# Advancing Methane Intelligence: AIRINS Ecosystem for Detection, Localization & Quantification - Event: NestGen ’26 - Published: October 2, 2026 - Length: 32:19 - Watch: https://nestgen.org/library/watch/piTUmaoIR64/ > 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 Methane is one of the hardest emissions for operators to manage well. Leaks are invisible, often intermittent, and spread across large, remote assets, so knowing methane is present is only the first step. Teams also need to pinpoint where it is coming from and how much is escaping before they can prioritize repairs and report emissions with confidence. In this NestGen showcase, AIRINS will share how its ecosystem connects methane detection, localization, and quantification, and what teams should think through when building drone-led methane monitoring programs that turn field readings into action. ## Transcript **[0:01]** Welcome back everyone. My name is Pranjil and I'm part of the solutions engineering team here at FlytBase. We are now back in the physical AI showcase blog. And then this next session is about something genuinely exciting. This session is about something you cannot see with the naked eye until now. Methane. We're calling it advanced methane intelligence. The errands ecosystem for detection, localization, and quantification. And that title is really the whole pitch. Most tools today can do one of those three things. Well, today's guest is going to show us what it looks like when a single autonomous system does all three. Before we get started, uh I would like I would love to know who's joining us today. And please drop a quick note in the chat letting us know where you're joining us from and what aspect of drone sensing interests you the most. uh if you have any questions please drop that in the Q&A tab and we'll tackle them uh during the session at the end of the session and also this session is being **[1:04]** recorded so if you miss anything don't worry we'll send a copy to all the registered attendees afterward and don't forget to share your key takeaways on LinkedIn and Twitter using #J26 we would love to hear your thoughts joining us today is Shalu products applications specialist at Arens where he leads the technical coordination of Ain's gas detection sensors and runs field demonstrations across the target industries. Arens built the sniffle 4D family of gas detection payloads, compact sensor modules no bigger than a hockey puck that can lead up to that can read up to 10 different gases gas parameters in its single flight. Their newest system Medscan goes a step further. It doesn't detect a methane. It localizes exactly where it's coming from and quantifies how much is escaping. All from an autonomous drone flight. For anyone who has had to send a technician **[2:07]** uh with a handheld sniffer or OGI camera to work a pipeline or facility parameter, you'll immediately see how that matters. Please welcome Shalu, products and application specialist at Arens. Sean, welcome to National 26. >> Thank you. >> If you could just introduce yourself. Yeah. >> Yeah. So, first I want to share my screen here. >> Yeah. >> Okay. So, I hope I hope can my beauty is presenting properly. >> Yes, it visible. >> Okay. So um hello everyone uh good afternoon. Thank you for joining uh Nesten and also Aaron's presentation. For today's topic uh it is advancing methane intelligence which is Aeron's ecosystem for methane detection localization andification. My name is Shan Leo products and applications specialist at AMS. **[3:10]** So today I will divide my topic into three parts. The first is regulatory insights and our solutions. The second is field experience and use cases. The final part will be OGMP 2.0 and quantification. First of all, just in case you don't know, so errands, our formal brand is solubility and we focus on the environmental monitoring solutions mounted on mobile carrying platforms such as drones, robot dogs and vehicles. So our main focus is the gas detection sensors. So here are the hardware and software we all we developed by ourselves. So the first one is the laser based remote sensing method detection system. We call this map scan. The second is the multi gap sensors for like H2SO2 and other gases. We call this sniffer foldy. And the third one is the speed. It's mounted on the drones and to collect water samples. And the final one is the AI. **[4:11]** Arensto AI is the software platform we used to analyze all the data we captured from our hardware. So let's move on to the first part. It is regulatory insights and our solutions. Since our main focus is the missing. So first we need to understand why we need to um inspect missing so so much. So um methane the first impact is about the greenhouse effect. So methane is like 28 times more powerful than CO2 over the past two years which may give rise to the global global warming and also for the left safety issues. So if there are methanol links on site and you don't know which is very dangerous for your onset uh operators and also for the assets loss and property loss. So it is very straightforward when you are emissing methane to the air it means you're losing your money. So after so many factors so um there are a few like regulatory frameworks um they require the customers to submit the report about their **[5:15]** meth. So there are a lot of but we just abstract three main like regulatories here. So uh EPA OB and EU methane regulations. So these two regulations they focus on more on the uh LDAR program. So LDAR means leak detection and repair. Uh it requires the customer to um do the pipeline and the gas facility inspections about the missing and when the missing is above like a threshold of 500 ppm they need to repair it immediately and submit the record and the maintenance record to the authority. And the third one is the OGMP 2.0. It is not like a mandatory um requirements but it's more like a voluntary framework. So the end customers could join as a member and when they reaches at the level four and level five they will be regarded at the gold level standards. This is very important for the uh reputation. So it is like the uh UP which is the authority of OJM 2.0 zero they required in **[6:17]** customers to submit the annual methane uh emissing report like how many tons of methane you are emissing in your facility and submit the report and data to the UP annually. So they will review it and give you a level. So these are the three uh these are the three frameworks for the method detection. So previously the conventional method monitor methods u there are two main methods one is the OGI the other is like the sniffer. The sniffer could be a handheld or a portable or it can be mounted on the drones. So they are very straightforward. They will tell you like uh for example the OGI which is optical gas imaging. They will tell you what the gas plume looks like. So it actually visual the gas plume but method itself is um uh is invisible and or you can use the sniffer to tell you the point concentration levels of the of the component. But uh the limitation is that uh they um they should be like very limited the uh detection range. So both methods they are valuable for close range inspection but uh **[7:22]** neither of them is ideal for a rapid or large area screening because their uh uh distance is quite limited. You should be close enough to the leak source or you may miss some tiny leaks. So um this is why uh we have our own sniffer methods but we still uh decide to move on to the laser based detections. So first we could have a little bit look at our own sniffer uh uh methods solutions. So we call it sniffer 4 and the two plus. So the totally white one and the white orange one. So there are two uh standalone systems. The white one is the sniffer fold multi gas. So it is like a base unit. We could install different uh gas sensor modules into this base unit. So it could be sensitive to multiple gases that our customers desire or it can be um methane only unit. So the orange white one is only sensitive to methane. So it is uh is a uh 200 ppb. It can be mounted on different platforms like the smaller uh drones especially **[8:25]** DJI's platforms like this is the dock solution like doc 3 uh M4D and M4D and also M40 and MOE for larger drones it is also feasible you can mount it like uh different uh sniffers onto the same DJI M400 or M3 uh 50 just uh the the the unit itself is totally the same. you simply need to switch the uh integration kit. So it can be mounted on different platforms or you can mount it on like robot dog like unit tree or like on vehicles they are all feasible. Okay. So so the reason we move from the sniffer to um the remote laser based detection is uh like three main factors. First is the operational safety. It is like sniffer is a contact based solution which means you need to fly close to the leak source. You need to fly close to the like for for example the pipelines or the components but if you are too close to the pipelines this may cause like some safety issues because when you are doing the flying the tan may uh change fastly. So it is not **[9:30]** safe enough and also like the second reason is that for some gas facilities they have um very strict uh requirements for the uh explosion proof. So sometimes your drone cannot fly into the facility. In this case the close based detection is not feasible but you could still consider using a remote sensing for laser based detection. You could use like the camera's it gimbal to still cover this whole area. Use it gimbals to do a 2D screening so that uh when your drone cannot fly into you could still hovering outside but you could still screen this whole facility. And the third factor is that its efficiency. So for some large area facilities or landfills or for very long pipelines if you're using sniffer uh it may take a lot of time because you have to fly every corner every point of this facility or pipeline. But to use a laser based you could use this gimbal to do a fast screening. So the efficiency will be improved a lot a lot. So these are the vectors we decided to move **[10:32]** on to um uh laser based remote sensing system we call it met scan which is short for meing scanning. So its principle is like mcan will uh emit a laser into the open air. When it reaches a reflection service, the laser will be reflected back and we will calculate how much laser emitted, how much laser reflected and calculate the difference or the gap into the methane concentrations which is path integrated concentrations and the wavelength the laser's wavelength is tuned at 60 uh 1650 nanometers which is uh especially for the meth measurement. So it's only sensitive to methane. So let's have a look at the hardware design. So uh on my scan it has two cameras. So one is the telephoto cap camera, the other is the wide angle camera. And here the uh the orange rectangular shows the laser emitter which is the real laser will be emitted. And the the huge uh camera here is actually a lens to receive the **[11:32]** reflected laser. So this is how it works. And right now we actually have a mascan family. So here on the left one is the mask V1 which is what we just see just now. It is a larger sized of um methane detection. We call it Mascan V1. So it can be mounted on larger drones like DJM400 and M350. And also we have another version of MScan. So uh they use the totally same detection principle but just a smaller size. So it can be mounted on smaller drones like uh DOC 3. So uh we have two versions uh another one F and another one D. So F here is like a forward facing the D means downward facing. So they can be mount uh they can be mounted on smaller drones and used for oil and gas segments. And also we have our own meteorological system. So uh so if uh if you uh do the methane detection and the wind data capturing at the same time they can be combined with each other and finally output the **[12:35]** quantification which is like the flow rate the emission rate like how many kilograms per hour for your facility. So um they are from like different uh level of detection. So use meascan only you can get the localization detection of the uh methane concentrations combined with the uh meteorological data like wind direction and wind speed you can finally get the quantification like the kilograms per hour for your methane emission. So uh it's like uh the larger one the smaller ones the meteorological system and also we have a portable like a handheld version of meascan. still a laser based but you can observe it in a more close uh proximity uh point of view. So we call this a meascan family. So the total workflow will be like you use the hardware whether it's meascan v1 n1 or handheld to uh do a remote screening to capture some methane data and all the data will be automatically uploaded to our software analytical platform and then you **[13:38]** could do the analysis and uh hotspot identifications on our software and you could generate a report from the analysis uh results and then if you find there are some very tiny leaks you are not very short of you could use a closed range validation uh device whether it's uh it's our like handheld version of meascan or your own portable sniffer or our sniffer they both for the close range validation this is the whole like uh workflow so after introducing our own solutions uh we like to share some field experience and use cases because we did a lot of on-site demonstrations and technical support for our global customers so we like we'd like to share some representative use cases here. So the first one is what we did um last year in December in Indonesia. So um the mission is like there is a man-made leak you see here is a pink gap cylinder. It is releasing methane to the air and our task is like to screen this whole area **[14:40]** and uh generate report to tell the customers what we found and where we localized the methane leaks. Okay. So basically you could see the scanning patterns is look like a sector. So we call this sector scan. It is like the drone is hovering and does not move. It uses gimbal to cover this whole area. So it create a sector shift of scanning. This is uh especially you uh suitable for the gas facilities which does not not allow the drone to fly over. You could hovering outside the facility but you could still use our sensor to screen this whole area. We call this sector scan. So from the results on our software you can see uh it gives you the geo reinforced data like longitude latitude and also the methane concentrations and this realtime imagery because we have two cameras. So it tells you the u wide angle camera and telephoto camera. So you could always know the whole surroundings looks like and the specific look of the lick sort. So you can see uh we just localized methan licks **[15:41]** from the pink cylinders. So um the detection distance uh is here is like um 50 uh meters above ground level. So it's very precise localization and after you get the data if you have your own 3D model of your facility pipelines landfields you can upload your 3D model to our software you can then overlay our data point on your uh 3D model. This could make the total data looks more uh intuitive and understandable. So which is some of the functionality we want to highlight here. And another um uh output you get from our software is the report. So there are three main parts for our report. The first is like the basic information like at which country at what time you did the screening and the second part is like the parameter settings like the speed the area and how many um hotspot suspected hotspot area was identified by our software and the third part is about the details of every hotspot like its **[16:46]** concentration longitude latitude real-time imagery and also the Google map links. So uh it's very uh it's uh generated automatically. So which could be very convenient for you to submit to the authority or use it just like a internal record for inspection. Okay. So the second case is like uh what we did just this year in June I believe it's like in Texas and upstream like gas processing in a storage facility. So you can see here the scanning patterns is more like a belt. So it's different from the sector scan we just saw. It's like the drone is flying does not hover anymore just flying alongside the whether the facility or the pipelines and the laser is uh traveling downward and do a this kind of scanning. We so it will create a belt shaped scanning. We call this belt scan. So uh within this screening uh we find some periodical venting from their vaults and also we find some very tiny leaks. the **[17:47]** operators that did not know before. So they just use our mess scan to identify the tunnel leaks uh using our laser based uh solutions and uh measto can also be used on landfills. So this is a use case we did uh last um August in France is a French landfill. So the uh the area for the landfield is like 40,000 square meter and we use mascan to do this is also a belt shaped of scanning to cover this whole area is finished within just half an hour less than half an hour and we identify some suspected hotspot areas. You could see um the green dots here means there is no meth detection uh detected and the red dots within the green dots means we identify some hotspot areas. So you could see there are very high concentration readings. So according to the real-time imagery we identify there are actually some disconnected pipelines. So we use the uh concentration reading and the real-time **[18:51]** imagery as evidence of the world fire damage to the landfill. Okay. So the customers they are very uh satisfied with the output and also you can not only show it in a um a 3D uh point cloud you can also use it as like other um visualization mode. This is the interpolated visualization mode. So it tells you like the dispersion trend of the methane gas. So you can see the the green area means there are no methane leaks but the red area means the more concentrated areas and is dispersion trained of the messing. Okay. So here are some like use cases uh we want to share for some um target industries like oil and gas segments pipelines gas facilities and also the landfills. So the final part will be the OJM 2.0 and quantification. So the um the output I just showed before they are all about the concentration readings of the methane and also the localization of the uh of the messings. But **[19:54]** quantification like the flow rate emission rate of the methane like how many kilograms per hour of methane is totally another topic and more complex one. So I like to introduce OGMP 2.0 level four level five a little bit. So, OJM like its authorities UP it requires the end customer to submit the uh methane report to them annually so they could review it. Level four means source level measurements. It is like you need to measure your every like valves like uh flanges like the compressor, your flare stack, your venting stack and uh uh generate a report for every component and submit to the UP. Level five means set level like top down measurements. It just is like you need to submit the total uh emission for your whole facility and after they get your level four readings for every component they will sum up all of your components and compare the total emission to your level five measurement results and they should be aligned with each other. So how errands so how we can support this workflow it is **[20:59]** like we have our mcan for the remote sensing uh methane detection and also we have our own meteorological system it will capture the wind direction and wind speed. When we combine mapping datas and the wind direction and wind speed we could finally output the quantification like how many kilograms per hour not only for the component but also for the whole facility. So we support both level four and level five at the same time. So we call this algorithm SRQ. This is our own developed algorithm. So SRQ means source resolved quantification. It's workflow is like this. So you need mascan to do a preliminary screening to get the mascans data, the methane data and also you need our meteorological system on the site. Uh it capture the wind data at the same time. And then you need to upload the meth data and wind data onto a software. So our software will process and find every likely elix source area and then it will analyze how many kilograms power for source A, how many for source B, how many for source C and then it will **[22:04]** output the final total set emissions. So in a simple word so SRQ help answer two questions is where the methane come from and how much uh every leak source is emitting. So previously if you're using sniffer for the quantification you could probably get um the total quantification for the whole site but actually it does not know where the leak come from and where is like every leak source it is emissing but with our mess scan you could uh get the um how many kil kilograms power for every league source and where they are located at and also the total set emissions. So it help uh resolve two questions at the same time. So um all of our algorithms and uh uh calculation they were achieved by the control release facility we used. So you can see here we use the component to release the controlled methane to the air from uh 0.3 kilograms per hour to above 10 kilograms per hour using this component. And all of **[23:05]** our results and outputs will be uh uh will be like uh compared to the real control release uh volume and make sure they are aligned with each other. So we did a lot of test here. Uh we just show some of the test results uh here. So when the uh emission rate is above 1 kilograms per hour, the relative error here we showcase uh is just within 25 uh percentage. So it's very precise for the uh localization. So here are the two cases uh two two cases we did uh for uh world bank funded uh landfill project. So we use our mascan and aristo AI and our meological system for the quantification for the landfill and their project is that they use uh our solutions to get the total emission rate for the landfill and when the uh flow rate exceeds 150 kilograms per hour they could use the methane to generate bale gas power. So **[24:08]** this is basically their um project here. I'd like to show a little video of our uh quantification workflows uh how it works. So we have uh the control release. It is like 1.7 kg per hour and then you could use the max scan to do a preliminary screening and also the wind data and output uh and uh input the uh import the meth data and wind data onto our software. You could do the visualization by yourself and uh then you could do like the 3D visualization. Okay, here the structure visualization or 3D voxels visualization on our software and combined with the wind data here is like we will give you like a wind rose patterns. So they could finally output the kilograms power. So but it t takes some time to do the calculation which is usually an hour to do the calculation. Then they go finally tells you how many kilograms **[25:09]** power for the emission rate and also tells you where the uh location of the leak source. Okay. So you could also do like 3D plume visualization because you know where the uh lick source is and you know the wind direction and wind speed. So it tells you like the 3D plume visualization or a 2D visualization and you could see the plume with the windfield visualization along with the uh kilograms power for every source. So if you have multiple dick source they will tell you every source where they are located and the 2D gas with the visualization mode. Okay. So this is how the solution look like from like the uh lick source localization uh concentration readings and real-time imagery combined with the wind direction and speed. So we can finally output the quantification for your methane emissions. Okay. So I think that is all for my presentation here. So thank you for joining and **[26:14]** listening. So if you have any questions please feel free to ask me. Thank you. >> Thank you Sean so much for your presentation and it's genuinely interesting to know that how laser like remote based detection can be utilized to you know from a safety perspective how you can cover larger areas and all the methods that we mentioned sector scan using gimbal apart from you know adding and visualizing your data onto the 3D model and finally reporting. So it genuinely forms a you know end to end solution to detect methane using laser remote based detection. I think we have a few uh questions uh in the chat. All right. So I can take one of them. So we run a facility with hundreds of point leak points. Can the drone autonomously survey all of them on a schedule and flag only the ones that need attention or does u someone need to review every reading? I think it's more of can there be alerts that **[27:16]** can be set. >> Yes, because the uh screening speed and efficiency of the mascan is quite quick. So uh your facility if if it's like not too large facilities, it can be within just one morning. >> So actually we did uh we did a use case just uh this month in September in Middle East. We did a quantification for our own customer. They want to do level four and level five quantification and we did several sector scan because they do not allow us to fly into the facility. We just do a several I think it's like 10 10 sector scans and combine them because our software it um it supports the merged functionality. So we did 10 sector scan and merge them as a whole function and output the report. So the total uh screening just take one morning from I think it's 3:30 uh it's at 8:30 a.m. to uh 12:30 to 12:30. So just within a few hours we did the whole screening and the calculation of the quantification it takes just one night because the because the volume of the **[28:19]** data is too large. So we calculate the level four and level five quantifications for our end customers. They already submit to the OGMP 2.0 zero and they are all authorized. So just for from our experience so several sector scans just within one morning. So I think even if there are a lot of uh leak sources but it can be finished just within one day. Okay, that is my answer. [snorts] >> That's uh great. Another question. So we operate in a region where methane reporting regulations are tightening. If we deploy this now, will the data format and accuracy meet the standards that are likely coming in the next two or 3 years? >> Uh yes. So actually it depends on how the requirements is leveled whether it's uh just require the localization of the lick source or it requires the quantification. But all all our solutions they are all uh commercially visible. So we can we can just provide whether you want mascan only or you you want it combined with our like quantification package which includes the meteological **[29:23]** system. So they are all commercially feasible. So so as long as you have the requirements you can just reach out to us so we can see what we can support. Yeah. >> Yeah. Final question. So how does SRQ source level work for individual source level measurement by drone when the sources are too close together to segregate some plants specifically offshore the sources are not spaced out? >> H so this is a very good question. So for the onshore uh uh facilities, gas facilities, we definitely uh recommend our meascan because mascan is mounted on the drones but uh it laser is very powerful. So it could still get very validated methane data from from the uh component level or so or set level and the meteorological system is actually deployed on the ground. You you can use it on the drone but we recommend you deploy it on the ground. So the wind data from the ground is representative and the methine it uh the laser actually reaches the uh commotion level. So the meth data is also **[30:25]** validated. So combine the methane data and the wind direction wind speed data. So we can finally output the onshore so gas facilities data. But for the offshore so because uh if you are using a laser based so it needs the reflection but uh on the water surface you could get no uh reflection. So for the offshore gas facilities so we recommend using our sniffer. So it is like a mass balance uh quantification method. So you actually need to create a visual wall and uh combining with the wind direction wind speed. So they can finally output the uh uh quantification. But for the offshore if you're using a sniffer actually it does not tells you where the leak source come from. It the lick source they are kind of based on our assumption. So it cannot localize the lick source very precisely. But this is the only method we could use for the offshores. So, so maybe you can uh consider combining this two uh these two methods based on your scenarios whether it's onshore or offshore. But even it's unshore we definitely recommend using uh our mask scan because it's very very **[31:29]** efficient. If you're using sniffer I think it may takes a few days to do the same um quantification report as what I mentioned for our like middle east cases but using mascan it should be very efficient. Okay, I think that's my answer. >> Yeah, thank you so much for your answer. Uh, so I think this hope answers your question, Chris. So, thank you so much, Sean, for joining us today and the details you've shared, you know, have been really amazing and people will be taking away a lot of things from this session and thank you for the audience to for joining and asking these really amazing questions. Once again, thank you all. Stay in enjoy the rest of Nation 2026. >> Yeah. Thank you for organizing such an amazing event. Thank you as well. >> Thank you. >> Thank you. Yeah.