# DJI Dock 3 Deployments at Scale, and a Sneak Peek into What's Next - Event: NestGen ’26 - Published: October 2, 2026 - Length: 30:29 - Watch: https://nestgen.org/library/watch/NJUHaQkteuM/ > The transcript of a recorded NestGen session, for reading or for use with an AI tool. It comes from the video's > captions; automatic captions can mishear names and terms. ## About this session Taking a drone-in-a-box program from a single site to deployments across hundreds of locations worldwide surfaces lessons that only show up at scale, and points toward what needs to come next in the hardware and software stack. In this NestGen live case study session, DJI Enterprise will share how DJI Dock 3 has been deployed at scale across customer sites worldwide, the lessons learned along the way, and a sneak peek at what's coming next for the platform. ## Transcript **[0:01]** Hello everyone, welcome back to another very interesting session here at Neschen. Uh do let us know in the chat where you're joining in from. I'm Bishali, your host for today and I am a host for a session that needs very little introduction. Uh when they launch a product, the whole industry collectively stops and pays attention. And today we are not just talking about what's already out in the field, but we have a super exciting sneak peek coming through. So please join me in welcoming Michael uh solutions engineer at DJI. Michael has spent more than 8 years in the industry and he lives at the sharp end of it. He works hand inhand with energy customers across North America, oil and gas, utilities and wind, helping them actually put these technology on to work on real sites. And if you've ever watched one of his deep dives on the Matrix series, you already know he explains the hard stuff better than almost anyone. Today, Michael is going to **[1:05]** walk us through the DJI Dog 3 at scale. All day here at NestGen, we've probably watched pioneers after pioneers showing their deployments across energy, ports, mining, public safety, and every kind of application you can imagine. DJI Dog 3 is the machine running underneath so many of these stories deployed across thousands of sites around the world. Michael will take us into through what running this at scale really looks like and then give us a sneak peek at what DJI has coming next. I cannot wait. But before I pass it over to Michael, I have a few quick housekeeping rules uh while you settle in. Uh please feel free to type in any question for Michael uh during the session in the Q&A tab and we'll get to them at the end. The session is also being recorded and will be published afterwards. And if you do capture any of your favorite moments uh during the Nesgjen session, do share them on LinkedIn using hashtag Nesgjen. A few lucky people might have a surprise in store. **[2:08]** With that, a very warm welcome to you, Michael. I will pass it over to you. I'll stop sharing my screen and you can now take it over. >> Thank you so much for the introduction. I just want to make sure you're able to hear me loud and clear and you're able to see my screen. Yes, everything is good. Go ahead. >> Perfect. Thank you so much. All right. Hello everyone. My name is Michael Lee and uh I'm in I'm the drone in the box specialist for DJI Enterprise. Uh today I'm really excited to show you guys how the DOC 3 is changing remote operations all the way from routine inspection to time critical public safety uh use cases. I will keep the presentation about 20 minutes and leave about five minutes for questions at the end. So, here's the plan for today. First, we'll take a quick look at the DOC 3's key capabilities and specs, and then I'm going to be able to show you two real world use cases from solar hotspot inspection, uh, and DFR, which stands for drone as first **[3:08]** responder. Uh, they're very different applications, but they rely on the same strengths, the automation of the dock, repeatability, fast response, and reliable data. So, we're going to start with the environment. So, the Doc 3 is IP rated 56 and the drone itself, the Matrice 4D and the 4TD is IP rated 55. The dock is able to operate in minus30 to 50° C while the aircraft operates from minus 20 to 50. It also supports strong wind resistance during flight, takeoff, and landing. The point is not just the rating. The point is essentially that this system is designed to stay outside and remain available instead of waiting in a warehouse for someone to deploy it. That permanent availability changes the operating model. Teams no longer need to spend the first hour transporting, assembling, and checking equipment before they collect useful data. They could just take off. [snorts] Endurance. Endurance also matters because every extra minute means more **[4:11]** useful coverage. Anyone that's flown drones before knows this. The Matrice 4D series can fly for up to 54 minutes with up to 47 minutes of hovering. And the dock is able to recharge that battery from 15 to 95% in about 27 minutes. That combination supports frequent missions with short turnaround times. So these are controlled test maximums, but they show the operating model. So longer missions, faster replenish, and less downtime. It gives operators more time observing set or target and keep a safer reserve without sacrificing any useful parts of the mission. Now let's move to our first use case solar inspection. Typically a solar farm is more than just panels. You're going to see inverters, electrical systems, batteries, even a substation nearby and the supporting infrastructure. We normally combine two inspection types. So you have your thermal inspection that finds hot spots and **[5:13]** underperforming modules. You also get a visual inspection that checks vegetation uh any of the racking, wiring, roads, towers, and physical damage. One flight can give the maintenance team both performance data and visual context. A temperature anomaly can tell the team where to look, and a visual image often helps explain why it exists and what should happen next. Why use a dock instead of a traditional drone program? Well, first is automation so that takeoff, landing, charging, and mission execution happens without a pilot on site. The second would be the availability. Missions can run on a schedule or on demand around the clock. Third is consistency. The same route and camera angles produce comparable data every time. And finally, efficiency. One remote team can support more sites with less travel and preparation. So traditional inspections often depend on who's who is flying, how much time is **[6:16]** available, and when someone can travel. The dock removes much of that variable, and the result is more coverage, better consistency, and less downtime. So here's a here's a typical workflow. First, you would need power, network connectivity, and enough open space for takeoff and landing. So, typically, we've seen customers use connectivity with a hardwire Ethernet, Starlink, or another suitable link. Getting this foundation right is important because automation is only valuable when the site is designed for reliable daily operations. Before detail inspection, we normally create a current map or 3D model. A satellite image is useful, but is often old or may not show recent changes. We usually have a quick mapping flight to give us a current planning surface, so later routes can be more accurate and safer and easier to plan. It also gives operators and maintenance teams one shared current view of the site instead of separate maps and assumptions. So, this could be done once the dock is set up. You just do a quick flight and you **[7:19]** process the model in Flight Hub 2 and then you can use that model now as a mission plan. Terrain follow also matters. So like when a site is not perfectly flat, hills, bmss or any drainage areas can change the distance between the aircraft and the panels. So terrain following is a useful tool that we have that keeps a more consistent height above the ground, protecting clearance while keeping image scale and thermal quality more consistent. Not to go into too much details here, but at a high level, the value here is not the individual flight settings. It's the ability to produce repeatable decision ready data at scale. So DOC 3 standardizes coverage, camera angles, and positioning across every mission. With consistent capture and RTK precision, teams are able to compare the same asset over time, locate anomalies quickly, and send maintenance crews to the right place with confidence. This repeat repeatability changes inspections from one-time snapshots into a continuous view of asset health. When the drone follows the same routine every week **[8:23]** or every month, teams can see whether a hot spot is stable, getting worse, or has even disappeared after maintenance. This makes trends easier to identify and maintenance results easier to verify. During each mission, you're going to be able to collect RGB and thermal data. thermal is going to show that anomaly temperature pattern while the RGB provides context behind it such as dirt, shade, any any debris or physical damage. So, this is a good video of uh showing some solar panel uh inspections. This is able to combine both views and reduce any unnecessary follow-up visits and give the maintenance team stronger evidence for prioritizing repairs. So this this video right here is going to show how Flight Hub 2 brings the operation together in one place. So the operator is able to monitor the aircraft mission progress and live camera output remotely. The routine flying is automated but the operator can always take control and maintain situational awareness at any time. So **[9:26]** if that's weather, airspace or uh site activity changes, the operator can intervene at any time. Next step is moving from scheduled inspection to eventdriven response. So with autodispatch in Flight Hub 2, this is a system that's able to get an external alert such as a SCAD event. It's able to trigger a predefined mission automatically instead of waiting for the next schedule inspection or even sending someone to the site. The drone can verify the situation within minutes. So you can see this is the auto dispatch and we're able to set this up. We can have a expiration time, task name, and basically once that trigger is received in FlyHub 2, you can tell the drone and the dock exactly what to do. So what altitude to fly at, return to home altitude, this can all be set so that everything can be automated. [snorts] So right here, they're still uh playing with the settings. **[10:31]** I'll fast forward it a little bit, but once everything is essentially uh set up, you could save it. And then from there, if we do get a trigger, the drone will take off and do the actions that it was programmed to do. So this is an example of it taking off and then flying to its allotted location. Okay. The ne uh the next uh workflow I want to show is the POI workflow. It's another simple example. So when a specific point needs attention, the aircraft is able to fly to that location and perform a a targeted action such as an orbit or focused data collection. So this is going to be an example of that. The details are all **[11:32]** configurable as you can see here, but the business value is pretty simple if there's a detected issue uh rapid verification and a consistent record without building the mission every time. So, for example, once it's all built out, let's say there's an asset right here, we were able to rightclick and fly up two, and the drone will take off and start doing a POI. So, you can see POI right there. It'll just ask for a title. From there, the drone will start taking off. and the drone is is approaching and once it gets there it will do its action point of interest. **[12:35]** So you can see it circle Okay, a typical output that you might see is some soiling, cracks, shading or damaged components and even thermal signatures can be used to detect hot spots. After capture, the RGB and thermal data can move automatically into analysis platform such as sitemark or into the customer's own storage and analytics stack through FlyHub 2. So this removes any manual download steps and standardizes the handoff from flight operations to analysis. DJI does not force one analytics tool. The platform is designed to fit the workflow of the customer that they already have. The key message is pretty simple. The dog 3 connects detection, verification, action all in one workflow and customers get faster awareness, fewer unnecessary site visits and more consistent evidence for maintenance decisions. It is not just about automating flights. It's about helping **[13:35]** teams identify issues earlier, respond faster, and operate more efficiently. Now, we're going to switch into from scheduled inspection to urgent response. So, you may have heard this term thrown around a lot now in the drone industry, DFR, or also known as drone as first responder. I'm going to dive a little deeper so you can have a better understanding. This slide shows why the DOC 3's 10-second takeoff matters. The updated logic shortens the selfch check process, keeping GNSS ready and supports emergency launch without unnecessary waiting. So, for inspection, this is going to save time for public or private security or safety. It can change the outcome of the incident. So, I'm going to show you side by side, dog two versus dog three, how long it takes to take off. **[14:39]** So you can see the doc 3 is already in the air about 10 seconds maybe about 20 seconds with the doc 2. So significantly quicker and get to altitude also quicker too. So, it took about 30 seconds to get to 100 meters uh with the doc 2. Looks like uh we didn't even finish the video, but it does take a little longer. Okay, DFR, drone first responder. You'll see this a lot. Public safety teams face several common problems. So, incomplete information from the initial call from the dispatcher, uh, limited manpower, long travel distances, and officers already covering many roles. So, a drone arriving first and giving the team a live view before ground units commit is pretty useful. I'm going to give you an example. A caller may report a person with a weapon. A zoom camera can help confirm whether that object is actually a weapon **[15:39]** before officers arrive. cha changing both the level of response and how the scene is approached. The drone can also assess urgency or clear call when nobody is present. So the dock turns that capability into an ondemand service rather than an occasional drone flight. So four capabilities that are going to stand out for the DFR program using the DO 3 fly to location. So this gives the closest available aircraft moving quickly. So the virtual cockpit is another. You're able to give the operator remote control of the aircraft and payload. And then third is autodispatch missions. So this connects alerts from systems such as CAD or sensors to predefined missions. And then flight safety combines geo fencing, aircraft alerts, weather information, and dog cameras. So each capability solves a different part of the problem. So response time, uh, decision quality, workflow integration, and operational risk. Together, these are **[16:41]** going to make the system faster, more useful, and manageable from a centralized command center. So, I'm going to break down each one of those that I kind of talked about. Fly 2 is pretty simple. Uh, you search for an address or rightclick any point on the map, and the closest suitable dock is going to launch. The aircraft can be airborne as in little as 10 seconds and can reach a point um reach a point like 10,000 meters away in roughly 100 seconds. So the operator can issue fly two while the drone is already in the air or if it's still in its dock. The value is immediate. So it's going to be able to get eyes on scene before most ground resources can arrive. It's going to give dispatch and field teams time to choose the right response rather than making that decision with only a phone call and a map pit. Once on scene, you will have virtual cockpit. So, I'm going to play these two videos right here, but the virtual cockpit supports a dynamic response. So, the operator **[17:44]** can switch between wide zoom and thermal views using the laser rangefinder. They can add the spotlight or speaker and take manual control when conditions change. You have AR overlays that can show street names, businesses, and house numbers directly on the live feed. A web link can also share the live stream with command command staff or responding teams. This web link can doesn't need to require an account. Can just have a password or no password. So, it's easy to share that live stream. Thermal can review information that is difficult to see in normal video. In one fire response, infrared imaging can help identify that the roof was becoming unsafe, allowing firefighters to leave before it collapsed. So, this is not automation replacing the operator, it's automation getting the aircraft there so that the operator can focus on the decision. We have this video on our DJI Enterprise YouTube channel. This kind of shows a **[18:44]** little bit about how that the Asheville deployment shows what it looks like in practice. So, after Hurricane Helena, a public private partnership helped bring DFR capability to the region. They had 13 docks supported rapid coverage with 10-second launches and arrival on scene in around 30 seconds for some missions. When roads and flooded areas were difficult to reach, drones provided overwatch, helped direct officers, and support the recovery effort. The same network can also run repeatable mapping routes to document cleanup and infrastructure conditions. So agencies could revisit the same locations and share updated imagery without sending additional crews into unsafe areas. So this program was overall able to reduce the number of personnel needed to fly missions across the city, improve situational awareness, and help build community transparency through consistency and visual evidence. The important lesson with the the Asheville case is not one individual statistic. It is that the doc network can scale from immediate response **[19:48]** into recovery, mapping, documentation, and long-term coordination. Across both use cases, the same platform capabilities keep appearing. So you're going to see that rugged hardware, remote operations, reliable communications, repeatable data, automation, and integration with other systems. So solar teams use those capabilities and inspect more often and compare results more o over time. You'll see public safety teams use them to arrive faster and understand a scene before sending people into it. In both cases, people remain in control but spent a lot less time on repetitive setup and more time acting on useful information. The use cases are different, but the operating model are all the same in both use cases. To bring it all together, the DOC 3 gives organizations more coverage, better consistency, and less downtime. It's going to automate the repetitive work, keep operators informed and in control, and helps teams move a lot faster and make safer **[20:53]** decisions. Thank you for your time today. I do have a teaser video that I want to show about the next doc iteration and then after that we have about 5 minutes for questions. Okay, I'm going to stop sharing. Yep, there. Perfect. >> That was really exciting, Michael. We can't wait to see what DJI has in store in the next iteration. I'm sure it'll be wonderful. Um, especially we everybody is kind of waiting on it. Uh, we have a lot of questions from the audience. um especially around um because you showed the public safety use case. I think there's a question around 911 call integration for DFR operations. Is that something that's already in place? Um so **[21:55]** yeah, go ahead if you would like to take that. >> Yeah. So we do have different programs. Um, from what I'm seeing, a lot of DFR programs are is that they'll get a 911 call and then the dispatch also ha uh is able to share that with, you know, either um I guess you could say a third party company or like a a supporting entity that's able to uh handle it. So, the dispatch is able to share the address and the details really quickly. and a third party that would I showed some of the features such as the fly to search the address. They're able to quickly go in there and either click on the map or type in the address and the and the drone is uh able to take off. So that's kind of the there there's a few ways to do it. I've seen um that is one of them. So I would recommend kind of exploring and uh either you know speaking to maybe other other different agencies and seeing how how they're doing it. But the integration is is available, >> right? And there is another question around is it possible especially for large **[22:57]** scale operations from for one dawn to take off and you know uh dock on another uh and move uh on for its operations. So is that feasible? >> That's a really good point. So if you have two houses or two docks and one drone, they are able to uh dock hop. So uh you're able to take off from dock A and fly your drone and then land in dock 2. Uh but there can't be the thing the catch to that is there can't obviously um you think of it like a garage if that second dock has that second drone in it it's not going to be able so you need two docks one drone one one drone um to do that so it is possible uh known as doc hopping from one dock to another >> right and this is one question that I had is just given all the dog three group deployments which is the industry that you think has seen the maximum adoption what is the use case that you know today dog 3 is being used most for? Is it more public safety focused or security focused or like completely uh different? >> I would say **[23:58]** both. I just see a lot of use cases. I think really popular right now is is DFR or security. Um not even just not even just like police departments. Um just for security like you know imagine um a lot of like new new build houses or uh you know properties or construction where you know maybe people do come on site and steal copper wire whatever equipments that's super common is having a dock uh being deployed is is amazing because it's able to essentially I was showing some of those uh auto dispatch if we're able if you're able to get an alert or an alarm it's able to quickly send uh an alert to Flight Hub 2 and the the drone and doc the the the drone will take off by itself and fly to that location that maybe it reached it got an alert or an alarm. Someone will get an email notification, text, or someone could just manually uh take control on a computer and quickly see what's going on. If it's nothing, you know, we've seen wildlife um it's not not a big deal, but if there's theft, quickly uh dispatch or call 911 and and report the crime. So, uh it's not **[24:58]** going to the next day and saying, "Wow, all this is broken or stolen." minutes instantly when that happens they're able to uh essentially call the authorities and stop that. >> Right. And um especially there's a very interesting question that there's somebody who's deployed two dockings two DJI do 3es today and they want to go to 20 and they just asking for what is the biggest thing that could have that could break when you go from the the first few pilots to uh at scale. So what are the things that they should look out for? That's that's a good question. Um, from a job like that, honestly, it's uh having two and 20 uh are very similar. I would say uh probably what I would look out for is depending on what your Flight Hub 2 plan is. You may uh depending on how much you live stream uh you might want to upgrade and that that'll be something that you might look into uh because from 20 to two to 20 docs that might be significantly more. Um, I would **[26:00]** think about pilots. Um, you know, if you're having only one or two pilots, uh, now you're going from 20 docs, you a lot of a lot more situational awareness. So, that live streaming. Um, just making sure you're set up in Flight Hub 2 where you're separating your devices with associating the binding your right your right uh devices to the right projects. Um, it could could be a little overwhelming. So, organization just setting up up your success with organization. Who's who's which user is responsible for this or that. Uh, that's what I I would recommend from going from two to 20. But overall, the system doesn't really change too much. Your missions and photos, they all go into your media. But yeah, I would just say organization and uh making sure who has the right admin or organization uh rights in Flight Hub 2 and taking a look at what plan you're using um from FlyHub 2 and see what what makes most sense for you. >> Right. And I think before we end it, there's one question that's mostly on a lot of people's mind which is around data security and data residency. So how does DJI support with these critical uh data residency requirements **[27:05]** uh especially from the the critical data that uh the docs capture the drones capture? >> Yeah, that's a great option. That's a great that's a great uh question. So in Flight Hub 2, there are many many operations to in Flight Hub 2. There's many uh ways to operate the the I guess I would say in doc 3 uh with the drone there's many ways to operate the device. I mean it's not necessarily just used through FlyHub 2. Uh we also have FlytBase. We also have an on premises version. So there are many many options depending on what uh type of uh security or uh uh I guess you know requirements you have. We do have different ways to operate this device and this system that will meet your security uh requirements. So like I said like uh different play of two FlytBase on premises we do have our uh cloud or S3 buckets. So we can have all the information just go all your data and information from your photos videos everything could just go straight to your cloud bucket or your internal system. So there are many different options that you can explore. I know everyone has different security requirements and IT requirements. I think um that's **[28:11]** something that you definitely can explore, look look more into and see which one fits your your your uh company and your use case best. >> Right. Uh I think we've covered most of the questions, Michael. Uh and we are almost up on time. So before we close the session and we wrap it up, do you have any last minute thoughts [clears throat] around where is this headed? Like how many what's the data talking about? How many flights are you collecting today? We saw recent stat on DJI doc covering some [snorts] 8 8 million flights a month if I'm not wrong. Maybe correct me if I'm wrong, but there's so much happening. So, do you have any last closing thoughts? >> Um, no. I don't I I would say with the with the DOC 3, it's really exciting. Um, as you can see, we have a teaser. Uh, October 14th, the the next generation of Doc is coming out. Um, as as you could see from the video, um, it looks like, you know, maybe some type of uh multiple uh battery or battery swapping. So, I think that was like a big uh pain point with **[29:13]** with the the previous do the doc iterations was uh was uh the the charge time. I mean, it was quick though, 15% to, you know, to 90 or 95 was very quick. I think uh in in the future you could see in the teaser it looks like um they may have more more than one battery or you know so I think that's that's I'm really excited uh for that for for for a lot of people like they would go fly and do a mapping mission and uh they they would essentially have to come back and and and charge. So uh super super excited to see uh what happens. So um yeah that's all the closing closing remarks I have. >> Thank you Michael. This was wonderful. Thank you so much for spending time with us today and uh sharing all the insights around DJI Do3 and what's in store and what should we look forward to on um October 14th. So, yep. Everybody, if you have any further questions um please uh drop in in the chat and uh maybe we can get send it across to Michael to take it over. Uh but yes, we are already up on time and thank **[30:17]** you again for joining in. We'll see you in the next session. Uh so yep. Bye everyone. Uh see you soon. Bye Michael. Thank you.