# Improving Port Logistics with Automated Pile and Volumetric Analysis - Event: NestGen ’24 - Published: April 6, 2024 - Length: 31:31 - Watch: https://nestgen.org/library/watch/sabDf-NT2tA/ > 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 Discover how drone technology is reshaping port logistics in this insightful video from NestGen 2024. Delve into automated pile and volumetric analysis, showcased by David , CTO of Octocam Maps Explore the evolution of Amelia software, designed for efficient data collection and analysis. Uncover the potential of drone-in-a-box systems for repetitive tasks in expansive port environments. Gain insights into the challenges and opportunities facing enterprises adopting drone solutions. Join the conversation on regulation, data analysis, and technical implementation. To know more, visit: ===== Subscribe to FlytBase YouTube - Follow Us On - LinkedIn - Facebook - Twitter - Instagram - ## Transcript **[0:24]** hey hello everyone and welcome to another session at nzj 2024 Maybe before we begin Let's uh like start engaging with our audience so maybe we can start uh talking about or tell us about how have you liked these other sessions at nesen and do show us some love at the nesen wall of love so and also maybe you can let us know your thoughts in the chat box it will help us curate better sessions for you in the upcoming edition of n gen so before we begin this session let me ask you a question how many of you over here are joining us at nesen from cities with the active ports maybe we let's let's take a five minutes pause and let's see where our audience are coming from David okayy hello Tango hi Rohan **[1:28]** okay okay hi Nigel from us okay let's start I think your people are still joining let's begin so like it's like it's multiple folks from the audience here today obviously want to learn uh regarding the ports how we are CH solving those problem with the boort cities and well then it seems like this was quite an interesting session for you let me just wait a couple of minutes more hey hi Mike hi Jason right let's begin so I believe every might everyone might have seen the session's topic but let me just quickly introduce the topic what we are going to talk about so today we will be talking about how the process of monitoring and inspecting ports with large containers is being automated to help quickly detect files and provide valuable data **[2:29]** such as surface height and volumes of each file with just a click so please welcome our speaker for today today's exciting and informative session David F the CTO of OCTA cam Maps being Ed being an Industrial Engineer David found OCTA cam Vision in 2011 during his third year of college providing a wide variety of closed range of aial services all around the Europe and they have worked on Renown renowned companies in such as Netflix Marvel Studios NH HBO math among others isn't it exciting so several and after several years specializing in field data collection David found OCTA cam Max maps working with drones liar Technologies for terrestrial and aerial uh mapping and topography and the and then inciting EX in fact you know they have inspected **[3:34]** 15,000 wind turbines all around the Europe and map thousand of hectors isn't let's maybe everyone obviously would be excited to hear from David so get ready grab your favorite cup of cup of mocha because we are about to begin in the next few minutes right so but before we begin I just want to have a couple of housekeeping roles so that we can can have a a better session and we can get most out of it right so please feel free to type in any question that you might have during the session in the Q&A chat box this will they will be answered at the end after and at at the end of the session the session is getting recorded the session is getting recorded and will be published soon we will also have OCTA cam booth in the Expo session please make sure to visit visited and during the networking and Booth visits and obviously David will be there to answer your **[4:36]** queries and solve the challenges that you are facing right uhuh finally please remember to capture your favorite movements of the session and share them on the Twitter LinkedIn by using a hash NJ and you know a few lucky winner or people who who share might have a surprise in store and that also there's there will be an ongoing uh poll during the session make sure you uh you participate and it would be great if if you guys uh participate and will get to know more about our audience today last but the not last but not the least I would also like to thank our title sponsor DJI and our go sponsors sunflower labs for their valuable support during NestGen 24 and with that I would like to over share it uh the deck to **[5:38]** David and David the stage is yours all yours thank you VC I'll start sharing my screen yes please let me know if you can see yes David I can see that okay great so thanks for the introduction vinc um d David F and today I'm going to be talking about improving Port Logistics with automated pile and volumetric Analysis um so as as Vin said um I studied industrial engineering my my but my background is more into into business related with drones I started my first drone company in 2011 it was called ticam Vision uh and a few years later in 2015 uh we we started with with octocam maps actually with with ticam maps we offer uh Services drone services and uh we are starting with with software development so in **[6:39]** terms of of services we work um as vinc said a lot um uh in in the wind and solar inspections uh and we also provide a lot of mapping services with lighters um we are not that experienced with with software we started our first software two years ago it was called rcloud it is still running and what we do is um we compare uh Bim models for construction with Point clouds um then about a year ago we started developing Amelia uh which was from an R&D project that we won from a steel company that we will talk now so let's begin how did we start ailia project so um about a year ago a steel company uh from Spain came to us um as they were facing um lack of of stock control in their open space um storage facilities here you can see some pictures of of how the company look like um imagine like a huge company in which they have a lot **[7:42]** of raw material in in Open Spaces um a truck came in with um all the stock uh on it um and they wait that truck but still um they are moving inside um um their facilities the uh the stock piles so they were facing a a huge problem and they wanted uh someone to to try to solve this problem for them so they came to us uh and they say hey guys can you please try to see if you can manage any kind of solution using drones and we said okay let's think about it so how did we solve this problem first of all we uh used standard drone technology um to get detailed photogrametry of their open storage areas and then we developed a software which was the first version of Amelia that was autonomously detecting piles and providing valuable data such as surface hate and volumes of all of those piles so that's that's something that uh you can already do with some sofwares but it's uh not autonomous you you **[8:43]** need to um select that manually and it's a usually a tough process so we wanted to make it very very simple this is how the first um version of the software look like a year and a half ago um here you can see a white line that's that's the region of in interest and then you you get those um file detections with with all that um data so how is the steel company using this technology first of all what we did it was training their employees to fly drones so they were not at all experienced with any drone technology so we acquired their first UAV for them it was actually a a mavic no um DJI M um Phantom 4 rtk if I'm not wrun uh we made them becamea drone operators just to fly in their facilities so right now what they uh doing mainly is flying more or less every two weeks something like this so they use um our software to get fast and precise um data so with this solution right now **[9:50]** the company became totally autonomous so they are not dependent on thread companies to fly drones uh that's a really step forward for them um they have improved a lot their control over stock files and they have reduced cost and timing so um just to understand you a little bit more uh the audience I would like to know if um you guys are using any drone the box systems if if you are just you can click it on on on a um if you're not but you're considering that it's purchased to improve your operations you can you can click B um if you think that it will never improve your operations you can hit and and if you don't know this about this technology right David so the polls are live audiences right you can head to the sessions polls tab uh maybe we can we'll wait uh maybe David for a couple of seconds and then let's **[10:53]** see okay let's see and the results are quite interesting David okay okay so we see that almost yeah you're considering its purchase mainly to improve your operations but and some of them are already using it great so uh regarding Dr box systems um we uh are right now after this first software version that we did we started working with uh join a box systems um we did um a quick overview and understanding which were the top players in the industry uh we um realized um FlytNow FlytBase were we doing a great job H we also found um haa I personally went to to shenzen to to meet them and we finally U decided to start working with the d80 uh so right now the equipment that we have is a h d80 with the DJI mavic 3T and 3E and flight based software um so so um **[12:00]** we are using this technology but not in a repetitive way because uh we are facing a lot of legal problems at least here in in Spain Europe General so um that's the the second PLL I'd like to ask you so for those of you that you are using our pretending to use um thr a box systems what is your main challenge uh if it is regulation data analysis technical problems or or others now again the polls have live guys let's wait again for a couple of second and let's see maybe David I would also like to vote this time and uh and obviously would want to hear from you more on that yeah um I agreed always with regulation I think I think everyone agrees with that regulation is it's is a huge thing and also date analysis which I think it's it's very interesting too that's what what's our **[13:02]** goal our mission is to um to make everything full autonomous so I mean Dr a box uh systems um allow us to automate that but the data analysis is not that automated so that's what we want to do with themia all right okay so let's continue so how how did we evolve um so after let's say the success of the first version uh of of a milia uh in in this in that steel company uh we we started developing um ailia but in let's say a little bit more good-look and easy to use um willing to offer a commercial version of forend users so um let's say that our business model is a software as a service and you can pay for um using this software monthly so our first customers were expected as as we said to be mining and still companies but we found um interest in another sector which was Sports so at least in in Spain uh in **[14:04]** Spain Sports bulk solid minerals terminals are very very common um so we discovered that uh ports used to invoice concessionary companies daily or the space used um so of course this is a very let's say slow inaccurate and also quite expensive process and and it takes a lot of time for them to to do this thing so we said okay with ailia have uh that thing let's say solved quite easily and we can talk with ports to see if we can um start working with them so I'm going to show you a bit more how does the zilia solof look like um this is the last version which is um kind of more good look more good-looking than the the first one we shown uh you first just log in with your email and password and then you access that that general information where you can you you have a map H and you um have those different areas um that you have been flying once you hit one of those areas you can go a little bit **[15:05]** more precise um this is the port of taragona it's a port located 100 kilometers South Barcelona uh it's one of the important ports in the Mediterranean Sea and Spain um and so well here you can see different spaces where you can fly or you have been flying with uh with Amelia um so actually the the software itself is very very simple you can create a flight you just filter by data uh and then you upload those pictures um that you that you did with your drone uh you can upload if you use the Drone in box systems or um if you are flying just a manual drone with standard photogrametry so um once you upload that data um the software itself creates um um a 3D uh with a with a photogrametry and and then you just need to to hit or create a what we call a Roy which is a region of Interest that's the the white line and then the software autonomously **[16:08]** detects those files that are inside that um that region of Interest so um when you hit um those uh files you get the information um in terms of um area uh volume uh and hate okay so um regarding the the algorithm uh we developed our own volume analysis algorithm so um of course um we thought that what we did was great but we need to make sure that what we were doing was okay so we used groundbase scanners for comparing results um I don't know if you're used to that but um this is a pretty standard technology that is being used many in engineering companies uh it's a very very very precise scanner we actually own Lake RTC 360 because we are quite used to do lots of services with this in in Spain in our company so we we profit that we had that technology to scan those **[17:09]** files that we flown uh uh also with the Drone and then we compared the results uh with the Leica which is extremely precise around 2 millimeter so we finally tested and saw that the average Precision of the of Amilia with drones and photogrametry was around a 5% which is quite reasonable um so that's why how we did it um here you can see a a capture of of how the software looks like when you when you hit the 3D uh map so this is standard photogrametry that we got with a low density Point cloud and uh in case you have different piles also ailia detects them uh individually and once you hit one of those piles you can see here um well the area the volume and and height this is uh something you can download uh and and fill it in in Excel files uh so it's quite useful for that daily work uh in case you got those different files you can even select and and place some **[18:12]** ID or even more information uh even if you if you know exactly the density of those files you can have the um the weight of those of those files too so um regarding ports um How is right now Port of taragona using this technology in in the case of Port of taragona they were already trained and operating their own drones so that's something that um it's not very um usual in Spain usually big ports they used to um hire third um companies that um use drone technology for flying in into their facilities but in case of taragona they own different drones uh and they operate them so uh for them it's a great tool because they are just providing and getting good software analysis um to be completely um independent so um right now they are almost uh flying daily um from Monday to to Friday because of that um **[19:14]** um invoicing that I said so they use the software to get fast and precise data for invoicing customers so with this solution they have improved their control over stock piles and of course they have reduced costs of measuring and timings uh here you can see two pictures of the testing that we did um right now they are not using Dr a box Technologies but the first time we went there we showed them the the Box system uh operated by us um and we proved that it was uh working perfectly with uh with Amelia uh as you can see it's a it's a haia d80 with a mavic uh 3E um flly now over and and AIA for data analysis um so that's how the the the space looks like when when we did the test in so what are right now the next steps that we will be implementing in Amelia uh with uh with ports technology so first of all uh we will be adding some vehicle surface measuring at least in Spain there are a lot of ports um in which um we got boats with lots of vehicles and they Place those **[20:18]** vehicles in in big surfaces and uh ports is are invoicing customers regarding the area that they are they are using so it's very easy to fly the Drone uh to uh hit that data on to a million then to get those surface uh area that's one thing and then also container accounting that's something uh it's quite important um so we tested both things with Amelia for example here you can see um like like a simple um surface where there are lots of of vehicles placed uh you just we we just created a a region of Interest then we got the surface of of all those we need to of course improve that and and get more accurate data but right now uh it is working uh the same thing with with containers um so we are actually onto um different research and development um projects together with the port of taragona to search for some funding and and keep on going with with the software development so as a as a **[21:19]** summary a conclusion of of what I said uh we can say that automation with drones and and so is plenty of opportunities and Port is an example of of a market Niche but there are a lot of unexplored new businesses so um together I would say that drone box systems are ideal or the ideal hardware for for repetitive work uh in closed range areas however um we are still facing lots of problems at least for for legal implementation and and let's say a consequent democratization of of those systems so um uh that's my presentation I'll be glad to answer your questions if you have them you're muted Pak ohk thanks a lot for highlighting that but no thanks a lot David uh I'm pretty sure I got to learn a lot about uh Port Logistics right I was working with couple of customers around the world right and I'm pretty sure I have a **[22:22]** better knowledge in terms of domain right what are the challenges that they're facing in and I'm pretty sure that's what also audience would love uh would have loved about it okay and uh I can see there are a couple of questions coming in and let's maybe wait for a couple of seconds and open this round uh for the audience to drop their questions in the chat SE sections of the session tab maybe uh meanwhile the audience is dropping those questions David obviously we had a nice two two amazing polls obviously like one is are you you guys starting uh looking for to adopt drone in a box systems right the other was about what are the main challenges so what from your perspective what do you think I'm pretty sure regulation is one in Europe it's very difficult that is something we are coming up but what are the other challenges that are in or kind of while deploying those kind of dib Solutions in an Enterprise Port so maybe if you want to shed some light on that well um yeah for of **[23:28]** course regulation that's that's one of the first and then you know ports are quite extend extended area so um if you're operating in an a A3 um area usually if you do not have any B Voss um um accreditation then you will no longer fly more than 500 meters radio so um that's quite a a closed uh area so um that's one of the things that uh uh that should be maybe I don't know um maybe a a vtool system in a Drina box should be an option for ports because they're us usually big big big places at least parcel for example you have like 15 kilometers or 20 kilometers from side to side of the port so it's a huge space usually with with a standard drone or you may uh face or have different drone a box Solutions or uh you have a different drone on it to cover more space uh that that **[24:28]** could be a thing right yeah completely agree on that like we all would obviously help us in covering large area in a smaller time in just one flight so I think one uh there is one question from Nigel uh he's saying is the objective to count the number of container I'm I think it's a half of the question uh maybe he he he wants to know what is the objective to count the number of containers like what is the end goal behind this yeah actually what what we will provide is is um you just create that region of interest and then um within that region of interest you will know uh the area that all containers are being um you know um not placed and then if you have more containers on top uh you will get the number of containers because actually there are like two sizes of containers 20 and 40 feet um so the software will autonomously detect whether if it's 20 or 40 or 40 and then you will you will get the accounting **[25:31]** so number of containers area occupied um and that's it maybe David it may also help in space optimization of that particular Port because you are able to scan the area right this could be one of the use case I hope yeah depending on on the ports and how we're using them imagine that that you have a private company that's operating in a in a in a port they might be um willing to know how to place the containers in a in a better way in order to avoid paying taxes U that could be um a mission that ailia could provide for um but that's what we I mean we are not into the ports business we are more into drone technology so that's why we try to ask our customers what do they need to um to fill their needs and make them hire our services that's what we want to maybe like a follow-up question like when you said obviously it's you're working on a own technology not on a port business right so **[26:31]** what kind of other use case that aelia software uh can be utilized where it can be utilized for so right now apart from that steel um um company that we work with and and ports uh we are talking a lot with mining companies um because they are facing a lot of problems and usually they are they are used to fly at least three four times a year with with standard drones they they get their topography uh and they do a lot of volume controls but that's uh quite challenging because I would say it's it's easy to get that volume detection with with a flat um surface but when when you're in mining uh you got an irregular um surface and that's a little bit more challenging so basically what we are developing right now is we will compare two flights so we do a flight uh day one and fly day two and then the software what it does is uh you you have you compare one with another and you get the places where uh you remove that surface and you get those um volumes in cubic meters so you can **[27:34]** compare one with another and that's what we are implementing that's a little bit more complex but it's it's working more or less right now like a change detection of that particular side right so there was one more like I would say just two of them more so when we talked about postprocessing right so can you also highlight what uh like how much time does it takes to postprocess that particular site or is it done instantly or well no we we are using different servers of course that depend it depends a lot on on on the on the service that you're using if you have powerful uh then then that's going to take and of course the number of pictures that you have taken uh so um for example for a standard area it takes usually around 15 to 20 minutes to get the photogrametry fly and then once you just select the Roy the region of it's quite uh fast I would say within minutes or even less you get that that um analysis so it's **[28:35]** quite quite fast yeah right obviously um so I think there is one more question along it because we talked about drone and a box solution so what kind of features would the end user would need right or expect from that docking station solution right in order which maybe OCTA cam or Emelia provides like apart from the standard what kind of other things that is required to make sure or for this to have this deployment done for these kind of applications right well what we must uh work uh still with that is like to to for example connect directly um the Drone box software fly control with a media so that you do not need to manually select those pictures and then upload that to the software we want this to be a completely um autonomous um fature uh and that's something that we first need to to to fix the regulation to be able to sell those those R box systems and once we know that this is doable then **[29:38]** we'll be looking after that and try to make it completely fully autonomous so we want no Pilots we want know Engineers to make it data analysis we want everything to be autonomous and we just want the the guys from the port to um to get the reports every morning with with all the details and data that they want that's the idea I can see a lot of questions coming out there David maybe um I would request audience so we are already out of time but uh the audience can head over to OCTA cam booth and David would be available there if they need a oneon-one session or they need to know more about OCTA cam Ms or aelia software people can head to the sessions booth and maybe talk with them so now it's the time to end the session I was thinking we can go a little longer but I think it's already at time thank you thank you all for joining us today we hope you enjoyed and learned something new and let us know how did you find this these sessions or this session specifically I'm already biased **[30:40]** with David so also we have one more kind of a news for everyone who is joining nesg uh we have also launched the nesg community an exclusive invite only Community uh where all drone autonomy enthusiasts can will be joining and communicating so may please find the form in the chat section uh and you can join the wait list and further coming up uh next few uh minutes we have various interesting panel discussion lined up make sure to attend those and with that uh we hope you have en enjoyed a great uh n experience looking forward to a couple of more sessions team thank you David and it was nice to have you here thanks a lot thank you team thank you byebye see you byebye