Beyond the Routine

Hello, everyone. Welcome to today's webinar. We are so excited to have you here to share more about Spot and Orbit features of our upcoming release. My name is Amanda Ganano. I'm the associate director of product for Orbit working across all of our Boston Dynamics robots. And I'm joined here today with Jess. My name is Jess Park. I am a product manager for Spot focusing on facilities management That encompasses industrial inspection, safety, and security. Awesome. Well, I know we already said we're excited, but I'm especially excited to be here today sharing all kinds of new features. We have things that are really cutting across all of the different users of Spot today. Yeah. Five one is really jam packed with features, so I'm really excited to dig in and share what we have in store for everybody. When we think about the different themes that are coming out of this release, I feel like there's three categories. The first is that we're working on features that are helping our enterprise customers continue to scale, add new inspections, and also help them scale in really dynamic environments. We are adding more AI features on top of some of the ones we had in our latest five point zero release, really across the spectrum. And then the final thing is that we have a refreshed new payload for data capture on Spot that we're excited to announce today, too. Absolutely. Before we get in, would you like to introduce to the audience what Spot does, especially in the data collection space, since there are thousands of spots out there working today? Yeah. We have so many different spots out in the world, really around three areas that people are using Spot today. The first is researchers. So Spot is built as a platform. We have a suite of APIs, and we have developer tools. That mean people all over the world are adding on to what Spot can do. So more and more capabilities, not just what we talk about today, but what the research community is really going out and taking to the next level. The second bucket is our public safety customers. So they're using Spot to do things like investigate suspicious areas or go into spaces that are really unsafe and humans shouldn't go, things like explosive and ordinance device disposal or investigating suspicious packages. And then the third bucket that we're going to spend some time talking about today is our facility management and facility intelligence group of customers. They are using Spot to do autonomous data collection, gathering all kinds of information from the suite of sensors that Spot has, and helping them run their facilities more efficiently and better. Yeah. So tell me more about this data collection. How are our customers using Spot for inspection for their preventive maintenance use cases? So Spot today for our facility management kit comes with a number of different sensors that give Spot human like and superhuman capabilities. So today, Spot is equipped with a camera that allows it to take really high def RGB photos that customers can capture and take the same image of over and over again at a really high quality rate. They also can do that with a thermal camera. So they're taking thermal images for things like pumps and motors to understand how those things are changing over time and if something might be wrong. And then the third sensor that Spot has is an acoustic sensor. So this means that Spot is doing things like listening for compressed air leaks in environments that are really expensive when they're leaking and could be detrimental to the facility. All of this rich information that Spot has is then being collected in Orbit. So Orbit is our software platform for all of the Boston Dynamics robots, and you can use Orbit and Spot together to review all the data that we just talked about but also manage your fleet of robots. So send Spot on autonomous missions to go and collect this data and then also be able to take all of that rich information and pump it out of Orbit. I know that we talk to so many customers about the problems of data silos. And Orbit really helps make sure that, whether it's through APIs or webhooks, you're pushing that to all of the important systems that you need. We see that for work order management systems, and we see that for things like data leaks for our data scientists. Yeah. I think we really focus on integrating robots into the existing workflows so that customers don't have to worry about the whole hurdle of adopting fancy robots. One area that we are really focusing on is unlocking the value of all this vast amount of data Spot is collecting using AI. So I'd to learn more about that. Can you tell me more? Sure. So we, of course, wouldn't have a Spot in Orbit release now in, you know, twenty twenty six without AI. Our past few releases have included a lot of different features that are continuing to make Spot in Orbit better. In our latest release, we released Orbit AI Visual Inspections or Orbit AI VI. And these allow users to be able to have Spot take a picture and then ask a question of that picture in Orbit using natural language prompts. So things like, is the door open? In this latest release, we've built on top of that initial launch by adding additional more advanced models into this new Orbit AI VI learning version. And the learning piece is really important because as we were talking to customers, we're starting to hear that they're adding more inspections and want this model to get better faster. Now with Orbit AI VI learning, you can actually have those models be updated without having to update the software for Orbit and Spot, which we are really excited about to be able to give this back and put it in the hands of customers. And we've already seen some customers leveraging these features as well. Right? And just seeing them getting value out so quickly without having to, as you said, not updating software, they just get value immediately, which is so powerful. Yeah, it is really powerful and exciting. And I think we have been able to work with lots of different customers about the use cases they're seeing, and I think it falls into a couple of buckets. One is inspections around safety and security. Is the door open? Are the fire extinguishers there? I'm always amazed at the number of customers I talk to where the fire extinguishers get up and move away. I'm not really sure where the fire extinguishers are going, but in a real life facility, you know, people are making sure that the safety equipment is here to keep their people safe. And now, instead of a human having to do it, Spot can go do that work for them. Similarly, these inspections can be used for materials in material management, so things like debris on the ground. We have a lot of customers who are looking for everything from sunflower seeds to large piles of rocks and debris in certain areas that could pose a risk or a health risk, and Spot can tell a team member, hey. Go check this out. It needs to be cleaned up. And in that similar category, one area that we've added now to this new AIVA learning inspection is being able to check for pallets. So we have a number of people who are thinking about, where is my material moving around in a warehouse? And Spot can now go say, Is there a pallet here? Is there material on the pallet? And how many are in that area? All of that information can tell a team, Hey, how do I move things throughout my facility? Should I have my team go clean up these pallets, or are they able to put more material here? And then the last bucket is asset inspection. So in this new model, we are now able to look at sight glasses. And sight glasses are often not digitized on sites. They can be glass orbs, or they can be long cylindrical tubes that are filled with oil. Maintenance teams have to go and check these on a regular basis to make sure they're at an appropriate level for the machine and aren't falling below a level. If they do, they need to go fill it. That has historically been a really hard visual task, and somebody has to go and look at this constantly. Now Spot can tell you exactly the percentage level of these different sight glasses, send an alert so that a team member can go fix it without actually having to do the rounds. That's incredible. I've walked around some sites, and there are so many of these sight glasses at various locations. So the fact that Spot is now able to capture data and give insights to our customers is very powerful. Definitely. The other theme I'm hearing from what you just said is Spot is now going beyond industry inspection and adding value to other user groups like security, safety, as well as materials handling. So I'm really excited that we are expanding the scope of who we are helping with spot robots. Yep, absolutely. So we are adding more and more inspections for all kinds of different customers and all kinds of different people that are working on a customer site. One of the other areas that we are adding additional data to, and I think a lot of our acoustic folks will be very excited about this, is the ability to see deeper change detection on the acoustic readings that Spot is taking. So now, in our most recent release, we've added an algorithm so that Spot is able to go and take an acoustic reading and then, over time, start to show patterns in the data. If there is a change in the reading that Spot is taking, it's able to alert and show that there may be a potential anomaly happening within that particular reading on that piece of equipment that Spot is doing. What's nice is that you're able to see the changes in orbit in a spectrogram, and that's really simple for anyone to use. But for our more advanced acoustic engineers and people who are doing this day in and day out, they can go in and also tune the charts and what they're seeing and the patterns that they're seeing to be able to ignore things like, hey. The pump is off during this time, but during this time is where we really wanna be able to see any kinds of changes. So this is a powerful feature and not just something you can see in Orbit, but, of course, data that you can pump out into another system, which means you could apply additional AI analysis on top of it or make sure that it's in a work order management system or a data analysis tool, like we kind of talked about at the beginning. That's very exciting that we are not only unlocking data value out of visual data, but also acoustic data. So truly unlocking the power of this multimodal sensing data collection robot and providing all the data power to our users. Yeah. I feel like we've just gone through so many different types of increased inspection capabilities now that we are able to do with Spot. How are we helping our customers accelerate all the new kinds of inspections that they want to take? Absolutely. So last year, we released this feature called Site View. What Site View does is essentially visualizing the site in three sixty degree format. So without users having to step their foot in the door, users can just scroll and navigate around their site from Orbit from the comfort of their office. We've added more features on top of Site View so that users don't have to pick up the tablet and walk around the robot inspections. Now from Orbit, users can virtually author inspections for visual inspections, thermal inspections, acoustic inspections, as well as gauge reading. This truly accelerates the deployment time frame. And also, it can just adapt to the changing environment that customers might have. So say customers have new gauges that got brought up, instead of having to travel to the site and adding these inspections, they can do so from Orbit, which is very impactful. Yeah. You know I love this feature. I'm very excited about it. And I love it because I hear so many customers saying, oh, well, now I want to add more inspections than I had originally because it's so much easier to go back and add more. So it's great to see us continuing to invest in an area that people are excited about, too. Absolutely. The other very exciting aspect is really unlocking the power of multimodal sensing capability. So instead of giving thermal image, visual image, acoustic image all separately, what if we unify all of that to present the data on a single asset in a very cohesive and comprehensive fashion? So now we have multimodal capability where instead of having to have Spot taking all different captures and users having to review that data individually, we are integrating all of those data together and present to the users the comprehensive multimodal data in a single pane, which means that users can get insight on a single target, whether it's visual, thermal, or acoustic in a single place to gain broader and more cohesive understanding of what's happening to their asset. You've talked about how these multi modal captures are making it easier for customers to get a diverse amount of data all in one capture. Is there a new payload that's making this possible? Absolutely. We recently announced a new SpotCam, SpotCam two, which has more power than our previous model of SPOTCAM. So the new SPOTCAM two has four k PTZ camera, twenty five times optical zoom, and also it's integrated with IR camera with the visual camera. What is really exciting and unlocks some of the features that we mentioned earlier is it has an extra accessory bay where customers can insert either SV600 or L642 for the acoustic sensing. And SPOT CAM two is moving around two ten degrees to capture this multimodal panoramic dataset to essentially create a site view that customers can see and sense and hopefully in the future here as well. That's exciting. It makes sense why now we can add all these new inspections SiteView because it's so easy for this SpotCam two to be able to move around without having to move the entire robot body around. So I am very excited to see how this continues to progress and what new things we can keep adding to it as well. It sounds like we've, of course, talked about data and updates that this allows us to do with collecting more types of data, but this also creates a new footprint on the robot as well. Can you talk about why that's significant too? Yeah. Because we freed up where the acoustic sensor used to go by integrating all these modalities. We can now add an arm to a Spot that has acoustic sensing capability so that Spot can sense in multiple modalities and manipulate the environment at the same time with a single robot. Yes. And with this new capability of the Spot Arm and SpotCam two, there's all kinds of opportunities for more industrial inspection solutions too. One of the other features that we've added for our AI inspections is that you can now utilize the gripper cam on the Spot Arm as well as the body cameras on Spot to be able to use those visual inputs on the AI visual inspections, which means that you can take pictures with the gripper cam or the body cam and then ask those same questions about, is the door open? Is there a fire extinguisher? So now we have all kinds of additional uses for industrial customers to add new sensor inputs for AI inspections too. I'm sensing a theme where Spot collects more data, multiple types of data, and do more with the data. Yes. And the other theme that I think we're hitting on, too, is that we're helping Spot and Orbit tie into more dynamic and changing environments, too. That was another key theme of this release. Do you want to talk a little bit about how we're doing that too? Yeah. Spot can open doors now. So we've heard from customers over and over. I see value that Spot is bringing. I want to expand the scope of where Spot can do its work. So we've built a new robot behavior, just like the existing crosswalk behavior and staircase behavior, for Spot to recognize a door, wait for the door to open, and safely pass through the door to continue doing inspections or whatever job Spot might have. It's super exciting. We've tested this with eighteen beta customers, and they've opened the door two thousand five hundred times already, which is super exciting. And it's validating that we hit the right feature and delivering value to the customers already. It's pretty amazing. You don't think about how often you go through doors day in and day out. But in these facilities, there are tons of doors dividing all kinds of environments. With this kind of feature, now Spot can go even more places that humans can go to in these industrial places, so it's very exciting to see. Yeah, no more interns opening the doors for Spot. No more interns hired by Spot. That's right. And there are new behaviors. We introduced two new behaviors to adapt to the dynamic environments that our customers have. Oftentimes, customers have janitorial staff or maintenance crew who need to do some impromptu work, and they don't necessarily work with Orbit. So what we are releasing in five point one is what we call special fiducial behaviors. So there are special fiducial behaviors that people who don't interact with Orbit can use so that Spot can either find an alternate path instead of going through the path that it was already on or sit down safely and power off. This will be really helpful for when there is like an ad hoc maintenance work happening and the maintenance crew doesn't want Spot to walk through that area because it has all the materials that are kind of sensitive, then by placing those fiducials, Spot will intelligently figure out, Okay, I need to take another route because there is some work happening. Let me find another path to get to the next task. Or there may be some situations where there is an emergency where customers want Spot two just pause the submission immediately and not interfere with the emergency mitigation situation. In that case, they can place another special fiduciary so that Spot knows, Okay, this is time for me to take a break. Let me sit down and power off. Yeah, this is great. This is really a way to integrate into existing routines, whether it's placing out a cone that you would typically place out during a maintenance or an event that Spot needs to be around. This just fits into a new group of users to be able to use Spot and integrate it into the work they do every day. So, yeah, what we are really excited about is we are thinking about users beyond the people who actively interact with our robots as well as our software. How do our customers, employees interact with the environment where Spot is in? With these behaviors and the visual inspections we discussed earlier today, we're also really helping Spot support safety teams and security teams, too. Absolutely. So with the AI features and a lot of insights that Spot and Orbit can generate, we are helping more user groups other than just the maintenance teams in the past. So we built a new feature where Spot can do the security missions during what we call third shift. So customers want to utilize a robot more twenty four seven. So outside of the factory for operating hours, they were asking us, are there ways to leverage Spot? And one big use case that surfaced was security rounds where Spot is walking around the facility at nighttime to see whether there are any people who are not expected to be on-site during that hour. So this is super exciting because customers don't have to do anything like new or special, just toggle on the security mission so that during a certain hour that they specify in certain areas, Spot will be extra pair of eyes to monitor the site, even when there is nobody in the facility. That's great. It sounds like we're expanding Spot to help all these user groups with really easy, minimal changes to Spot's routine. And then that data is still all accessible in Orbit, too, for all these kinds of people, or pump it out into the new security systems for these groups, too. And that's been a theme throughout Spot journey. So Spot found its first product market fit in the industry inspection use case. So it helped the maintenance teams and reliability engineers to make sure that Spot is finding any failures, equipment failures early to reduce the downtime. Starting with that, customers quickly saw a lot of other ways that they can use Spot. So now Spot is helping the safety and compliance teams to make sure that the environment is safe for the employees to work. Also, this recent security missions help the security team to keep the facility secure. And the other thing is given all of the data that Spot is collecting, it also helps the data science teams to gather the data repeatably and also for them to do the analytics, create the analytics tools, as well as build their own AI models. So Spot is not only an inspection tool, but now it's becoming a companion or, if you will, coworker to multiple various teams. Yeah. I love it. And especially because the data that a robot is collecting, it's repeatable. It's high quality with the sensors we've been talking about, and it can do it at scale and at volume faster and easier. That's exactly the type of data and data volume collection that's going to be needed to power the AI models of the future. So Spot is really a tool to help customers continue to innovate and innovate faster as well. I think we've made it to the end. That was a lot of features that we talked about in the latest release. We had a couple of key themes. One was, of course, the new payload for data collection, which now enables us to do all kinds of multimodal inspection and continue to build new AI capabilities. All of those new features help our customers scale and continue to scale in really dynamic and changing environments. So at the end of the day, this is continuing to add more value to the existing value case that Spot has and a lot of our customers have. It's been great getting to share what we are excited about for this new release, but now we're gonna just pause, take a break, and answer some questions from the audience. We won't get to every question, but we'll try to answer as many as possible and follow-up outside of the webinar. When you want to ask what Spot's doing for us, it's gathering data that we didn't have before. And it allows us to build a database filled of records of critical equipment statistics that we can go and eventually use to diagnose or prevent machine failures, to help keep uptime, and even safety of the people that work around the machines. We try to do daily machine checks, weekly machine checks, but as your maintenance team gets pulled on hotter tasks, more urgent matters, there's times when those routines and daily tasks can be skipped. You can miss something critical in between those times. We currently have SPOT doing inspections throughout the facility for critical equipment or category one. It's utilizing its SPOTCAM for thermal inspections, but also the Fluke SV600 for acoustic imaging. It's looking at things like motors, electrical components, and it's really capturing the current status of our machines and equipment. There's also a personal sense of comfort that I know that when I'm not here, when my people aren't here, the maintenance crew is thin, there's somebody out there looking to keep an eye on things. Several times a year, we're making big moves throughout the factory. Keeping up with that is very hard. It would probably be a multi week operation with multiple people. As soon as they're done, we'd probably need to scan again because we chained something. Having the dog walk through the factory and routinely capture the facility will provide a lot of insight. You go and you set these missions up and then say we want to run it every quarter, every week, every day. The ultimate value is time. Time is what it saves. All right. We have some questions here from the audience. The first question that people are asking is how is the original SpotCam different than SpotCam two? Yep. SpotCam two is unlocking some two major kind of possibilities. One is instead of doing individual frames, it's able to collect two ten degree, three sixty degree panel images so that customers can get more of a holistic environmental insight of their site. The other one is it's getting us closer to the continuous inspection. So traditionally, we've been doing a lot of discrete inspections, meaning that customers have to know where the assets that they care about are and alter the inspections at their specific locations. But with SpotCam two, it's unlocking the possibility of the continuous inspection. So customers don't have to predefine the locations of where the inspections need to happen. It evolves over time. The environment evolves over time. And Spot is being able to collect the data in that fashion to give more insight to our users. Very exciting. So it's a refresh today, some new features, but also a huge promise for the potential of what this payload is going to bring. I know we are both very excited to see that and keep sharing more as we develop it, too. A second question that customers are asking is, what's the return on investment for Spot? What does it look like when I have it in my facility? For a number of our customers, that ends up being under two years that they see return on investment and value back in what they've implemented Spot for. Our Spot customers are seeing this ROI come from a few places. One is to prevent unplanned downtime by collecting all this information about assets and equipment. Another is energy savings for things like the compressed air leaks that we talked about earlier with the acoustic sensor. A final thing is one we spend a lot of time today talking about, which is safety and compliance as well as security. So combining all these different use cases, we're seeing customers able to get value back from Spot in under two years. Yep. I recently heard from a customer who is so excited about adopting Spot because Spot is able to give eyes and ears in areas where there's very low foot traffic. So instead of having to send humans to those locations, Spot is going there routinely and get data and give the insight as to what needs to happen next, if any. Yep, saving people time is another huge value add for Another question that we're hearing from our customers is, what's the life expectancy of Spot? That's a great question. So Spot got commercialized in twenty nineteen, and those spots are still operating today at the customer sites, so definitely more than six years. What is real value to our customers is we continue to improve on our software, and customers get those improvements for free. The other thing is we have this thing called SpotCare, which is an extended warranty to help our customers manage any of the hardware issues, which has been very impactful for customers who run Spot in very harsh environments and really utilize it twenty fourseven. That's great. So it's true. The old dogs can learn new tricks. Those spots that have been operating for a really long time can continue to get more value out of these features that we're talking about today and continue to have our customers have a clear path towards upgrades too, which is great. The next question is, does Orbit work for Stretch and Alice? I love this question because the answer is yes. Stretch and Atlas will both have Orbit as a key component and as their software platform as well. So when customers start today with Spot and Orbit, they're already on a great path to continue to add more and more robots from Boston Dynamics into their suite of robotics within their facility. What are some of the harsher environments where Spot is being used? What type of maintenance is typically required for a Spot? The harshest environment that I've seen is a cement company. It's very dusty. Cement is everywhere, and the visibility is pretty low. And also, it's outdoor, so the environmental factors also play a role. So we help our customers to figure out how to best utilize Spot in those environments so that it requires minimal maintenance to the extent possible given the environment. In terms of a typical maintenance, we need to make sure that Spot can see clearly, so wiping down the lens and also footrest might wear out pretty quickly depending on the environment, so changing those footrests. And we do this cute thing called spot yoga that makes sure that spot is well calibrated, and we also have spot eye exam to make sure that spot is seeing correctly so that it can continue to have the right visibility in terms of navigation, as well as the data capture. So Spot can help users who have it see all those things in the tablet when you have Spot in front of you. But all of this rich information is also on our support site too, which is available for customers to go, ask questions, and continue to care for Spot over time too. Our final question is, can I integrate custom AI models into the AI visual inspections that we've been talking about? Today, Orbit has the capability to add extensions onto it no matter what your Orbit option is. So Orbit is deployed in three different ways as a virtual machine, as a physical server, or in the cloud. Each three of those have a different way that you can add an extension into it. So you could integrate another model that you have potentially outside of Orbit and Spot to tie that information together. And that works closely with our Orbit APIs and the webhooks that you can use to push data out. So using all those tools, we have innovation teams that are able to tie things together that they want to see also in Orbit. Fantastic. That's all we have time for today. Any questions that we didn't get to during the q and a portion, we will follow-up with you after. For any more information that you're looking for about Boss Dynamics and our products, please go to our website and stay in touch because we'll have more updates for you soon. Thank you.