we made more capable in 5.2, so that it's a single pane of window. Security team doesn't have to go between application. They're getting all the data, but it's collecting as well as other sensors and cameras are collecting in one screen. That's a great jumping off point. So speaking of third party connectivity, what are other ways that can Spot react to a changing environment? Let's just say, motion sensor or motion alert or some other PLC control. I? That possible? Yeah, absolutely. That's something that I'm really excited about, especially in the security application. Spot collecting data about the machine health is only valuable if the machine is already running. And these machines don't run 24/7. So what spark can do using all these integration capability is be informed that the machine is running and go collect the data. And the other more instant application is let's say a fixed sensor detected some anomaly. And that alert can fetch Spot to go and collect more data so that the inspection team or reliability team can send in the right experts with the right tools so that they don't waste time. The similar concept applies to security application. Imagine CCTV detected some motion, but it's not clear what motion-- what caused that motion. Then the CCTV can fetch robot to go to the place to collect more data so that security team can get more context around it before calling the police or dispatching a human security guard to the site. So before we take our break and have Asylon come out, I mean, one thing I also wanted to talk about is on the robot integration side. So our friends at Asylon have pulled it in into their ecosystem and built this great managed security service platform. Can you speak to how our APIs and Spot really works as a platform for folks that are thinking about building their own nn of solution? Yep. So Orbit as a software has a lot of integration capabilities like API and webhooks. The robot itself has integration capability. So we have a lot of SDKs that our partners can use to mount their own payload. And that's exactly what Asylon has done. So they have what's called Pop Pack that is collecting data around the robot's environment. So yeah, robot integration is also super simple. So if the customer has their own sensors that they want to mount on the robot, they can fully leverage that capability as well. So I've had the privilege of seeing Spot used in a lot of different ways. And these new features that we've been rolling out, I've already seen in action. So we just released a case study a couple of months ago with Coca-Cola Euro-Pacific Partners. And in that, they're using it to check lockout, tagout, other safety and compliance applications. But I think arguably, the more exciting recent use case was at the World Cup. Jess, can you talk about that? Yeah, who has lost sleep over World Cup? What is really exciting about Spot's deployment to World Cup 2026 is that Spot was able to do the perimeter security, meaning collecting the data around this environment to look for suspicious backpacks and people that are not supposed to be on site at a specific hour to alert to the security team that there is some security breach happening that a human needs to take a look at. And it was a great success. So GB Jones, who is the head of security at the FIFA World Cup 2026, praised Spot, saying that it increased the security team's efficiency. So that's a really good testimonial that the team leader, the head of the security, saw Spot operating and had a really good experience for his team. The other aspect that is really fun for me is, some of the people who were around the robot was very curious about what the robot was doing and just being integrated with the robot operation into their day-to-day. So that was also a fun part, as somebody who works for a robotics company. Great to see what we're working on in the real world in such highly visible moments. I think it's pretty awesome. Well, we're going to take a quick break. And when we come back, Matt's going to take the stage with our friends from Asylon. [MUSIC PLAYING] And we're back with our partner, Asylon. So let's do a quick round of introductions. Dennis, do you want to start with you? I'm Dennis Crowley. I'm the chief growth officer here at Asylon. Been with the company for several years. But have been a long-time investor, advisory board member, and big supporter of what we're doing here. Cool. And then do you want to go? Yes, I'm Anthony McCarty. I'm the senior vice president of the government portfolio for Asylon. I joined the company just earlier this year. Prior to being with Asylon, I spent a 30-year Air Force career doing operations. Awesome. And then Brent? I'm Brent McLaughlin. I am chief operations officer and one of the co-founders for Asylon. Background's in engineering. And so did a lot of the initial integration of Spot into the Asylon portfolio. But today primarily charged with getting the robots deployed and keeping them operational for our customers. Awesome. Thank you guys. And we've obviously been long-time partners with you, so it's good to finally have you on here. So let's start. We did the first segment on Spot and the different applications we have. But let's talk about you guys. You've built something on top of our platform. And I guess between CapEx and overhead, they're challenging for customers. So how do you guys deal with that, with the different customers that you guys have? Yeah, I think it's a really good-kept secret at Boston Dynamics, but robots are hard. You guys make it look really easy. But we have a saying. Anyone can deploy a robot once and make it work once, but getting it to do 100 or 168-- a patrol every hour, every week after week, is really difficult. And it takes a lot of investment. It takes a lot of expertise to do that. And so Asylon has built this model where we try to alleviate that burden from our clients. We can come in white-glove service, fully managed. We deploy the hardware. We deploy the software. We operate it for our clients. We maintain the robot. It's great technology, but it still fails. All technology fails. And so we just completely relieve the burden for them on that front of it. And then they also aren't stuck with a CapEx model where they have to invest a significant amount of capital upfront for something that right now, a lot of people aren't sure is going to work for them or not. It's just where the industry is. And so we really relieve that burden, and get it streamlined for them and up and running. And they see value really from day one. We have a system operational within a couple hours of being on site. We're doing security patrols day one for them, which is unheard-of for a lot of robots in this industry. And they can start to see that ROI, that value, that return on investment, out of the gate. Yeah, and that's awesome. Let's talk a little bit about the system, too, real quick, just to give everybody an overview. So I'm sure people are familiar with the Boston Dynamics Spot right. We have our different software solutions. But what is it that you guys do on top of the robot that that customer then gets to use? Yeah, it's a great question. So Spot is a very mobile, dexterous, dynamic robot. Has a great software package on top of it. Asylon basically came in. We strapped our own payload on top of that, put our own security payload. So we have a 20x optical zoom. We have a integrated thermal sensor in there as well so I can see heat signatures, so I can see warm bodies hiding in foliage. I can see warm engine blocks on cars. What's really novel about what we've done is, it allows us to operate the robot completely remotely. And so we have robots from Hawaii, across the country, all the way up to here in Massachusetts, that's operated from what we call the Robotic Security Operations Center, or the RSOC, which is at our headquarters just outside Philadelphia. And I can control a robot, really, from anywhere with an internet connection. And so we've taken the phenomenal Spot platform and really unleashed it. And now I can operate it from anywhere. I can stream video and telemetry in real time to anyone with the right access, of course. [LAUGHTER] And really, it unlocks a lot of potential for the robot, that I don't need a security guard or an airman walking behind a robot with a tablet to control it. Now it's much more dynamic. You have economies of scale. You have a lot more efficiency and it allows a robot to do a variety of mission sets. Right, and you couple ground and air, correct? We do, yeah. So that's actually where we got our start. Myself, Adam, and Damon, my two co-founders, we met at MIT doing aerospace engineering. And so we wanted to do drones. Everyone wanted to do drones in 2015, 2016. And so we started there. We picked up an Air Force grant, a research and innovation grant, that said, hey, we love your software. We love your drone. It was a different robot, of course, which is why they were having problems. They had a different ground robot platform. They're like, I have to have an airman walk behind this thing to get it to do anything. That's not scalable. I can't use that. Could you guys integrate? We're like, yeah, we can. And so we did that. It worked really well. And then we went to commercialize it. We reached out to Boston Dynamics. We needed the really top-tier support to make this scalable. How do you take a robot and deploy it at an enterprise level across the country-- across the world, potentially-- and be able to support clients in that type of environment where it's a no-fail mission? Security doesn't get a lot of glitz and glamor unless something bad happens, like your IT team. No one really cares about IT until they can't print or their computer doesn't work. And security is the same way. It's fine until it's not. And so you need a backstop of the really great support team at Boston Dynamics. Asylon builds on top of that with our support and our own operations team. And that's been really powerful for us. Yeah, I'd imagine so. You have a wide breadth of customer types. So do you have any testimonials, Dennis, like commercial customers, potentially, that are worth talking about? We've got a long list of great customers to talk about. And as Brent said, they go from coast to coast, north to south, across the United States. The first one I always think about is, as a big sports fan, Mercedes-Benz stadium. We've been operating there for almost four years now. They had a challenge. So the old footprint of the Georgia Dome stadium became a public park. It was closed to the public from dusk till dawn. A lot of vagrants. A lot of issues of thefts, petty assaults. Security officers were walking it, patrolling it. It's a tough environment. You're getting yelled at, spit on, thrown at. Not the place you want to work. So what do you do? You call in sick. You look for other opportunities to go from there. We put what we call DroneDog Spot with the Asylon platform in. It started patrolling. And amazingly enough, the challenges and the problems disappeared. People went down the street. It's got a strobe light on top that will drive you crazy if you're just trying to hang out. It doesn't care if you call it names. It doesn't respond. It can see what's happening. We can talk through it. We can have pre-programmed messages. We can hear and interact with someone remotely. If we see something that's a concern, we can call the on-site security personnel to come respond. What we've done is take advantage of what the robots do well. And then supporting them in the back end with the RSOC that Brent mentioned earlier, where we have humans that are trained to operate the robots but are also trained security professionals. So we have humans doing what humans do well, which is making decisions off of good information and situational awareness coming from the robots. And we have robots doing the dull, dangerous, dirty tasks that humans don't like. Another example is-- we've talked about in the past, Bayer Agra. We've been out in Hawaii and now California, walking out in the farmlands where they're testing and growing along large property lines, doing the routine, dirty patrols. We've got construction sites. We have a great video we use. It was January. It's 23 degrees. It's cold. It's snowy. And the robot, walking back, it has mud up to-- I'll say its elbows. [LAUGHTER] The ultimate visual of the dull, dangerous, dirty. No human wants to be walking in 23 degrees, snow and mud, and doing it every hour consistently, in and out. We've got corporate campuses. My favorite is, we have a corporate campus here in the New England area. And when I go to the site, it's posted at the front entrance as employees are coming in in the morning. And you see employees just waving to it. It's part of the deal. And when I'm down in our RSOC in Philadelphia, I'll often see the guys there, see someone wave to it. They'll grab the remote, give it a little head tilt, give it a little personality and charm. People are asking, hey, where's DroneDog? What's going on? Everything OK? I haven't seen him on his patrol. Oh, no. He's responding or doing this or that. It's become part of the team and part of the personality in the campus. And that customer specifically says it's been a great tool for them in recruiting personnel because they see the organization as forward-thinking, not just for their security staff, but also for the company as a whole. So definitely it seems like there's an effect above and beyond security, too, for the customers that you guys are offering or provide. And one thing I just thought of, too, while you were saying that, is you have a tremendous ability to scale, too. Because you're operating-- you said, it's out of just Philly, correct? Yeah. But you're doing things globally just from your local area, which is pretty impressive. And then I guess you guys don't just do commercial, right? You do defense, Anthony, which is why you're here. Do you want to dive into that a little bit? Yeah, that's exactly the right in. I would argue that Asylon is a dual-use company in the truest sense. Half of our work is in the commercial sector, but the other half is in the government sector, working across the services with the Department of War and some of the other government and federal agencies. And the versatility that Denny talked about of the systems, both the Guardian airborne system and the DroneDog system, have been very appealing to our government customers. As a matter of fact, DroneDog and the Guardian systems are deployed today, operationally supporting security operations in some of the Air Force's no-fail missions that are going on today in some of the most austere locations. So I smile every time Denny tells the story of the versatility of DroneDog and Spot and what it will do for the customers because that exact value proposition is why the services have such true interest in it. And the applications are a bit endless. In addition to security operations, just earlier this year, again, on the Spot platform, but with our Guardian small UAS system as well, we changed some of the payload out that Asylon was utilizing those basic platforms with. And now we're on a contract to do depot-level inspection on aircrafts for the Air Force. And so think of it in terms of, you have a 747 or you have a C-5 aircraft with a DroneDog underneath the aircraft. And the Guardian flying above the aircraft doing digital mapping of that aircraft taking what would take the manual operator-- that 22 to 24-year-old that they're putting in a 65-foot cherrypicker up to do a manual inspection of an aircraft tail. We can take that down from four to six weeks to almost five hours. And so that is an incredible impact on the overall readiness of the services and the overall readiness of the Department of War. And that is why we are seeing a consistent interest in the ability to be able to grow that opportunity to additional locations, not just within the Air Force, but potentially with other services. Nice. And on that inspection task, too, I'd imagine you're reducing the time, but you're also reducing the risk. You're putting a young kid up in the air. Now you don't have to do that, which is probably a good value prop for them, too. And not only are you doing that, but you're building consistency in how those aircraft are inspected and how that information is funneled and channeled in to the experts. And you're freeing those service members up to do the things that only humans can do associated with depot-level maintenance. Yeah. I think that's the key to everything, is the robots do these dangerous tasks or routine tasks that we don't like to do. I spent enough time on my days as a security guard. My theory is, I walk at midnight and do a patrol. I see everything that's there. I walk at 2:00 in the morning. I see what I saw at midnight. I walk at 4:00 in the morning. I'm not seeing anything. I'm just walking around and going from there. The robots are seeing it. Now think about it. It's a security guard. It's a-- Service member. --a service member who has a body camera on. It's recording everything. From the legal liability standpoints later on, the value proposition is outstanding. But the quality of the patrols, the recordings, the pieces-- and then the response from the human in the loop who's actually making good decisions, who's not worried about fight or flight if something dangerous happens. They're sitting comfortably in Pennsylvania in a heated or air-conditioned room, depending on the time of year, being able to make good decisions and notify the right people to respond and react. So would you say you're basically doing repeatability, reliability, and then putting time and distance between basically dangerous situations-- or just situations in general-- and the responder, essentially? Exactly. Denny said something I think is worth mentioning, and Brent alluded to it-- is the robotics as a service model with RSOC up in Norristown. We basically are packaging up that same setup, and we're deploying it on premise to the government locations, taking into account what their common infrastructure requirements are, their operational security requirements are. And so we're packaging it up and tailoring it to each of the services and that specific mission need. And I think that's where we're seeing the greatest value to the service members, in the ability to maintain readiness or maintain overall protection of some of those no-fail, critical missions. It's a powerful tool. My favorite story of all-- that construction site I mentioned earlier. We have a video. The robot is on its routine patrol. Looks to the right. It's in thermal, as Brent talked about. Sees three heat signatures of bodies. Alerts the operator, who takes over operations and starts walking towards them. They see the robot coming. It's 2:00 in the morning. It's the middle of the night. They run back through the fence line. They disappear where they came from. At the last second, you see someone, a body reach back in. It looks like they're holding a cell phone. Because even the bad guys want to get a picture of the robot because that's how-- I got caught. But hold on. Let me take a picture of it. That's awesome. That's cool. That's really funny. Nothing more incriminating than video evidence of me trying to steal from the site. Who are the guys? Well, how about the guy who posted on Instagram, the robot on the construction site? We know we have a lot of questions coming in, and we'll take a minute in a little bit to get to those. But there are some questions that we get asked a lot. And Brent, if you don't mind answering-- so robots for security is fairly new. So how would you address that with someone who's coming to this? Yeah, that's a great question. We get that a lot. People new to the industry, new to technology. And technology and security isn't new, but there's always new technology and security that's coming out. For us, we deployed our first DroneDog based on Spot in 2021. We're basically five years into running patrols with DroneDog now. We've done 350,000 security patrols. We've done over a quarter million miles-- 250,000 miles-- of DroneDog. This is real. It's a technical readiness level nine, like we talk about in the government side of things. We're out there producing value every single day. Our Robot Security Operations Center, that RSOC, is doing between 400 and 500 patrols across the country every single day. We've done over 7,500 security events. And a security event for us is anything from an open door to a trespasser to a fire to a potential theft or a thief. We've caught all of those. We've been part of all those types of events. And so our customers today-- and they're across the country-- are seeing true value, true ROI, that return on investment already every single day. And I guess a better way for me to have phrased it is probably, it's not new, but it may be new to you because you've clearly been doing this for a while. And you talked about ROI. How would you, or all of you guys, describe that to your customer base? I think the way we've priced it and put it together in partnership with Boston Dynamics-- and Brent talked about it earlier, the robot as a service model. So the initial fear of hey, is this going to cost me a few hundred thousand dollars to try this to see if this works or makes sense? We come in, and we charge much like buying a security officer. We charge on per patrols per week. And you're going to find the price range is going to be similar to that of buying a security guard. Actually, you're probably going to see it less than buying a security guard these days because labor keeps going up and up and up from there. So you can see that ROI immediately. And we also de-risk the challenge of trying to find that ROI by the way we've priced it and put the model together and the structure. It's a service. You're not buying a robot. You're not licensing software. You're subscribing to a managed service that is delivered by the robot and supported by the software and managed by a human in the loop. And all of that comes together in one package for the end user. So you're lowering that barrier to entry that most customers wouldn't really want to, like you said, spend a ton of money out the door to see if they like it. And as we're seeing the rising cost of labor-- but more importantly-- and you see it in the government side and the commercial side-- is the supply of labor. Just, the bodies aren't there. Yeah, across the government and the federal agencies that we're partnering with, across my entire career, you always hear one thing, is, the federal agencies and the government and the services are undermanned, under-resourced with what they're asked to do. And so the true ROI on the government side of things is that we can provide that consistent and persistent enhancement to their operations, whether it's aircraft inspection or security operations, with these resources that really free up those limited human resources and personnel that they have to do the things that only humans can do because they've got that consistent presence of the DroneDog and the Guardian systems out there, executing the mission day in and day out. That's awesome. If I can-- and we talked a little bit about Mercedes-Benz stadium. And I want to talk about it a little bit more because it's really just a phenomenal use case and a really great story. They originally came to us. They met us at a trade show. And they talked to us for a few minutes and walked away and came back about half an hour later, like, we have this seed of an idea. And they basically have a statutory requirement to provide security where the old Georgia Dome is, which is now called the Home Depot Backyard. It's a beautiful park. But it closes, as Denny mentioned, at dusk and doesn't open till dawn. And so you can still access it, but you're not supposed to. And they'll come and escort you off. And they came and said, I have a security team. And he's like, candidly, it's mostly women at the time, overnight shift. They're not armed. They have a flashlight, maybe. And they're walking around this dark, urban environment stumbling upon issues. And he's like, could the robot take over those patrols for them? And we said, absolutely it can. He's like, I have zero interest in replacing the human. I'm not trying to get rid of my human guards. He's like, I just want to make them safer and give them a better quality of life in their job. And so at the time, they wanted to trial it for a little bit, which goes back to this robotics as a service model. They don't have a huge upfront cost to sink into this to try it. And so we came out. And in the first two weeks during a normal patrol-- it was around midnight, we come across an individual. We actually picked him up on thermal. We saw his body heat, and so we called it in. They knew exactly where he was. They knew what he was wearing, what he had on him, what he was doing. They deployed two security guards to his location. He could see the robot. It was very obvious that he was being monitored, which is different than CCTV cameras. I think all of us today, we're desensitized to cameras, but we're not desensitized to robots yet. And so in a way, it was a de-escalator because he knew he was being monitored. The security guards knew that they had that backup as well as this robot being there and filming things. They were able to talk to him. And he left. He wasn't doing anything malicious, nefarious. It was a hot day in the summer. It was a beautiful park, and he stopped there. But you can't be there. You've got to leave. And so everyone waves at the robot. He walks his way. The two guards go back inside. And we do the patrol about 10 minutes later. We find him again. He actually didn't leave. He just went to a different part, a darker part. But we pick him up on thermal again. And so we call it in, and they're like, oh my gosh. And so they sent the two guards out again, and they actually escort him off-property this time. And we get a call the next day from the security director. He's like, this is perfect. He's like, this is exactly what I wanted. My people are not stumbling upon issues and potentially violent or dangerous situations. The robot is giving us so much situational awareness that now I can respond appropriately. I can see if they're armed. Like, do I need to call law enforcement? Or can my security guards go out there and deal with it? They've been champions ever since. And just all of the resources you talked about in there-- I just think of one of our clients that we're doing exterior patrols. That's the primary duty. But as part of his emergency response plan, should there be a situation that occurs, you can call us up. We can teloperate it. We can walk inside. His whole idea, active shooter, the first responder on site is going to be the robot. Which he wouldn't have if he didn't have the robot on site doing all these other duties. It's there for routine tasks, but it's also there for those emergency response and creates that additional value prop. I think this has been a great conversation. Thank you guys for coming in and having this. You guys have been a great partner from Boston Dynamics for a long time. We're going to take a little bit of a break, and then we'll answer some of your questions. [MUSIC PLAYING] And we're back. So we're going to get to a couple of your questions. The moderators are in the chat answering questions live as well. And we're keeping a couple for us to discuss here as a group. Let's start with, where do you see security applications going for Spot? This is really a little roadmap-y, a little vision in the future. Maybe start with Jess. Yeah, so in terms of roadmap, we want Spot to be the conduit between the Security Command Center and the person or people by the robot. So one of the things that we're working on is a two-way audio capability so that the operator at the Security Command Center can communicate with the person by the robot. And what I'm really excited about personally is what I termed Sarah's walk home late at night. So today, in the current state, in the parking lot or parking garages, there are static blue light phone that you can walk to to ask for an escort to walk back home safely. Imagine you're Sarah, coming home late at night from work. And nobody's around, and it's really dark, and you don't feel safe. What you have to do today is, you walk from your car all the way to the blue light phone, being exposed to potential threats, pick up the phone, get the person come, walk back to the car, wait for a stranger. And you have to verify that person is an actual escort. And when you're walking back home, you have to engage in an awkward conversation. I don't know about you. I don't like engaging in a conversation with somebody that I don't know. What I'd love for Spot to do is, as you arrive in the parking lot, as Sarah arrives in the parking lot, Spot recognize that there is a person who might need an escort, walk to the car, and just wait for the command, whether that person, Sarah, needs an escort. When she does, Spot walks with Sarah back to her apartment, back to her home, while collecting the data so that somebody from the Security Command Center can monitor the situation and intervene if necessary. So Sarah is never exposed to unnecessary security threat, and she didn't have to engage in a very awkward conversation. So that's the future that I'm really looking forward to. I think for Asylon, like many companies today, we're investing a lot in artificial intelligence. How can we utilize Spot and the sensors on Spot to detect anomalies and other security events? Today, we have a human in the loop, and I don't expect that to change moving forward. But it still relies largely on the human to detect those types of events. And so how can we use computer vision models and large language models and things to catch those types of anomalies, to just enhance that ability of the human operator? We have some of those initial capabilities today, but we're just putting a lot more resources into AI this year. Something's that's a little novel about Spot-- and it's a pro and a con-- is, it's basically, for security purposes, more or less a mobile security camera. And there's a lot of value in that. And we've seen a lot of value in that. But a lot of the security analytics that exist today on your static CCTV cameras, they rely on the camera not moving or moving very little from a fixed point. And so a mobile camera just has a lot of dynamicism to that and makes it a little bit more difficult. But we're tackling that head-on and showing a lot of good progress there, too. One question that's come up a few times in chat-- and actually, I've received this question, too, at trade shows-- but what happens if something happens to Spot while on security patrol, if it's damaged or even stolen? Yeah, so for Asylon, we have unfortunately had that happen a few times. But that is the beauty of our model. The client generally-- unless it's something particularly egregious from their end-- they're not on the hook for that. We're there in a day or two with a new model. We'll swap it in, and the patrols are back up again. No-fail mission. For a security asset, we need it up and running. That's the most important thing. So we handle that. Generally-- knock on wood-- we've not had anyone try to steal the robot from any of our client sites. We do have some that are a little bit more public. And so it's something we're conscious of. We take a risk assessment for that. But the robot is a little heavy. You're not going to pick up and run with it for very long. At least I can't. I don't know what you're going to do with it. The pawn shop isn't going to give you a whole lot for it either, I hope. And so we haven't had too many issues with that. But I've had people square up against Spot that try to fight it and that sort of thing. Spot, we don't fight back, obviously. We're an observe-and-report asset. But for the most part, Asylon handles that, and it's not an issue for the client. I guess another question that we've seen come up over the last 30 or so minutes is, how long does it take to deploy something like this? And maybe we start with third-shift security or industrial customers. How long does it take to deploy Spot? Yeah, our typical deployment looks like this. So a sales engineer goes to the site to scope out the facility to see what needs to be detected, what is a parameter, and what kind of sensing capabilities are needed for Spot to be the most effective. And then our field engineers come to site with the robot and record the map as well as other inspections. And as of 5/1, the authoring capability got a lot easier. So before, you had to record everything through tablet. But now we can record other things through SiteView, which means that you just do it from the comfort of your own office through the Orbit. That takes about a week, plus minus some validation period so that the field engineers can leave the site with the conviction that the robot is not failing during missions. And we have SAT, where we monitor the mission's success rate over time to make sure that everything looks great and hand off to the customer. And how about on the Asylon side? I think Asylon, we're very similar. We didn't talk about it too much earlier, but we have a doghouse for Spot. It's this clamshell design where Spot can go in and charge. And so we ship all that to the client's site in a big wooden crate. We have a deployment technician that shows up the day of deployment. Within a couple hours-- the doghouse-- it's very easy, plugs into your standard outlet-- we're plugged in, and we're running missions. Usually two to four hours, depending on the setup and environment. For us, our deployment technician is usually on site for about three to five days, depending on the complexity of the site. Our very first patrol-- again, that two to four-hour range-- we generally ask for someone from the security team to walk with our deployment tech and the robot who will be controlled from our office in Philadelphia. It's someone who knows what they want Spot to do, what is the actual mission. And we've talked about this. We've had pre-meetings. But there's something about being on site with the robot that really just changes that dynamic. And so we walk with them. They point out everything. And we call it scripting. We script our own automated patrol. And then our deployment tech over the next several days will be on site varying hours, depending on the day, to help train our operators back home. And then they're off, and we run everything remote. And so like Jess mentioned, we can also rescript and modify the patrol as necessary from the comfort of our own office and our own home, and we can do all that remotely. But it's, again, very quick to first patrol and generating value. As a final question, people are asking, how can we get started. So how can someone get in touch with Asylon if they're interested in your solution? Yeah, we'd love to reach out. asylonrobotics.com is our website. From there, it's very easy to schedule a demo, schedule a chat with someone from our team. We can talk through what your needs are. For our perspective, we don't want to deploy something that isn't going to provide value and isn't going to solve a problem. Spot's cool. Robots are cool. But they're not toys. They're actual tools doing real missions, providing real value today. And so we'll be up-front if it's a good fit or not. It generally is, so no worries there. But we want to be partners in that perspective. We want to be a trusted partner on that front. And so we'll talk through what the capabilities are, what we think we'll be able to do with DroneDog or our Guardian system, and then work through that process with them. Yeah, and in a similar token, if you're interested in some of the solutions we've been talking about from Boston Dynamics Direct, you can reach out via bostondynamics.com in our sales form. And similarly, we'd be happy to evaluate and learn a little bit more about what you're trying to do with Spot. Thank you. Thank you to our panel. Thank you to Asylon for being here. Thanks, guys, and we'll see you next time. [MUSIC PLAYING] PLAYING]