# From Detection to First Response of Wildfires with Automated Drones

- Event: NestGen ’25
- Published: March 7, 2025
- Length: 43:03
- Watch: https://nestgen.org/library/watch/BTGlwFlATyc/

> 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.

## Transcript

**[0:20]** good day everyone my name is rudr and it is my honor to welcome you to the session titled from detection to First Response of wildfires with automated drones today we will explore the Innovative use of dock and drone technology in forest fire detection and response a subject that is increasingly vital as climate change escalates Wildfire risks worldwide I invite you all to briefly share in the chat where you are joining from and what interests you about the integration of drone technology in wifi management your inputs will definitely help enrich this session uh before we dive into our content uh let me outline a few housekeeping rules uh to share your general comments or observations kindly use the chat section you may also use the Q&A feature for your questions towards the end we will have a dedicated segment to address your queries the session is being recorded and the recording will be made available within 24 hours if you encounter any connectivity

**[1:20]** issues uh during the session a quick browser refresh may help I also encourage you to visit the U UIC booth at the schedule time to explore explore more about the innovative solutions uh the booth uh timing is after a couple of hours from now additionally we would like to extend the conversation Beyond today's session uh if the content of the session resonates with you please consider posting about it on LinkedIn uh you can tag FlytBase and use hash njin to connect with others who are passionate about drone technology I would also like to take a moment to extend our sincere gratitude to our partner who have made this event possible uh first a special thanks to our media partner commercial UAV news for helping us reach out uh reach a broader audience within the Drone Community I would also like to thank our digital Spotlight uh sponsors avss and free fly freely for their Innovative contributions and continued support and finally a big thank you to our broadcast sponsors yonic drone land and

**[2:25]** drone nerds for enabling us to build bring this session to all of you thank you for your continued support it is uh now my pleasure to introduce our distinguished speaker Mr Michal sea CEO of U onic with extensive hands-on experience in implementing automated drone systems for critical infrastructure Michal brings a wealth of expertise from his work across forestry construction energy and Public Safety sectors today Michal will share insights from Real World deployment illustrating the tangible benefit benefits of deploying dog drones for forest fire detection and response without further Ado I now invite mishal and his colleague uh Miroslav chadii to to present and uh uh to take the stage and present uh uh and share the tangible benefits of deploying dog drones for forest fire detection and response uh thank you for joining us and I trust you will find this this presentation both informative and

**[3:27]** inspiring Michal uh the stage is yours good morning everybody uh I would like to uh start my presentation and uh I try to share my screen I hope it will work now okay okay so the topic is uh quite in uh important in this moment we just uh saw some uh wildfires uh in California few weeks ago so everybody now understand what the Val farest can do and uh I'm happy that I get the opportunity to cooperate with uh uh military Forest uh in Slovakia and uh prepare this kind of presentation about our experience how to how to manage uh the response to Wild

**[4:28]** Virus uh shortly about uon uon is a company which is leading leading provider of uh drone Technologies uh we are mostly focused on on Aral Technologies providing drone based solution for construction uh energy and Industrial application uh over the years we uh developed expertise in drone integration and Automation and data collection we are based in Bratislava and in Slovak Republic and serving mostly Slovak and Czech Market uh here you can see nice picture of the of the our uh uh our office in Bratislava uh and uh our mission is to improve the efficiency of safety and Precision in AAL operation uh we work uh with businesses in uh and government

**[5:31]** agency to comp incorporate Advanced drone system to into their workflow as an official partner of FlytBase uh we focus of providing secure and intelligent automatic solution for enterprise we had also several experience with integration sensors to drones data integration to develop specific GIS application uh one of our key Services is metan detection for uh Gas Distribution Company this something special uh in slovan ch Market uh about uh the the DJI dog brings uh another possibility so we focus into oura uh in drawing the Box system uh this fully automated drone system en enables now uh 247 operation with minimum human

**[6:32]** intervention uh and today I will show how to use the Technologies for Wildfire detection uh drone in in boox uh brings new possibility like remote operation automa Mission performance but also new challenges regarding safety and coordination in airspace uh shortly about uh uh military Forest uh it's we call it uh VM which is shortcut from the sloak uh sloak name SLO name of company uh VM is stown company responsible for managing extensive forest and agriculture land within the military training area uh with significant area about 63,000 hectares Willam operates through the for region branches uh understanding their operation is crucial as we explore how to drone

**[7:33]** technology can enhance their management practices uh VM has taken significant step uh by integrating an autonomous drone system uh to bolster monitoring and Security in zah military District uh on the map you can see the area covering 27,000 hectares uh uh has been a key military training space since 1950 uh the implementation of drone technology here is not only enhance security but also improve operation efficiency in monitoring vast uh landscape okay V fars uh okay

**[8:34]** sorry um now let's take about Wildfire situation in zah uh in a region of Western Slovakia this area is covered by large Pine Forest which is which are highly flammable combining with dry condition and strong wind uh the risk of wild fires is extremely high historical data reveals that a larger scale Fire have C significant damage in this region understanding this factor is essential for developing uh effective uh prevention and response uh strategies uh on the right side you can see some pictures or some some video from uh uh extinguishing the wildfire and uh on the next slide uh we will speak briefly about three major uh events uh or incident causing uh caused by wildfire

**[9:35]** in Slovakia uh in 19942 we have a larger vire in Slovakia in this time uh it was uh burned about 1,171 hectares uh in this region uh in 2007 uh there was a three hectare Forest destroyed and in 200 12 was another 200 of Pine Forest burned uh the the main disadvantage uh of this F in in this area is that uh the we have a lot of sand here uh and this uh Pine Forest which is which is very quickly uh burning so it's a really challenged to to fight with this uh this kind of fight uh so the the the problem is clear we Wildfire needs to detect early

**[10:40]** to prevent major major distraction if the if the wildfire start to start to rise and be bigger then usually uh we do not have enough uh um sources to fight with the fight with the fire so we need to be quickly in detection and react also very quickly uh what are the traditional methods for uh detecting the Wildfire uh we can use satellites we can use some uh Towers uh and of course we have to rely also on the citizen reports of course there are some limitation in the in the uh this technology satellite based technology is nice it's a large scale it's helping to track the fire but usually is not um so uh uh accessible for for

**[11:44]** the people who are managing the managing the fire and of course there are some obstruction uh due to visibilities and it's limited for early uh early warning uh Lookout Towers uh are good solution but of course they have limitation uh in in the way of um area which you can cover uh in the past they rely mostly on the human observation now uh there are several project where the human observation is replaced by uh modern uh technology like cameras air detection and so on but uh still it's uh limited with the area which which can cover this uh lookout tower uh citizens reports are very helpful they can be Qui quick but

**[12:46]** especially in this flat area where we are placed in in West Slovakia uh Sometimes When You observe observe Forest uh first you have to look on your yourself because uh If You observe the Wildfire usually you are too close and first you have to take care about your life and then uh trying to uh uh contact the the firefighters and give the information about the location uh what are the firefighters uh challenges it's uh it's really the to have uh very quick uh and accurate information from the people and we think that uh implementing drones in this uh uh decision uh decision making is very helpful uh because we have to understand that

**[13:46]** that the spread of the fire can be very quick and uh the response uh the rapid response is essential for managing the the fire so what is the solution uh we have to solution it's it's automated run we believe that it can helps so let's Deep dive a little bit to uh drones as a tool for uh managing the Wildfire uh drone give us uh view from the top of from the top from the sky so we have very quickly the response uh how big is the uh fire where what is the direction of the fire and so on so it give us more more possibility how to react if

**[14:52]** you if we uh uh connected with the automated drone missions uh then the then we have system which can regularly repeat the mission give us uh regular picture and detect the fire very soon uh okay what kind of drones we can use uh there are several drones uh which are which are uh relevant to this uh to Wildfire management we can use uh fixed wind drones we can use multirotor drone we can use some HB drones that that means Vos and so drones like this uh but but all all the standard approach with the drones means that uh we are starting the drones if something happened uh automatic drones have the main uh

**[15:52]** difference that the drones are flying regularly so uh uh we have big chance that we will uh identify the the fire in the start okay what help us with the Advan with the detection uh maybe 10 years ago we was happy that we can place some thermal camera on the Drone we have to use big drones uh sometimes over 10 kilos just to Implement some thermal CA camera and some good visual camera now if we can if you can uh use the the small drones like mavix 3dt or Now new matric 4 uh these are very small uh drones which uh can be operated from uh uh from any place they are very small

**[16:55]** and uh uh give us uh a lot of possibilities of course they are also significantly cheaper so uh the thermal camera is really the key point for detecting the heat anomalies uh then uh we can use some AI uh analyzes we can use satellite data which can help us and Al also we can use iot sensor but we believe that drones and AI detection is uh the most advant technology we which we can use now for accurate WiFi detection okay uh drawn in the Box uh it is my concept uh it's a significant advancement in uh automated vire response

**[17:58]** with the future like autonomous Lounge Landing the system ensure 247 read Readiness and for immediate deployment when the fire risk are detected uh in the the key benefits on in the of the draw in the Box concept is uh emphasing the its potential for rapid and effective wifire management uh I'm never tired for of watching uh the Drone the take off and land into automatic box it's nice to he see how the Drone is fighting with the wind to keep the right position to land precisely this is really not boring image and I like it uh until now we have ADD

**[19:00]** in in our drone business the cyber security was something we what we have to what we have to take to consideration but it was more um in relation to data processing protect protecting of our infrastructure however as we move towards fully autonomos drawn in the B operation cyber security become a critical factor in ensuring the safe and reliable drone missions drone not only collect sensitive dat data but also operating in uh complex uh environments where the security treat such as hacking and data Brides post significant risk flight based on premise provide a key solution for safeguarding autonomous drone operation

**[20:08]** okay uh a little bit missed in my notes uh in August last year we conducted initial test over several days to evaluate the actual radio signal and range and collect large amount of of video and photo footage to verify optimal flight parameters we had Pleasant 35 Plus degree uh Celsius in the shade also in the spine Forest there wasn't too much shade to begin uh with only the real test show uh you how the how is written only based on the real test uh you can you can uh check what is written in the techn technical specification we was happy that that uh VM asked us for this test and uh we perform it uh to show

**[21:10]** them that really the the DJI do and uh FlytBase is the right solution uh to implement it uh into uh a regular operation uh all the land managed by by the VM Falls within restricted airspace as you can see on the right uh side on the map which bring the specific requirement for the Drone operation it is not just uh about communication with ATC but also uh with other military units that may be May conduct the training exercise in the area uh uh test flies are not only way to uh verify whether the Assumption of the technical specification hold up in the real condition how far and we have to answer all's

**[22:16]** question how far I can fly will the Drone R lose the signal if the power files will the Drone um do what it's supposed to do what happens if the connction ction between the dock and the server is interrupt there are simple Never Enough testing for example we initially expect that uh charging would be the biggest immitation in uh performing repetitive uh uh Mission however we uh as we discover when using Starling and the flight based Cloud we couldn't upload data fast enough to um move the all the colle data uh to the cloud platform uh uh what are

**[23:16]** the why we do keep talking about the two and uh two to 2.5 kilom radius where the specification State 15 kilm range uh radio range the answer is simple 15 km is teral maximum under ideal condition based on FCC standard however in Europe we follow the which allows us use lower transmission power power reducing the maximum range to 8 F 8 kilometers additionally see additionally since the dock station is uh on the ground level and the Tre trees naturally obstruct the signal realistic operation range is about 2 kilm sure we could place uh the docking station in higher uh position on some uh Tower but uh this would limit the drone's operation capabilities the DJI dock has uh maximum wheel wind tolerance during takeoff uh and Landing for 8 meter for wind

**[24:24]** about 8 m per second but uh it can fly up in 12 uh um it can fly uh in the wind up to 12 m per second uh keeping the dun on the ground reduce the uh restriction caused by strong wind ensuring the more re relable operation uh and of course we believe that new DJI dtk uh will help us to uh uh increase the uh operation range with the function of relay station here is the uh key user requirement from the from VM and it was the custom map as you can see the map include a lot of colors lot of uh numbers and this uh is

**[25:26]** the identification of each uh uh Forest parcel and uh based on their management they understand that uh under uh under each number it's a history of the uh management of the of the forest and also what are the plans with the with the forest in this area so if uh VM would like to include the dock into standard operation they would like to uh manage Al also the Flies based on the uh their organization of the of the parcels okay uh it security challenges honestly say this is the nightmare uh of the on premise installation it security it feels like beginning uh

**[26:30]** being catch between two Milestones on the side customer customers closed and the high secur system on the other hand FlytBase with the their uh expectation uh strict it secur it security requirements a well protected infrastructure in VM and the same time uh habits of developers and their own vision of implementation this is the main challenge of the implementation uh the network architecture was designed to keep the individual components separate while allowing allowing uh for future system um expansion it was crucial to ensure that the system could be universally controlled from any any of the customers location uh the basic diagram uh may look simple but within the concept of the entire n

**[27:35]** network infrastructure it posed a significant challenge during the installation uh Beyond visual line of site operation in VM uh restrictor aace uh come with the specific operation limits the maximum flight altitude is 120 m above the ground level which ensuring the compliance with the SP regulation while still providing good coverage the operational radius is uh from 2 to 3 kilom from the DJI do 2 allowing the Drone cover the significant area without exceeding the communication limits using Advanced sensor like thermal camera and good uh Zoom cameras uh increase the total total monitor area extent up to 5 kilomet these include Optical and thermal imaging capabilities for precise Wildfire detection and

**[28:38]** uh situation awareness uh these parameters Define the safe and effective operation of the system within restricted airspace ensuring reliable monitoring while staying within the regulatory requirements uh operating BB laws in restricted airspace PR present several challenges particularly in airspace coordination and communication with different authorities uh first we have to be uh must have clearance and maintain Real Time Communication with Mal malaski ATC since we are flying in controller aace every flight must be coordinated to avoid conflicts with men aircraft this means Contin update and approval before each Miss altitude liit limitation also play a roles because we are flying at low altitudes Teran and uh dense forest canopy can

**[29:42]** interfere with operation make it precise planning essential another major challenge is signal and connectivity at extent dist distances uh maintaining a stable communication link within the Drone and CR within the Drone is critical any losses of communication could impact the mission so we need robust and relable infrastructure beyond that uh there are additional atic to consider uh there is a lzr 314 is an active airspace uh restricted uh that extend uh from ground level and overlaps within drone flight altitude this means we need to coordinate every emission carefully to avoid any interference there is a firing range called Zana operates in the same altitude Zone

**[30:47]** meaning the drones must be synchronized with military training activities to prevent the conflicts this may communication uh with not just 8 but also with the military units absolutely essential with proper coordination operation is restrict operation in restricted will be not possible that's why uh we are working closely with all relevant Authority and it's a key part of BOS implementation okay conclusion and next steps AI detection it's a it's a uh in this time uh to have presentation without a AI it's absolutely nonsense so I also part about AI detection implementing machine learning from real fire and SM detection reduce uh

**[31:50]** relies on human monitoring making detection faster and more efficient however where AI process uh Implement when where air processing happen is crucial live stre video is uh compressed to maintain range and minimize the uh drone control or delays if AI is applied only after the data exist DJI dog it can process video in HD or even and in good condition full HD quality leading more to more accurate uh uh fire detection the M3 M3 TD drone records the 4K videos in visual spectrum and uh 400 640 to to 52 uh pixels thermal

**[32:51]** footage but uh there are some limitation like uh we can uh transmit only one stream uh and this is challenging to implement good AI tools of course we can use the the AI tools for full uh 4K videos but uh this is not realtime analysis because the 4K videos are only available after the download to uh Doc and then upload to the to the server so we we have to rely uh on the H HD quality for the for the AI analysis and also uh with the limitation that if the operator switch from visual camera to to thermal camera then also the the stream which which have to be processed is is different uh but we take the AI as a next steps for for our

**[33:55]** project uh what are the the key takeaways from the from the deployment uh we approved that DJI do uh with M3 TD has proven valuable for continuous area of surveillance uh we uh we are still in finding uh successful uh uh uh air spare regulation uh we have some uh frame for testing and this pilot operation but but still we have to lot of do a lot of to do with the national uh authorities and the next steps is of course uh the Automation and AI integration uh where we see big potential of this uh of this project

**[35:00]** okay that's all from my side I and I look forward for the TS and questions and uh I' I now try to switch back uh to the live view okay thanks um thank you Michal for this very insightful presentation I open the floor to questions now uh I am looking at the Q&A section and uh we do have a few questions so let me start with the question received first which is uh so I'll just translate that question for you and uh uh the question is that given the area over which uh VM operates or in general the forested areas are quite huge uh for the purpose of forest fire detection uh does the rescue service have the uh budget for using a lot of uh dock based drones or is fixing a more reasonable option uh so it's

**[36:06]** basically a comparison between fixed wing and um you know DJI Dock and drones for uh uh forest fire monitoring overall extended period so uh do you have any thoughts on that um based on our cooperation with VM uh we have to understand little bit different position uh uh how we why why VM considered to use DJI doc uh VM is not uh it's VM is a company who is managing the forest so uh they decide that uh of course if they have to uh manage this this huge area with a lot of challenges the DJI doc can be helpful and the the financing the the implementation of the DJI dock is based on the budget of VM of course they have relation with

**[37:06]** the firefighters located in malat and in uh in in the whole region but the the the the monitoring is done within the VM organization and if something happen based on their uh their operational procedure they uh they start communication with the firef fire TS and uh uh initial the in and they are then initiating the fire five ter response to the Wildfire okay thank you for that uh I have another question from uh Christina uh which is regarding uh the temperature operating temperature ranges for the dock so have you faced any challenges with operating during high temperature or during snowy conditions uh what has been your experience uh based on our our experience uh in our uh climate condition the the the temperature range which uh allowed us

**[38:12]** to operate the Drone are um sufficient of course if you have a very hot weather in in the summer uh you have to calculate with a little bit longer time for charging the batteries because the throne when lands first should be cooled and uh of course there is a uh there is air condition inside but but you cannot simply charge batteries which are which are very hot and in the winter condition uh the low temperature of course make two challenges first is the freezing on the on the on the propers this something what we have to uh still fight some somehow with this and second is the not the low temperature but but mostly the strong wind uh especially for the landing and take off uh when we have wind which is more than

**[39:13]** 12 met per second simply we cannot start the Drone understood uh thank you for explaining that I have couple more questions uh so one of them is uh so have the forest rangers adjusted to using drones instead of fire towers and what do they think about this change uh okay okay uh we are using the drones uh here that that we they are in dedicated places and operated uh from from one place uh we are slowly slowly preparing also some uh Operation Center in VM uh which will take uh take all the operation in one place uh so in the future we expect that there will be a several dock station and there will be a central control room for all of these

**[40:16]** four areas which are operated by VM okay and dedicated only only for West Slovakia but also for the rest of the Slovak where where we is operating okay and the last question that we have is uh what was your uh biggest challenge during this project with VM was it Logistics was it uh you know on the deployment of on premise what was what has been the biggest challenge so far uh as as I mentioned the the the communication with flight based development team and the the it guys from LM and quite strict strict conditions uh regarding security and uh uh and uh uh information which can be uh uh on one side from VM of course they cannot uh provide us with uh very

**[41:19]** detailed information about their infrastructure and on on side I understand that that the VM the the flight based developer has also some uh some knowledge which cannot be uh just simply uh uh uh distribute so so and finding this uh right communication and uh right place uh for the implementation it's really The crucial Point okay so to summarize U you know integration of uh dock drones uh in a high security environment has been the biggest challenge yeah okay all right um thank you very much we have we have to understand that VM is uh is uh state of company but but their owner is Ministry of Defense so they have to follow all the all the military standards regarding the it security so this is the This Is The crucial Point all right okay uh thank you very much Michelle for your uh for the presentation for your

**[42:23]** answers to the questions and as we conclude today's session I would like to thank everyone for their participation I hope you found the found this session inspiring and it would encourage you to explore further applications of automated drone systems uh not just for uh Forest uh management but also for other applications and uh thank you all for thank you all for joining us today thank you