hello and welcome to strauss's webinar after the blast measuring blast performance with ai my name is kimberly sahu i'm general manager at strayos presenting today are brad gingell and john lewis blasting optimization experts brad spent the first phase of his career helping over 100 sites optimize drill and blast activities as a mining engineer with the world's leading explosives providers orca and dino he then expanded the scope of his support to encompass all commercial and operational functions by moving into management consulting with the top tier firm boston consulting group brad holds a bachelor in engineering mining engineering from unsw and an mba finance and entrepreneurship from carnegie mellon university john is a graduate from the university of kentucky mining engineering program upon graduation he worked for the explosive manufacturer nelson brothers where he primarily focused on project management and blasting optimization for coal mines and quarries these projects range from bucket fill optimization to cast blast analysis he is now senior manager of drilling and blasting for metallurgical coal company a couple quick notes before we get started please make sure to put your questions in the questions section rather than the chat box that makes it easier for us to make sure that we've addressed everybody's questions and to keep track of it we will be sending out a link with the recording of the webinar after the webinar is finished so please keep an eye on your email it should show up in a few hours if you would like a certificate of completion please email us at hello at strayos i will put that in the chat box for everybody we will be emailing answers to all the questions asked during this webinar within a couple days so don't be shy about asking if you have a question and we don't answer it here we will email the answers out later and without further ado brad gingel and john lewis thanks so much kim um hi everybody and welcome to this a new stream webinar after the blast now this webinar is uh focusing on measuring blast performance uh with new technologies like drones and artificial intelligence so we're going to be covering uh firstly kind of the just a quick overview of strayos uh the company um who we are to give you a bit more background uh on us and a bit more context then we're going to dive into the reasons for doing post-classes and hospitals so we're going to be talking about why it's important and uh and how it creates value next we're going to be uh deep diving into uh use cases uh specific to different types of mining um and specific uh to different kind of ways of creating value for operators so um we're going to be looking at fragmentation analysis we're going to be looking at muck pile shaping and we're going to be looking at cast optimization and going over some kind of real uh real-life use cases on how to apply these techniques using latest technologies finally we'll wrap up with a q a you can post questions throughout the webinar and they'll in in the questions section as kim said and they'll appear uh in that in that list you can also upvote questions and so basically i'm gonna look at that question list at the end and i'm going to look at the top voted questions and answer answer them uh and as kim mentioned any we don't get to answer afterwards in an email so don't worry if we don't get to yours today amazing so let's uh get into it all right so uh a quick background on strayos so we are a software company founded in 2016 so about five years old we provide visual ai software for mining optimization this uh this includes drilling and blasting tools this includes ai optimization tools um and this includes a kind of bespoke custom built uh photogrammetry engine um as well um just to give you a bit of context on that we operate in eight countries uh customers uh use us at around 550 sites um we have a global presence with offices in the americas in india we have people in poland and me and sydney here as well which on aggregate means that we can provide 24 7 live customer support around the world we've done uh 75 000 plus projects with 50 plus ai models you can see the stats down there we we focus on on on delivering savings in value for our customers in the drill and blast realm and beyond uh so to give a bit more detail on what we actually do um what our software does basically as you see there operators and engineers use our ai powered tools at each step from mind to mill what this means is we have we've built a suite of tools which are useful to people at each step of the value chain so from the surveying side we have our own photogrammetry engine that you can upload pictures on to build 3d model and then we have pip planning tools and uh and kind of basic survey measurement and annotation tools that you can use um to measure volumes distances things like this from there you can we have automated ai tools to uh to do rock mass characterization work out jointing and structures in the rock things like that uh you can then take that 3d model with your rock mass information um design a drill pattern and then upload or export and then re-upload measure while data to get a full picture of what's going on inside the rock and from there we have a full suite of drill and blast tools so once you know uh once you have all of that kind of uh situational context of uh about the rock um you can then tailor your loading and timing to that and uh and predict what the outcomes of your blast are going to be when it comes to uh well it's a mukbal um which is something that we're going to be touching on today we then have loading and hauling automated ai assessment of this site so um auto detecting the whole roads um the the grades crossfalls uh berm heights uh you name it uh things anything that's that can be related to uh operational performance or safety uh of the um of the operation um from there tools on the system the processing side with our fragmentation ai so auto detecting fragmentation another thing that we'll talk about today in more detail and then uh finally on the reporting side we have a automated stockpile reporting so basically fly a drone or detect the stockpiles and then track um how much material is in those over time and the value associated with those uh and then finally vibration and air blast management so this is um it's in the reporting section um because you can use this for record keeping and compliance but then also on the records that are kept and on the information that's put into the system that's then used to build models automatically build models um which can be used to give you predictions at the glass design stage so that's that's the that's the overview of the of this rios platform um as i said the high level summary summaries uh tools to help people enter the value today now uh the the kind of flow on effect of having all of these tools in one single place is that you get an amazing uh amazing data lake uh that allows you to do full value chain optimization uh and this is this is really special um because having structured data all in one place is is really the key to being able to do things like correlate mill performance with drilling design um even uh correlate your kind of loading and hauling um performance with uh say which type of geology um a blast was from or which type of product blasting uh bulk product was used and uh and here on the side i've got the the kind of the the real differentiators are that if you want to do this uh generally and you you're using a lot of tools that are not all in one place uh it requires a lot of custom integration between all the tools which is kind of avoided in this case um uh often uh data is stored in kind of silos in in hard drives things like that having it all in one cloud database makes it easy for anyone to access and collaborate on um and also having it naturally be structured by geospatial dimensions and time dimensions um are kind of necessary for for doing any time based optimization uh later on so here i've uh i've just got one more slide on this um summarizing those last two slides basically saying we build tools for engineers and operators that are useful for them and then those that information that we collect can then very easily be used for optimization and that's what we're that's the sweet spot we're trying to play in setting up to be useful for for people in their everyday jobs um but then also as a byproduct of that creating all this valuable uh information that's set up already structured in a way to be used for optimization um uh to maximize value mine to mill cool yeah uh let's let's get stuck into why conducting post blast assessments is important um so i'll start off with uh with a slide that you probably are familiar with a chart that you've probably all seen before in some form or another um because it's fairly fairly widely accepted um that uh to increase blast results you you generally take a bit more drill and blast cost and you reduce your loading hauling crushing cost um basically all of the flow on steps in the value chain and it's generally accepted that uh you you kind of do this up into a point um where you reach an optimum and at that point uh the next dollar that you spend uh extra dollar you spend on drill and blast uh stops delivering you an extra dollar in uh in the kind of the follow-on steps and so that's your optimum and that's where you stop and i think it's uh it's uh the really the interesting part about this uh this chart is that it looks very simple um when you have it written there um but most operations tend uh everyone wants to operate at the optimum point but most operations uh in fact do not operate at the optimum point um they operate either somewhere to the left of it or potentially somewhere to the right of it as well and the key message is that it's impossible to know where you are on this curve if you can't measure your blasting results because it's impossible to know where you are on that x-axis and if you don't know where you are and you're not at that optimum point then there's always going to be a cost saving that you can achieve in your total costs by moving towards that optimum point and that's generally quite significant when you consider that all of the other steps in the value chain all of your other costs associated with loading hauling and crushing usually being quite uh you know large chunks of your of your total site costs so um it's a drill and blast cost tends to be uh quite a powerful lever in influencing your overall total costs and so i'll follow up the uh you know one of the most commonly shown graphs with one of the common most commonly uh quoted quotes um the real simple thing is you can't control what you can't measure if you don't know what your if you can't measure what your post blast results are you don't know where you are on that curve and so you don't know which direction to go you don't know how much money is being left on the table and that the only the only way to start optimizing your total site costs is to start measuring your blast performance and so why doesn't everyone measure their blast performance if it's so important well traditionally measuring blast performance has been quite difficult this has been for logistical reasons um like uh interrupting the rest of operations uh so for example if you uh you know you fire a blast you really want to re-enter as quickly as possible because every minute that a truck and that a loader are not operating digging out that shot when they could be is money lost so you want to really want to you don't want to delay anything to get back in and resurvey like you had to in the past with traditional survey methods and also once you are loading you don't want to stop people loading so for example on the fragmentation side getting back in there to take photos manually you know involving stopping loading operations is just something that is not worth it is not worth it to do because of the cost on the operational side so there are logistical challenges um with conducting post-blast assessments uh on the other side uh there are technical difficulties in the past um where for example conducting uh conducting assessments of throw or mug pile shape would require uh kind of specialist expertise you know you you might get the the standard survey model but you need to have an engineer who knew how to do that particular piece of analysis using uh using specialized cad software uh and and at the end of the day you'd need that person to be doing it the same way in every single uh after every single shot your numbers were standardized um on the uh fragmentation side a lot of the time the fragmentation software comes from third parties so it's not integrated with the uh with the blast design software um so that's been a kind of a specialist uh skill off to the side um and the result of this is that a lot of this post blast assessment work is has often been outsourced rather than conducted in-house um by quarry mine operators themselves who are the ones who actually stand to gain the most from this data cool so that's the so that's the the reason why it hasn't been done before um so kind of the the cool news is now um a lot of these uh a lot of these challenges have been addressed um by new technology so uh these days uh the two key technologies that have kind of disrupted this have been drones and ai uh artificial intelligence so uh what they've what they've really done is uh is made this uh it's much more streamlined and uh and simple and much more accessible to everyone um because really to get the most value out of this you want it to be a continuous improvement process um where it's kind of built into your day-to-day workflows and it's something that you can do after every single shot um and somewhere track and manage all of that data in one place to identify trends and then ultimately optimize operations so uh to go into more detail on how um these new technologies have come along and uh and and kind of uh fixed up a lot of those challenges uh drones have now allowed surveying to be done without disrupting operations so if you remember in that past slide one of the issues was once you've uh once you've fired a blast you want to get back in there and start digging as soon as possible well now you can start you can get your your loaders moving in and at the same time as soon as you fired the blast you can fly the drone straight over the thing and uh and take some photos within a few minutes uh come back and you've got your survey data uh next um the analysis so uh like i said on the on the previous slide um it used to require kind of expert um uh technical skills uh to do a lot of these analyses um but now thanks to ai a lot of these can be done automatically um so this this doesn't involve having a having a a drill and blast specialist there to do it it's a it's it's basically a box tick um on the upload and then the the ai um can automatically assess things like uh cast percentage and uh fragmentation things like that and then uh the final point um is that uh now that the data capture side and the analysis side has been streamlined um the future blast optimization has also uh been able to be streamlined so this is something that ai is also really good at looking at all of the um the the results of that analysis from your previous blasts and uh and then telling you okay this is how you should design your future blasts in order to get the best uh the best results or the most the optimum results to achieve the best cost total cost outcomes cool so we've uh we've gone over why um why it was difficult in the past why we can now do it today much more easily so what are the benefits that people are seeing uh essentially sites um that we work with that have been doing this that have been analyzing their blast results they have been seeing uh improvements um in in kind of two ways um production rates uh so allowing them to impact the throughput of their total mining system and improvements in the cost efficiency um so on the uh two examples of that uh i'll use improved mug pile shape um you can increase dig rates with certain um pieces of equipment um which is the increasing your production rates and then your cost efficiency side you can be reducing wear on your loading equipment um uh to uh to reduce your total cost at that step of the of the chain and uh and also improving safety um if you have uh uh in some cases uh like lower uh height muck piles uh and on the fragmentation side for example um improved fragmentation uh can boost plant throughput and in in certain kind of metal mining context for example and that can just translate straight to extra dollars um of revenue uh and on the way if you uh if you reduce the wear on your liners and grinding media um then that also reduces your costs so you're basically getting a two for one deal you're getting improved revenue reduced costs and finally again in the uh in the in the middle context uh improving yield and recovery along the way um is a is again kind of you know free money at that at that step of the of the value chain um but this is also something that can be true for uh in a quarry context as well um reducing fines percentage increases your increases your overall yield essentially as well um which is another form of kind of free free money compared to what you were doing before nice so uh let's move on to uh the three use cases that we're going to talk about today um we're going to talk about fragmentation optimization which is basically achieving achieving particle sizes closer to the feed requirements of the plant uh and in particular the primary crusher uh we're going to talk about muck pile shaping so that's matching the mug pile shape height to the optimum specification of the digging equipment um looking at different types of digging equipment and we're going to look at cast improvement um and this is where john's going to take us through um kind of an in-depth example um of a case study that that he's done in the past uh and i guess the uh the high level um summary for this is uh essentially throwing more material to the final position uh reduces the amount of re-handle that you have to do and therefore reduces your overall costs great so uh let's uh let's dig into the first of the use case deep dives fragmentation optimizing fragmentation is a balancing act there's obviously nothing no rocket science here obviously cause fragmentation can slow down your slow down you plant increase wear you can get require secondary breakage if you have too much oversize um and sometimes oversize can be uh so large that it's not worth breaking and you uh reduce your overall yield from the shot um on the finer fragmentation side for certain operations you get more fines you can then lose product there and and basically decrease yield and also the other obvious cost of finer fragmentation is um is often higher drilling and blasting costs because you're drilling more holes using more powder uh often and the key to finding the balance here is surprise surprise matching the requirements of the plant so what this means is finding out uh what your current distribution is what that looks like and finding out what your plant's optimum feed distribution size is and then kind of running a continuous improvement cycle to basically adjust drill and blast parameters measure changes and work towards an optimal point where those two distributions match each other as closely as possible so uh in uh in theory this uh this sounds simple and makes a lot of sense in practice um there are often a lot of challenges associated with doing this um so on the measuring side uh like we said before you have to capture the data uh properly quickly um you have to then use that data um you have to store that data in a way that allows you um to assess trends over time and link that back to blast results uh blast inputs so that you can you can see the effects of your changes so that's on the on the kind of the difficulties in the left-hand side on the right-hand side it means you have to go and talk to the people in the plant and you have to find out what what their specifications actually are um and a lot of the time if the plan hasn't been run uh outside its current parameters in a long time um then they the plan operators might not have the full picture either of how the plant's going to react to to different distribution sizes of rock coming in um if for example you've been kind of doing things in similar rock um with similar blast patterns um the same way for the last five to ten years for example so the uh the answer for technically how to do this and why this is uh why this is easier now is that uh drones and ai uh have basically come in and made this process a lot faster and a lot safer so i kind of split this into the three phases um of planning taking the photos and then processing them um and uh and showing the kind of the more traditional methods um we're kind of dependent on waiting for gaps uh in loading operations because like i said before you you really don't want to stop loading um and uh methods that that uh kind of require a manual scale and kind of single image at a time processing uh tend to be prone to errors from things like shadows and and things like perspective error as well whereas now when you fly a drone over because each of the photos are geotagged it creates an automatically scaled 3d model with no error from perspective things like that no human standing in front of the dig face potentially getting hit by a rock and and uh but it's not just drones um that you can use uh obviously camera mount equipment mounted with cameras avoids that issue as well um from the safety perspective um and also uh the ability to just use uh geotagged iphone photos um from a cab um also means that you can be out of the way and uh and not have to uh not have to stop loading operations uh in the taking photos section um we've got uh yeah humans walking in front of the of the of the dig face was definitely a big safety issue um retrieving manual scales is something that's another safety issue and it's something that just takes a lot of time of uh of kind of specialist contractor contractor hours which um is a is just a significant cost um for the uh for the study whereas now flying a drone from a safe location in a few minutes um dramatically cuts down those costs and those safety factors processing side the uh so kind of earlier um versions of software that use edge detection by uh by highlighting like darker versus light zones are prone to prone to shadow errors and kind of manual adjustments that were required were quite time consuming whereas now kind of fully automated ai fragmentation analysis is just a tick of the box that can come back in in 30 minutes um and a kind of a web a web portal rather than um having to have any kind of standalone powerful computer so that's uh that's basically the the way that drones and ai have improved fragmentation in the process of doing fragmentation um studies um definitely feel free to reach out if you have any more questions on that one um happy to happy to go through that in more detail we'll now jump into mukpile shaping so uh first i'll kind of explain what i mean by pile shaping um essentially every different piece of equipment has a different optimal mug pile shape so for example loader may prefer a lower flat mug pile an excavator may prefer a taller steeper muck pile and uh and the reasons for this are just the due to the structural way that they they're designed and broken where they sit while they while they dig out that muck pile um you basically want to minimize the movement of that piece of loading equipment as it's digging um let me move on to the next one so optimizing uh optimizing the muck pile shape the way that this kind of creates value is you can get uh you can get a lot of extra productivity out of your loading equipment if you can get the the mug pile to the optimal shape so uh every piece of equipment has kind of a maximum dig height um imagine in example one that you have a mug pile that is just slightly higher than um then twice the the maximum reach of your excavator um or shovel um potentially and then so you have to doze the top of the mug pile down um to get it to that to that level or potentially take a third uh flitch off the top um with your excavator or shovel uh then let's say that you can do something like um predict a mug pile shape change the blast parameters and get that blast mug pull you know to be maybe it's only a few meters lower you may remove the whole need for one dozer on site um or you may you know cut down the total amount of time it takes uh to load out uh one uh one shot um by a third because you that excavator does not have to make one uh one extra kind of pass um along the top of the pile so that's a that's kind of an example of of kind of reducing the mug pile height uh in example two we'll look at potentially increasing the mug pile height helping um say you have an excavator in the bottom of example that has to move back and forth across the whole length of mukpol getting it stood up higher may allow them to stand in one place and dig the whole thing out without having to move as much so that's essentially the kind of two examples of of how you can uh optimize your uh your mug pile shape for specific loading equipment um to improve productivity at that step of the value chain so what is uh what is coming out with new technology that helps this well uh drones and ai again are the two key factors so basically being able to one measure the muck pile by taking drone pictures like we said at the beginning means that you don't have to interrupt digging operations you can just come in straight after the blast take a bunch of photos and uh and then turn that into a 3d model of the entire mock pile um the ama comes in where uh now i've always uploaded those pictures tick a box and you can automatically detect the boundaries of that mukpl and automatically connect it with your pre-blast shot design and generate analyses like the kind that you see here um automatically measuring trough drop center of mass movement and your average throw things like that um all aspects of your mug pile shape so rather than in the past having to do this with kind of an expert with uh with cad tools specifically set up for this kind of thing um it can it's just all automated um with the uh check of a box uh on uploading some drone images uh so that's that's your measuring side um also on the predicting side ai comes in um where it's dramatically increased the speed at which you can do uh mukbang predictions so what this allows you to do is again last parameters and then quickly go back and re-uh re-predict your mock pile and then go back re-predict your readjust your parameters and uh and ultimately get a very uh get your ideal uh mug pile shape um in a very short amount of time on the design front and this is something that you can calibrate with your actual measurements um done in that in that first step so yeah overall as a as a summary mukbar prediction mukpil measurement is something that's now now able to be automated which can generate significant value for mining operations um at the at the loading step of the value chain so next i'm going to move this onto cast optimization and here i'm going to hand over to john yeah so cast optimization thank you brad for the introduction so like they already said uh previously i worked for um an oraca joint venture nelson brothers and during my time uh cast optimization was a very big thing because i mean that's kind of what uh surface coal mines are going for so getting right into it uh on this first slide here so cast is kind of your biggest lever for reducing strip mining costs as you can see from this diagram here on the left you have uh you have your bench which is your your your left line you have your muck pile and then you have your cast line so your cast line would be equal to whatever your your dozer push uh angle would be or traditionally even your angle of repose of just your rock so you can see at the bottom you have the cast to final position which means there is no re-handle required you do not to come in with another piece of equipment such as a dozer or an excavator or all sorts of uh you know i mean excavator haul truck and then move that to its final position all the material above your final cast line is thrown short which means now you have to introduce manual excavation which any time you introduce another piece of equipment even if you're just re-handling a little bit that's an additional cost so more material cast to final means less re-handling so reduce energy and fuel consumption that's money reduced equipment wear that's also money and increase production capacity of pit operation that means more money so now you're spending less money to generate more money so that's why uh cast analysis is a big thing that you should be tracking if you are in a cash blasting scenario even if and granted even if you're not say you're just in a quarry it's just it's a good thing to understand from the mukpile shaping perspective like brad was just talking about so if you go to the next slide please so really there's two key cast metrics to measure there's the cast percentage so which is the proportion of material that was thrown to its final position in the low wall spoil pile as highlighted from the green so it indicates the amount of material that must be re-handled so that's just a percentage of the entire shot so now you can say you know 20 of it doesn't have to be re-handled 80 of it does you know flip the metric just depends on whatever your your wording wants to be for whenever you present this information uh there's also the center of mass movement so that's the distance that the centroid of the shot moved in the blast trust me you know i remember calculating centroids in like deformable solids you don't these pile shapes indicates how are all has been moved really well here further on your next after your then you kind of see that hey you're going to be need to be bit john i'm not sure if we've got you you want to try rejoining or something john while he's uh uh disconnected and reconnecting um i think that the the summary of the center of mass movement part um that we missed was uh that uh yeah calculating centroids is is tough it's a it's a very annoying uh geometric um uh kind of mathematical um i guess task to solve problem to solve and it's it's something that you have to do manually for multiple slices throughout a shot and so it's it's it's it's something that's time consuming and also something that you um need to have a kind of a level of expertise in order to actually accomplish uh and so this is uh uh something that has been it's important because it uh it it tells you how far everything's uh everything has moved um and it's also something that's traditionally been challenging so oh we've got you back canadian quarantine hotel [Music] so now i'm just gonna do it do this on the on the lte so hopefully that works better yes so uh man all right fun times so how to run uh cast optimization so first what we'll see here or go through the general way of doing it and then we'll get into kind of like how i did it uh from an actual uh start to finish perspective so preparing you're gonna kind of align your plan with the blast crew engineers management setting up your tools and equipment such as your drone your software you know if you need ground control points with a gps or if you already have an rtk drone the most important step is you have to baseline you you can't measure what you don't know kind of similar but if you just assume what is already there then you're just going to destroy all of your results going forward so you have to have an understanding of what you're already doing and what you know and establish the your kpis beforehand so you know how they are versus what you're going to do in the future so suggestion to fly the drone over your pre blast and your post blast bench for five plus baseline shots you also want to make sure that these are geo reference models so with either an rtk ppk or a gps unit um you could try to get away with not doing that but it's going to skew all your results and introduce error that you can't account for um you're going to upload you're going to input your blast design info you're going to check your cast percentage and your center of mass movement so now you know your kpi's on your baseline and then you're going to adjust you're going to make sure that you align your target cast profile with the engineers in production management you want to consider all sorts of things when you're shooting now there are certain situations where you're not going to be able to get an efficient cast like if you have a drag line pad that you don't want to just cover up with rock or if we just excavated a bunch of coal in front of us like maybe i don't want you to cast this panel as much because we still need to get this coal out so there are a lot of things that you need to consider uh going forward and just to make sure because as much as we want to optimize these things if you irritate your managers and the people running the production then what's the point of even doing the optimization in the beginning uh you want to test varying timing and products in the prediction model and adjust the design for the blast crew it's very important that we don't just go adjusting 17 different parameters all at once because you're not going to know which variable actually contributed to your performance increase so you want to adjust these variables say one at a time you know maybe start with a pattern change maybe start with you know a loading change so on and so forth and then you're going to measure so you need to fly the drone again pre and post blast just similar to what you were doing with your baselines and then track the performance increase and now the reason there's an error going back to adjusting is because after you kind of isolate one variable you get it to where you want it to be you then should go back and adjust more variables and see you know you can't you don't know where you fall on that line that brad showed in the beginning unless you kind of know okay i went too far or i'm too short and now you can figure out your optimal area so you need to adjust get it to your efficient area and then measure it again and so on and so forth for each variable so the way that that looks just fly the drone pre and post blast establish your kpis and then just keep doing the same thing and that's essentially all i did except i was doing it very manually which you know you're paying an engineer to go out not only go out in the field to collect data but then sit at a computer for you know three four hours putting together a pretty report um that really should be automatic and it's not that it's annoying because if you want to do that all day good for you but the company probably doesn't want to pay you to do that when they could just do it automatically so next slide please so this is a real world example of a cast panel that i was a part of so you kind of have your cast optimization prediction so let's just say for example and this the model predicted three percent more cast after making design changes and you can see uh the 3d model of the actual cast panel where you can put the drill holes and then the bench bottom and then on the bottom right you can see where those drill holes are you can see your uh your mock pile shape with the blue line you can see your cast line with the uh the yellow plane and then where it intersects uh at every drill hole which is you know kind of where the yellow meets the cyan and why it looks like a kind of a wave so on this next slide we can see this is uh i know this probably looks the least professional out of all the slides but i wanted to put it in here because it's it's proof of the reality of the information so as you can see here uh all i did before when i was trying to do it kind of manually is i have the pattern i knew it was a nine inch bit i know what my the top elevation and the bottom elevation of my uh bench was because it uh you know i think it says 59 foot uh drill so you know just do that and then i know what the ground elevation is so that allows me or allowed me to essentially triangulate my bench and then triangulate that muck pile to the ground so really the steps to analysis that i was doing um which should be automated are creating those individual boundaries around your uh around your mug pile in your bench triangulating the tops of the shot with the individual um you know bottom areas and those boundaries triangulating the bottoms within those boundaries so you're essentially snipping off some of that ortho photo to determine that volume doing the volume based on the triangulation so your muck pile to your bench and then generating your uh i call it x which is cross section and kpis per cross section again uh i was doing based on cross sections because you know i didn't have an automated 3d model i was doing all this in autocad and then determining the swell factor based on the volume information which is just um you know post blast over pre blasts of my pile over uh over bench all right so going back to what that post blast looks like for that cast analysis is now you can see exactly where those drill holes were in relation to um you know what we actually did and then now you can see the actual cast line obviously we don't have that little cyan overlay anymore because we have an actual mug pile there's um and so you can see that you know a cast improvement of two percent and this is going to be based on your real data this is not the prediction information so then what does cast improvement actually mean you know we talked about re-handle and i guess we can all think okay re-handle you don't need to introduce additional equipment so let's say we increased the the cast improvement by up to five percent over the next five blasts so we're not only we're increasing the production rate of 125 loose cubic yards per hour we'll keep it in the volume right because as a we're casting this so we're not running this rock we want it it's all waste so i don't know about most people but anytime that we deal with waste we just keep it in volumes and uh realistically decreasing costs by you know one one cent per loose cubic yard which is a very you know i mean if i get a decrease cost by one cent per least cubic yard that's a lot of money and it could you know even if it's 0.0001 or loose cubic yard that is savings and uh i'm pretty sure anyone would be happy with any amount of savings when you're doing you know lots and lots and lots of moving of rock okay uh overview john of um ever of cast optimization and your experience doing that project in the past um and i guess as you alluded to a few times um the kind of the manual process uh of measuring taking each of those measurements individually um using autocad to do a lot of the cross-sections um and calculations um is something that's now uh that now has been automated um using ai uh ai tools based on uh on drone models so um to kind of wrap uh wrap that up um it's great that that these days you can do that you can get those same results um you can get all of those same cost savings um without the investment of time um that it took in the past so amazing i'll um uh i want to thank everyone for coming in in closing i just want to ask one more question that i'll ask you to throw in the chat um which of the metrics and kpis do you use to assess blast performance that we haven't covered today because i'm i'm aware that this isn't a comprehensive of a post blast matrix so um yeah everyone feel free to to throw in the chat the other things that you measure after after you blast or the things that you would like to [Music] measure so i do yep yeah back right there also just a heads up guys it is 9 05 eastern time so we are five minutes over we want to make sure we don't um keep anybody overly long we all got jobs to do and or beds to get to so if we don't get to any questions today we will email uh we will answer all of the questions and then email out those answers to all the participants yep that sounds like a plan so again thank you guys for coming and we'll see you at the next webinar thank you
2021-06-12