# Smart Analytics Data Management for Automating Infrastructure Inspections

- Event: NestGen ’25
- Published: March 7, 2025
- Length: 36:36
- Watch: https://nestgen.org/library/watch/P29qxhiFU2M/

> 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]** hello hello everyone welcome to another session at n25 I am aushi your host for today and today we are diving into how AI powered analytics and data managements are transforming the infrastructure inspection think about it critical infrastructure like Bridges power lines pipelines required constant monitoring but with traditional inspection methods are slow expensive and prone to human error enter autonomous drones and AI powered analytics which are redefining how we collect process and act on inspection data but wait before we begin I have a few housekeeping tips for all of you to help you make the most of the session and make it a seamless experience for all of us first if you have any questions drop them in the Q&A box anytime we'll address them towards the end of the section second engage with us share your thoughts in the chat box and let us know what industry you work in and where are you

**[1:21]** tuning in from and last but not the least join the conversation on social media use the hash n and and post it on link in Twitter Instagram and let us know how you feel about the session and with that also a massive thank you to all our sponsors for n 25 who have helped us to make this even possible bigger your support has been a key element in helping make this na bigger and better every year with that I would like to introduce the speaker for the day today we have with us Danny Feller the VP VP of s at scopito a company at the Forefront of intelligent inspection of data management Danny will work us through how advanced analytics AI powered Automation and cloud-based Solutions are helping Enterprises automate inspections improve decision making and reduce operational risk welcome Danny on board let's get started

**[2:23]** thank you very much aushi and I would like to start by thank you very much to flight base and aushi for the last three four weeks of communications here it's great job you're doing and Awesome event thank you very much perfect Danny so we have a lot of things in store for us so over to you Danny cool so just uh to go through the agenda so we are all online um we we are going to go through a little bit uh on the how scopo started um where are we today where we see the future scopo is uh not great at doing presentations we prefer live stuff so I'm going to share my screen I'm going to give some uh live demo and our take on um uh how the workflow of inspection is done a little bit about Ai and then we have a couple of uh pre-recorded videos from a couple of our um Partners uh partners in crime I may say we are going to uh watch

**[3:29]** that and eventually leave some time for a Q&A is that cool perfect that sounds great awesome so just a few words about uh how we started scopo started little over 10 years ago as a a project spin-off by Ken FAL today CEO of scopo um Ken started the company with the idea of creating um basically the world best drawn for inspecting power lines the uh company quickly uh evolved into doing a project for uh energy net the Powerline transmission company in Denmark and and very early in the process we realized that there is a problem in handling the data viewing the data back in the day we gave them SD card with 5,000 images not a lot uh in today's terms but back then it was quite a lot they put that SD card in their in their computer and then basically want to drink coffee because it took

**[4:32]** minutes and minutes just loading those 5,000 images not talking about each time you wanted to go to the next image you had to wait few seconds time 5,000 that that took a whole day just reviewing images so we really quickly realized that the problem is not necessarily with the hardware of flying the drones and we believe that that's where great companies uh with the you know w footprint will be able to just take over and we wanted to focus more into the into the uh uh software into the data management the um the turn of what it happened was that compo the platform itself our c-based platform was an internal tool for us to deliver and share data with our clients we were fortunate that uh a year or two later after our start um the largest lens company in Denmark came to us and basically acquired all of the hardware part of scopito which

**[5:37]** made the made it uh very clear to us that we had the capital to run with the run with our ideas of uh the cloud-based platform so fast forward to today we are uh close to 10 people um we operate all over the world I'm going to share my screen soon and uh and and show you rough where where we have people uh people using scopo and uh as as I promised we are going to talk a little bit about the future so our idea inside scopo internally is that we wanted to make sure that at the end of time when everything is of course in in you know H when all the stars are aligned we want to make sure that you click a button and an inspection is done including a report including analysis including everything we're not there yet but we are moving there in a quite the big steps and I think I feel

**[6:40]** like the the market is definitely uh filling it so without further Ado I will uh start my um my screen here I will move you to the other screen so I can present on my screen I'll share my screen right here share entire screen and can you give me a little no when you can see my screen yeah your screen is not visible can you try again yeah sure I'll do that again um share screen Tire screen share make sure you're not connected through a VPN I'm not connected through a VPN

**[7:41]** so that might be a problem okay something happened here can you see my screen now yes it's perfect it's up live otherwise I think there is some issue um just uh give us a moment Danny coming online yes you see my screen now yes Danny uh we can see your screen now that's perfect all right so what I'm showing right now is a live map of all of the places around the world where scopo is being used you can see uh literally in every every country every continent there is some sort of usage of sco and just perspective the use is mostly for power lines but used wi widely for solar Farms wind turbines um cell towers those are the main areas then you have oil and gas and buildings and

**[8:44]** everything you can really imagine um people are using drones to capture data to to acquire it and need the analysis for it scopito is being used for cool I'm moving on to My Demo account in this account I'm going to show you um my um transmission distribution inspection a little bit of a demo this is how users how scopo users in inspect infrastructure and move from traditional methods into uh the future of visual inspections where scopo puts the uh the the the the weight into making everything more efficient so of course we all uh are looking at AI we all look at automation but we also need to be realistic and we need to do things today we need to inspect our assets today to prevent fires like we recently saw in in Los Angeles we need to make sure that we are doing things

**[9:44]** correctly today with the ability to use what we're doing today in the future cool so just to to go through it quickly I have an inspection here of around 5,000 images what you can see is a demo it's running right now on the cloud and hopefully it runs smoothly at your end when I click on an inspection it opens instantly and this is referring to what I just said before of how computer started we just loaded 5,000 images there was maybe a millisecond of lag there but that's what it that was it we didn't need to spend time just waiting for data to load on my map here this is an inspection you can see in Denmark and you can see the images are being automatically so after uploading the images they are being automatically positioned on the map based on the metad data from the images those images have already been analyzed and therefore you can see the different um color codes on the on the data but let's H

**[10:48]** look aside of that for a second the idea is that when you load images you can quickly go to what's relevant in this case I'm going to focus into this Tower 29 and I'm going to zoom into a certain area here when I do that the images that are relevant right here on this portion of the screen of this inspection are available for me so I can quickly uh select the image I would want to see I can go through my images just by clicking on my keyboard I don't need to spend time waiting for the images and then I need I can go quickly to different elements on my inspection and start reviewing what's relevant this is currently done manually scopo has over 11,000 companies registered throughout the map I showed before all over the world and no one is using AI in the sense of uploading images and finding issues as they are that is that might be a bump that I'm I'm throwing right here at you but that is

**[11:52]** the reality another reality that is that every annotation you create in scopo and annotations can easily be just made with a couple of clicks on of a button or a simple click if we are using our suggested issues all of those annotations can then be used to create Ai and being uh or preparing your organization to the Future to when the AI is going to be working in a a way that is a valuable so it it finds what need to find and B where it's um cost effective it doesn't cost more to run AI than it does using manual work but going back to the annotations so the idea is that you find the you find the the issues that you want to highlight in this case we are going to focus on this Scutter pin and let's just imagine that it's um it's missing the idea is that you can quickly select button start creating the annotations every

**[12:57]** um every operation I doing scopo I can also uh use keyboard shortcuts to increase my efficiency so imagine going through tens of thousands or hundred of thousands or even millions of images you want to save every millisecond because they are up so instead of me going with the mouse and clicking on that button I can quickly press D mark and then start filling this form here on the right hand side so I can quickly um create my annotation this case uh it prompts us the question of what the component is we'll just start typing copin and it will already pre filter for us again to make it all fast so this is a Corin the issue would be that it's let's say missing and a severity again I'm not an expert so we'll just select severity three here you can optionally add a remedy action a repair cost or a comment if you had it and simply create it once it's created The annotation is created right here on the

**[13:58]** image and you have a visual indication on the bottom when all of that is created and done you can quickly go to the next image and if you had the same issue or similar issue you can select it from your recently used issues again to increase your time of um uh analysis or make it more efficient you can also create your suggested issues which will automatically be connected to a keyboard shortcut so you can uh pin those up by pressing the plus buttons and you can rearrange them so you can make sure that the the buttons fits to your to however you want it to to work with your with your keyboard so if we have for example a broken glass insulator which is connected to our key2 and we go to this insulator string here and press two and Mark this broken glass insulator again just a demo when all the images have been analyzed we can quickly use our keyboard shortcuts to view everything and we can see that we

**[15:02]** have those annotations we have the ability to use these interactive filters where we can select one two or as many filters as we want to filter the data so we can work with it and essentially report it and create a work order in this sense we can uh start by saying okay I have 5,619 images that's a lot I'm not interested in all of them I only want to see where I have issues I can simply select this button and now I'm down to 216 images that's 5% of all of the data that is relevant still 26 images it's a lot it's 5% but it's still a lot for us to to work on so we can filter it even further say okay I don't want to see all issues I want to see the most severe three four and five severity issues by the way those can be um customized so if you work with different severities I don't know call it category one two three or whatever you call it scopo can adjust it to fit to your organization now we have

**[16:04]** selected three four and five and we are down to 11 images and we can be even more accurate saying I I want to see three and four and five but I also want to see my lighting strikes when I do that I have four Images left and that is what I want to fix that's what I want to focus on so I can see my lighting Sprite I can see see it from different angles I have the ability to link issues together if I see it from different perspective and when I'm happy with the filters that I've that I've selected I can generate a report there are three ways of how to share data you can generate a PDF which I'll show you in a second you can share a you can generate a CSV so you can export it to your Erp system or whatever system you use to create a work order or eventually you can simply give access to your uh counterparts to your stakeholders to use scopo so it is a multi-tenant platform you can create organizations companies users with different roles and rights it's all pre-built and

**[17:08]** it's all included with whatever um package you buy or use in scopo so this button right up up right here would say generate report I'll press on the generate report it will take into account my um filters I've selected we have the option of course to change the logo we can uh just a map level if we wanted do we have an overview of our uh issues and numbers of issues for severity with a place where we can write a summary if we wanted that then we have a table view of all of our issues if our assets have been shared with scopo in like I will show you in a second where we know the name of the asset and its location scopo will automatically assign the images to the asset so you don't need to remember which image belong to which asset you can simply focus on what you do best which is piloting a drone or

**[18:08]** acquiring images using a helicopter or handheld device whatever right and then the main part of the report would be a a image obviously with this issue the zoomed in part of the issue we have the asset which is connected to so the nearest Tower here would be 31 scopo makes a lot um of calculation to actually assign the correct Tower not just the nearest one that's a topic for another session I guess then we have the link to the location of that image we have the link to the location of the issue so those could be very different sometimes the location of the issue and the location of the image then we have the issue and its severity and very important a link to the image so if we share it with someone they can click on it given that we gave them the access to it and it would send them to that image they will be able to look at it and see if anything else needed to be annotated which was missing okay that's a very straight forward um we are slightly short on time because we still have the

**[19:12]** two videos we want to show so I'm going to jump quickly to our asset overview this view here allows you to see all of your assets and all of the inspections that were created for each asset at this zoom level you'll start seeing that things are happening so when I zoom to this area right here you can see different assets being shown with different colors the gray ones would be all of the towers that I have on my grid the blue ones would be the ones that are inspected and we can see that some where the data was not uploaded for the um for for these assets allows us the ability to Al also review the progress of our inspections and then if data was uploaded and annotated we will see it in red if we zoom in further we'll see the name of each asset we have the ability to search assets or group of assets the whole feder line or a specific to or simply click on that Tower it will uh note all of the inspection that you had on that specific Tower throughout the the history of uh your

**[20:16]** inspections uh with scopo obviously and then you can simply click on that inspection it would move you from the asset overview to the inspection view you'll be able to look at the image that was actually causing this whole FID line to have this priority last thing as I promised the AI dashboard I'm going to lightning fast go through it because we are short on time I would uh welcome any question regard it afterwards so the idea here is that scopo is trying to simplify this is a better better feature right now but you should see it you should see it because this is basically the thought behind everything you do in SC so each time you create an annotation you can see that scopo automatically count it and put it in a library this libraries become our libraries to train to train AI we are not saying that you are going to be able to train a 100% Mega good AI that will work from day one we are

**[21:18]** saying we are able to test your uh compatibility and assess if your data is good enough to continue improving it to reach to um basically fullblown automation so in this uh instance I have this insulator that we've annotated I can press the train here we'll have three phases we can Define the model give it a name in this case it doesn't matter I'm just going to quickly go through to the images scopo allows you to select the data where the images will come from in this case only one inspection I'll press uh next and then I'll move to see the actual cut out of The annotation of the inspection I can review it and if I find something that I don't want to include like this brown insulators for example I can simply deselect them when I'm happy with that I'll press train and scopo will run in the back end and start creating this Ai and give it give us a button for it in our front end where we will be able

**[22:21]** to Simply press analyze use human or AI analysis to run this um mod models since we're short on time let's move to the videos so I'll stop sharing my screen right here and give it back to you aushi so you can run the videos well I must have to say that demo was quite impressive it was very detailed I think the audience might have get the entire view how they want to proceed with their next steps for you know inspecting any kind of infrastructure just give me a moment I quickly share my screen right I'm sure there were some questions and we should probably attend them afterwards but let's focus on the video it's from slide four there you see slide four that would just start the video automatically re

**[23:30]** uh can you see my screen we can see your screen the video just stopped in second three start it again revolutionizing Utility Services through drone Innovation hi everyone I'm raid CMO at Kronos group at Kronos we specializ uh give me a moment I think there is some Network issue just let me just give me a minute I'll quickly fixed it sure couple of words about Kronos while we're waiting Kronos a longterm Cutting Edge drone robotic solution for inspections security and infrastructure management but beyond just flying drones we also have an advanced R&D division where we develop robotic drone based solution for Unique challenges like installing anti-b bird electrocution covers on energized power PS or placing bird diverters on power lines to protect wildlife and prevent outages we work with

**[24:34]** utilities energy companies and infrastructure operators helping them gather high quality data reduce operational risks and improve efficiency one of the key tools that help us do all of this is sco it's a game Cher in how we manage analyze and share their own inspection data let me show you how it fits into our daily operations let's talk about what happens after we collect all this drone data one of our biggest challenges is not just capturing thousands of high resolution images but organizing and making sense out of them that's where scopo comes in there is another GL a lot of traffic on H on flight Based Services today transmission tower substations and and distribution Poes after flight we upload all

**[25:41]** the inspection data directly into stedo the platform automatically organizes and tags images detecting potential issues like corrosion or damage isolators this is huge for our teams because it means they don't have to manually shift to thousands of images I think there is some Network issue at it also lets us compare past inspection with the latest Danny can you confirm can you see my screen I can see your screen it's just uh the video is kind of stopping and not working just give me I I just give it another try which helps our end users to track assets health over time and

**[26:44]** PR I think there is some Network my end prioritize maintenance and prevent failures from happening give me a minute I'll just quickly that's a shame how easy it is to generate report just give me a moment I'll quickly get my network fixed apologies everyone for this with just a few clicks we can create professional

**[27:57]** detail e I am back with my network let me just quickly try it again revolutionizing Utility Services through drone Innovation hi everyone I'm R Electric power is the revolutionizing utility services through drone

**[29:00]** Innovation hi everyone I'm CMO at Kronos group at Kronos we specialize in Cutting Edge drone and robotic solution for inspections security and infrastructure management but beyond just flying drones we also have an advanced R&D division where we develop robotic drone based solution for Unique challenges like installing anti- bird electrocution covers on energized power palls or placing bir diverter on power lines to protect wildlife and prevent outages we work with utilities energy companies and infrastructure operators helping them gather high quality data reduce operational risks and improve efficiency one of the key tools that help us do all of this is sco it's a game changer in how we manage analyze and share their own inspection data let me show you how it fits into our daily operations let's talk about what happens after we collect all this grown data one of our biggest challenges is not just capturing thousands of

**[30:05]** high resolution images but organizing and making sense out of them that's where scopo comes in take out work with utilities and electric companies for example we conduct large scale inspections cataloging and surveying assets like transmission tower substation distribution I think there is after flight I am so sorry everyone upload all the inspection data will the video links will be available for afterwards uh watching yes uh apologies

**[31:05]** everyone I think there is some issue on our end um the link will be available after the session ends we will be uploading all the presentations along with the videos on YouTube channels so all of you can access it through there and apologies again no I think that's okay let's uh be mindful about everybody's time uh the video links will be there there is a link to my uh contact details for further communication if anyone have any question or anything wasn't clear in my presentation H unless there is any uh questions right now I don't know aushi I know that we're over time but U if there is any question I'm happy to hear I see there are a couple of questions in the Q&A tab maybe we can take them for now sure okay the first one says how do AI powered platforms like scopito handle large volumes of infrastructure data effic efficiently that's a great question and I think that the short answer is that we

**[32:07]** have a a team of Superstar developers that basically created everything from the ground up to make sure that we deal with data in the most Innovative way possible to make sure that it's uh it's done in the right correct way I'm not a technical person so I cannot answer more than that understood understood uh I think we can take the next question in that case uh if you have any other further queries or so we have a booth live that is set up for scopo so all of you can you know directly head over there after the session and maybe you know have another reward with Danny there uh okay Danny let's take another question that says what industry have seen the fast fastest adoption of AI powered drone inspection and why do you think is that so solar solar farm solar PV is one of the uh one of the industries that is mostly using AI where I'm saying AI I'm assuming that people are talking about the the

**[33:12]** option to upload images press on a button and hope that they will be analyzed correctly and the reason for that should be if I if I um you know think about it in the same terms that people people people do is that it should be very similar every solar panel every inspection every should look the same in reality it's not and it's not showing great results yet we have been through a couple of iterations with our um Partners doing AI for for solar I see I think that we actually in the next two to three weeks going to be able to uh release the newest um uh version of our uh solar um AI um and hopefully that would work now there is a big uh you know there is a big expect

**[34:15]** expectancy in the in the industry to see AI working for everything we have been on it or at it for last three four five years now and we took a we took a stab it we're we're taking a step back and we are starting to reform our thoughts of how AI should be used and could be used to make the work uh better more efficient looking at the different models out there and it's not necessarily the the image um or the issue fall detection that we are going to to be seeing automation on it might come in the form of analytics or different uh different forms of um deliverables you can ECT using AI so we are we are looking at it from the helicopter or bar ey view if you will rather than simply looking at the specific um use cases for applying AI H that's very interesting uh next up we have how do cloud-based inspection

**[35:17]** Solutions integrate with existing asset management software so there are different ways um currently throughout all of the market what we are doing except of a couple of project projects where we are developing very specific handles to different softwares Erp softw or work order systems um the most common way is using csvs uh taking that CSV changing the content of it to fit to the correct way the work of the system would want to see it internally for that specific organization and thereby using that for the uh for the integration understood so Danny that was it thank you so much for sharing you know all of these valuable insights it was great the demo was amazing I think it was very detailed and it was very amazing you know the audience had so much to learn from the platform and from you thank you thank you and thank you everyone for logging in have

**[36:22]** a great yes byebye byebye okay