# Rewriting Nationwide DFR Operations with Automated Drones, Robotics, and AI

- Event: NestGen ’26
- Published: October 2, 2026
- Length: 32:58
- Watch: https://nestgen.org/library/watch/2wbwqZ6dYhk/

> 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

Drone as First Responder (DFR) programs prove themselves at one site or one city, but running them nationwide raises harder questions about consistency, oversight, and how many operators it really takes.

In this NestGen session, FlytBase will share how automated drones, robotics, and AI are rewriting DFR operations at national scale, and what public safety and security teams should think through when moving from a single pilot to a network of autonomous response.

## Transcript

**[0:00]** Hello. Hi everyone. I am Dh. I'm director of at FlytBase. Um uh welcome to this session on autonomous drones for public safety. Uh first of all just checking uh backstage if I'm audible clear. Okay thanks. So uh my personal background um I have been working on drones for about 12 years now. I um most of my career spent with FlytBase working as robotics engineer and then moved into product management. I work with our customers across security, public safety and many other domains understand their problem problems and um I'm also responsible for our product strategy. With that I would like to get started. So first of all let's understand uh why autonomous drones. uh everybody has been using drones but specifically why autonomous drones and by autonomous I mean docked drones for public safety. So the biggest advantage is there are 24x7 available. Uh

**[1:04]** it's an aerial response that's like always on standby. If you have used manual drones you know that you have to schedule uh a pilot to be on the site and do something. So usually anything that has a quick response to it cannot be done with those. uh but uh with this like a new avenue of use cases opens up suddenly. The incident trigger to take off uh that time usually with the latest talking stations is about 20 seconds which is really great uh and that's why you can reach any incident location quicker than ground teams mostly uh depending on the traffic scenario and all [snorts] uh which saves lot of that crucial time u like the for any incident let's say for any 911 call the later you reach you're going to lose lot of advantage it helps right size all the ground response. If a drone unit can reach the instant site faster then uh that live view uh that intelligence helps the um

**[2:06]** SOC guys or the ROC people to right size how many ground units do they need to do they need to send is it a false alarm or um like do they need not just one truck but 10 fire trucks because it's such a huge fire so that intelligence is very helpful every second or every minute uh that information is achieved or uh is obtained sooner is really uh helpful down the line. The other advantage is of course bird's eye view advantage and this is uh common with any drone operation. So with these advantages uh what are the different operational scenarios of for drones in cities. So uh public spaces uh public parks, tourist attraction, lakes, beaches, river funds, fairgrounds, there are a lot of crowded event uh crowded venues where uh you could have pre-planned or uh like a regular basis lot of crowd music festivals, sports stadiums, crowded shopping streets, malls, uh etc. There are places of worship, religious sites where you might have certain security scenarios.

**[3:11]** Then there are uh security sensit like sensitive sites such as security services, HQs, uh infrastructure, training facilities, communication facilities. There could be power plants. Uh there could be water treatment plants here as well. Um this uh that could be also categorized in critical infrastructure. Then uh roads, highways, critical junctions, bridges. Uh apart from that there's transit network, train stations, rail networks, airports. So these are just some example operational scenarios where autonomous drones are used and across these scenarios what are the primary use cases. So we see primary use cases as drone as first responder u DFR of course this is cat trigger launch uh arriving before ground units. Then there is incident response oversight and response coordination from command center. Search and rescue is locating missing persons in forest, desert, lace etc. Public events as we uh saw here um music events uh sports events, marathons etc.

**[4:17]** Then counter drone is another very important use case um like any rogue drones uh drones that are not authorized to fly uh how do you uh respond to them? Then patrols uh for spotting anomalies and issues uh where you know these are the things that are not usually are not usually covered by CCTVs because uh you can't you cannot have CCTVs everywhere as well as um most of the policing infrastructure also video analytics etc are geared towards u very specific uh people detection face recognition etc. not really uh geared towards identifying if there are any medical emergencies or any accidents. U so unless those events are reported there is no way for uh the law enforcement agencies to go and help uh or uh like address those situations. Apart from this there's critical infrastructure uh such as uh the let's say

**[5:19]** headquarter of police forces or training uh or communication equipment etc. which also needs its own protection. Uh and in those cases, those parameter protection use cases are also there. [clears throat] So let's maybe uh like double click on each of these use cases and see what like we can do in those u as autonomous drones. So drone as a first responder is usually u so this is a video by Cleveland police. uh they've been using uh docking stations to respond to various emergencies and these emergencies could be traffic accidents uh fire and other emergencies, thefts, robberies and depending on what situation you're responding to, there are different things uh that help uh for example uh in case of traffic accidents, it could help with post accident documentation, capturing a digital twin or a 3D model of that scene um uh full video like a PO UI video around that scene. Um that same system could also assist uh for

**[6:22]** traffic flow management. Uh if you need to redirect the traffic or anything um again response right sizing because some accidents if they have chemical spins or uh uh if they need to need hazmat then all those uh u issues can be identified sooner than later and then the response can be right right sized. [snorts] Uh in terms of fire and other emergencies uh again response right sizing is important. Then thermal analysis can be very helpful for planning response strategy like where exactly are the firefighters going to enter from etc. Good thing most of these autonomous drones now come packed with all these sensors. So all of these use cases are just rightly available. There's no need of like a special payload or a special attachment or a special version of drone just to cater to one use case. Uh in case of thefts and robberies, you can have you can track suspects uh per pursuit uh like have a live aerial pursuit. Although it it is still limited because of operational range of drones, but at least uh in crowded areas until certain distances you can uh

**[7:26]** certainly track and uh chase the suspects vehicle or person. So these are some of the u like detailed uh value ads that drones can do in this use case. In terms of incident response, it is all about uh uh it could be disaster management or terrorism incidents. So this is all about how do you uh use uh drones in your uh real-time crime center or national command center where uh everything is operated from there and you're able to coordinate response not only using drones but ground units uh multiple agencies and all the resources are working towards a single objective uh responding to that situation uh etc. The other use case is search and rescue and uh this is where drones shine a lot and u while we have seen lot of interesting use cases of drones um in uh one of such

**[8:26]** unrescy use cases but uh for common sites as uh major national parks or uh uh deserts um like really uh crowded tourist sites it is very uh important to have uh it's very useful to have a few drones drones stationed there uh which might come in handy in terms of search and rescue all the time and this is where drones can help u and there could be AI assisted thermal or visible live video search um search light and with a loudspeaker uh public safety agencies can even communicate with the person and coordinate the response. So this is where uh again uh talk drones shine a lot. Public event oversight is uh very interesting. Uh most of the um police agencies have uh already tried this with manually operated drones. um uh with a uh like a continuous overlap uh multiple drop drones uh you could use them to have a full

**[9:30]** cover of that entire area and with AI assisting in terms of counting of crowd u to avoid overcrowding to show if there are any u incidents that are happening any suspicious behavior uh pedestant traffic flow if you could look at that and understand okay what uh if there's going to be a cooking hazard here uh if the flow of the people who are entering versus who are not going out who are going out is not same then it might eventually lead to over uh overcrowding so all of those things you could do I'll also show how uh we are helping to some of the things but this is again another important use case and a very uh uh import uh every uh good one um which is like a major ROI for these systems. Uh counter drone response is another major use case. Uh automatically respond to rogue drone spotting events directed by counter drone. So we already uh a lot of

**[10:31]** public safety agencies already have these uh systems like um Done or any radar based system etc. And these systems are able to use RF, radar and several other sources sometime even visual sensors to identify any rogue aircrafts, drones etc. And many times these systems can also give you pilot location. In such cases, uh the doctor can go and uh look at those drones and give you live uh intelligence on who is flying. And if those if that person tries to um run away, then you can track them. Uh if they are there, you can talk to them [cough and clears throat] over loudspeaker. Uh so from a response point of view, this is a really good addition uh uh additional use case that these drones can serve. Then there's critical infrastructure protection. Um again uh live monitoring uh 24x7 line monitoring for prisons, security forces HQ, police training sites,

**[11:35]** communication equipment, sensitive tourist sites, sensitive religious sites, embassies and u lot of time u u these u uh perimeter protection task although you have lot of CCTV cameras they cannot cover entire area and during the night especially with uh thermal analysis if we are able to spot intruders or people or vehicles that are not supposed to be there and identify them uh investigate them immediately that adds a lot of value uh to overall parameter security use case. So with these use cases u u again uh petrols this is again sorry this is one of the cases I missed. So patrols for safety related anomalies. Uh so um anomalies is uh very interesting because that's not really covered by most of the um video analytics or CCD analytics today. Um everything is for valid defined uh either license plate or N&PR or ALPR or face

**[12:37]** recognition. But um anomalies could be very helpful in terms of identifying debris on road. Person trying to cross any access control highways uh which could be a major safety risk. Uh any person any vehicle that is like uh uh at the side of the highway maybe they need attention, they need help. Uh uh uh people could be fighting in public or person with weapon anomalous crowd behavior. Suddenly people start running and that might indicate there is some altercation or some kind of u maybe um some kind of issue that is uh the reflection of that and identifying all such an anomalous behavior is also very helpful in terms of uh um taking the next steps. So this is again uh and this requires continuous patrols. Uh but at the same time if drone system can identify such issues and alert public safety agencies then this also becomes a really good monitoring tool.

**[13:40]** So while there are these many use cases um and over the years we have seen many agencies [clears throat] start with one or two dogs and u uh they start with one or two dogs they try to go across more dogs and they start facing lot of challenges as they start scaling. So while there are those many use cases but at the same time these challenges are quite real and unless you solve these challenges it becomes quite difficult to uh um uh to get the ROI from this program and we have observed these challenges while working with public safety agencies across the world. Um some of these are agencies that operate nationally, some of these are just one city. Some of these are u um agencies um that are u private security operators as well. So we have uh experience working with all of them and that's where uh these are some of the challenges we understood. So I'll first share those challenges and then share more about what we are uh doing to solve those.

**[14:42]** So uh the first one is ROC expertise uh remote operation center. uh if you are doing private security or if you are an enterprise let's say an oil and gas company and trying to use docs you can externalize the ROC you can give a contract to some external vendor and they'll manage your ROC they'll have drone pilots they'll solve the drone pilot staffing problem maintenance problem etc but public a public safety agencies cannot do that uh they are they have to fly their own drones because they cannot share that video feed with anybody else so in that case uh the first big bottleneck that agencies face is how do they build this ROC muscle? How do they do hiring? How do they do uh all the SOPs? How do they set up all the processes and compl internal compliance uh audit trails etc. that becomes a major u challenge and uh many agencies like their programs are delayed just because uh there's nobody who would uh bring in

**[15:47]** all that expertise. The other challenge is responsible AI um u along with all the uh manual drone operation and of course there are pilots who are looking at the video feed etc. But several use cases are only possible when you use AI and at the same time AI use of AI in public discourse comes with several privacy concerns. So that's where uh there are some challenges. The other one is integration of drones in policing infrastructure. um using drones as independent tools um and uh not connecting them with the existing software systems uh makes them kind of obsolete eventually and uh uh they may not be able to offer value that they truly can. So that's one of the major challenges and then the fourth one is sovereign concerns. This is where uh um because due to the anti-globalization uh which is very unfortunate but it is what it is uh all the agencies are looking uh

**[16:48]** for uh infrastructure that they will not be denied in case of any uh issues with any particular uh country or foreign relationships. So this is where some concerns are and let me also share what we are doing in that direction before that like this is very important. The first challenges the ROC challenge that you saw the challenge also scales across these two axis. So one is when you go from one dock to hundreds of doc that is where everything becomes more complex the deployments become complex maintenance becomes complex because you have too many drones there's so many conflicting maintenance schedules there are piloting challenges you have u you have one several pilots that you need manage several flights that are that might be happening simultaneously etc. So overall that becomes a complex u operation to manage the other access is going from one police force to a national program. So several uh police forces that we work with are usually one agency that takes a lead

**[17:48]** and does the pilot uh for the full drone program. But then they are supposed to u like create a template for other agencies to follow the suit and that is where uh there are a lot of challenges. uh there is usually an expectation that whatever the first agency did they will be able to uh streamline everything abstract everything in terms of common infrastructure that everybody else can use and then there are some things that every agency will have of their own. So this horizontal versus vertical infrastructure that becomes one of the uh you know uh key uh issue there. So what are we doing to address these challenges? Um so one is u building safe efficient ROC practice nationally. So in this direction u so BLOS operations in areas have these many threat vectors. So fluttered airspace, rogue drones, ground risks etc. And for a pilot who is flying a mission, chasing a

**[18:53]** suspect vehicle at high speed in an urban area, they already have too much on their plate. And then there are also these safety threat vectors. So from a uh software point of view, we have been keeping in mind safety goals, human factors, operational goals. Our dashboards are optimized considering human mental attention, language limitations. So over the years we have done a lot of work on interface design and then um u how do we build in bacon fails um and how do we bake in automated response to various alerting uh alerts uh for example if you have uh airspace alerts how do we uh take the drone to a safe altitude automatically or how do we uh in case of any weather related events how do we alert the pilot properly so that they can go into whatever is needed to take care of that. So the whole interface design is done keeping in keeping that in mind. The other important thing is the training gap and

**[19:55]** this is where uh we really uh work with our customers in terms of helping them um of course create help them with training manuals scenario specific training where uh the training is not really about features but more about how do you really do an incident response or if you have on a public uh let's say uh a tourist site you have such a such scenario how do you navigate that so that scenario specific ific training um also helps operators uh really uh use the system and deliver the outcomes. Uh we have programs to train the trainers so that they can internally build an internal training program. We provide all this infrastructure for training consultation around setup practices SOPs compliance frameworks etc. So this is not something that is our primary software offering but we understand that technology is hardly a problem. U it is usually the people the processes the compliance and that is where we go uh above and

**[20:59]** beyond our basic software offering to help our customers. The other is key AI applications uh for public safety. So these are various AI um offerings we have for example detecting accidents we were able to identify these anomalies or detecting intruders where uh uh like there's a person who is trying who is climbing these wall here identifying that again these are various anomalies that that will be very useful in various use cases um in search and rescue identifying one thermal zone and like suggesting okay these might be people. It's very difficult to say that it's a person from like in that scenario but at least if we can aid the pilot and uh have them look at those spots then that might be helpful to the pilot. In terms of crowd monitoring uh uh like there is um during the live show you you could create such

**[22:00]** zones and in every zone you could monitor what is the crowd size and u you can measure the crowd that is passing from one zone to another zone etc. U again this is very helpful for u overall uh uh events and sports events, music events etc. Traffic analytics is very interesting in terms of identifying overall traffic flow, types of vehicles that are flowing, overall weight time, where is congestion getting uh uh where is exactly what is the reason behind congestion. So usually these traffic studies are done on CCTV cameras or using lot of expensive equipment or having people count lot of things etc. But now and it is still difficult because one CCTV camera has very limited view. But with drones now they have a huge vantage point and you can measure like 2 kilometer length of highways or roads and understand what are the different issues uh the traffic flow there and again this can be used for finding traffic violations like pedestrian crossing or like people

**[23:05]** taking long turn etc. [snorts] I'll be brief about the last uh two uh since we are running short of time but integration of drones in policing policing infrastructure is very important. Uh again uh there are several systems like citywide cameras, CAD uh VMS systems, alarm, RTCC etc. The drones have to be in >> now your fixed cameras keep watch on. >> For example, in this case uh this is all the CCTV cameras in the city on the drone interface. So that when the pilot is flying they will be able to uh go and look at any camera quickly and why along with the drone view they will also have the live view at that site. This helps with response. So particularly when this is operated from RTCC or u national command center at that time um you can have all your intelligence in one place rather than uh split it across various systems and then tying it in real time manually.

**[24:06]** The other example of this is uh how you could do your emergency response across different police forces. For example, there are this is one unified interface for all the all the teams to come together and uh respond to an incident together. On the map, you would see all the police vehicles or u even the drones and fire vehicles etc. everybody on one system and everybody can share their video feed, talk to each other, annotate, map and videos etc. This helps all the teams to come together and respond to an uh emergency or an incident uh together. Again, this brings that coordinated response and this coordination is exactly how drone programs can succeed and become uh that true infrastructure uh like part of the infrastructure that police use uh police forces use every day. If they remain isolated, if they're not connected with existing systems, then uh they just become a standalone tool that gets used seldom.

**[25:09]** So uh with that the last one is sovereign concerns and as understand understandably um many uh so these are beyond like the regular IT security compliance requirements. Sovereign concerns are about what if we go to war with so and so country and we are importing this hardware or this software from that country. What happens to our drone program and uh unfortunately that is like a real risk today and um uh in that case uh the denial of service or those are the primary concerns. In those cases what we have done is from a hardware point of view u any hardware that we use from a software point of view we uh sanitize that to ensure uh any third party cannot jeopardize your drone program. In terms of the software piece that we offer, we um uh basically provide locally owned servering cloud deployment. Um and these are locally owned deployed in a cloud sovereign cloud within your

**[26:10]** country and that uh usually deployed by local partner and the deployment is also owned by them from a licensing point of view. So that helps retain uh full rights u and uh continue your usage even uh during any such adverse scenarios. And then there is of course option of onremise air gap deployments where uh there's just no connection with the internet at all. [snorts] With this uh I wanted to just uh summarize we saw various use cases. Then we fa we saw like four major challenges that we have seen while public safety teams have been trying to scale their drones uh drone programs and how FlytBase is helping there. Um we have created a new offering uh because you saw there were several challenges that are very specific to uh public safety. We have created a new offering uh uh which is very uh focused on that. It's called Secura and u you can learn more about Secura uh at if you scan

**[27:10]** this QR code and uh with Secura our goal is to uh create dashboards, create systems, create integrations, EI agents uh that are that are specifically designed for public safety use case. Um happy to share more about it. Uh feel free to reach out to me. Uh my email is here as well as if you scan smaller QR code you can reach out to me uh there on LinkedIn. With that I'll be happy to take few questions. Sorry we are over time but I think we can take few questions. Okay. Which use cases should one start with if helping a city municipal department

**[28:10]** for a DFR program? Sorry, I'll read that again. Which use cases should one start with if helping a city municipality department for a DFR program? I think the primary use cases that you could start with are uh [snorts] especially around responding to events. DFR is particularly within that DFR is a use case. So basic CAD integration and then being able to respond, collect evidence, come back and like save that evidence that itself is a huge use case. The second lowhanging fruit uh would be uh I'll just quickly see if I can share the screen again but the um public event oversight is again very common and usually you would have these

**[29:11]** events all the time and if you can pick some stadiums or common uh uh some like famous public streets where there's usually a lot of traffic or people and or some kind of tourist sites. Um oversight of those uh is very helpful. Those are uh I think some of the use cases that you can easily start with and some of the other use cases such as counter drone response etc would need some more hardware integrations and you need like additional infrastructure to detect that but I would uh start with that and the other one is incident response. I think that becomes uh like a very uh basic fundamental infrastructure. So, integration with RTCC or at least some basic cameras and being able to involve other police personnel uh fire uh systems. It's very easy to do do that and u u involving all of them in one operation. So, those are I think some of the use cases I would recommend going ahead with.

**[30:17]** How to measure ROI for each use case. I saw a webinar earlier where Christian from Belgium police mentioned about a preemptive operations across cost line for use smuggle port we saw clear metric can you help me with either one of the use cases which you do about how to make these frameworks it's logical but considering what is the time can you help with metric design so we can certainly help with that uh I think Christian uh from police as well as uh I think some folks from uh NPCC u in UK uh they have been doing amazing work in terms of quantizing the impact their drone program is creating. We are uh of course not a practicing uh public safety agency ourselves but we talk with all of our customers and they have been sharing lot of work that they have done publicly. So that is a very useful uh framework and we can certainly engage uh on one-on-one basis to understand your use case what are the problems you're solving and uh being able to quantify those ROI metrics.

**[31:20]** Can you tell me whether any of your customers have already been using multiple docs for public safety? So there are several customers uh who are using u um se like multiple docs. uh so what I can publicly share which uh I think NPCC has already shared so in UK they're already I think roughly uh doing some I think uh I'm not sure but they have tens of docs u I don't know if I can share the exact number but you can really check it out you can just search BBC there are several BBC articles around that you would be able to find a lot of public information around that u there are uh the we also have a city state they are using about 30 docks and they're going to 100 uh docs uh next year. So um it's very common uh for public safety agencies to go beyond tens of dogs today. What are your hardware recommendations? I'm planning to use Skyio. Is that integrated with your software? So Skyio right now is not integrated. We are planning

**[32:24]** to do an integration with Skyio but at the same time uh we don't have a clear timeline. We are working with Sky but uh there's no clear time now when that would be integrated. Um at the same time if you have a particular uh requirement uh and a full uh project brief we would be very happy to work with you and see if we can accelerate that integration. All right. Thank you. Thank you for your questions. I hope this was helpful. Uh looking forward to speaking to you guys again. Thank you. Bye.