# Enhancing Mining Efficiency through the Combination of Drone Autonomy and AI

- Event: NestGen ’23
- Published: March 13, 2023
- Length: 41:30
- Watch: https://nestgen.org/library/watch/NidArsOgCzM/

> 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

Learn how drone autonomy and AI in mines improve efficiency and safety by automating repeatable missions and generating real-time insights.

## Transcript

**[0:08]** thank you good evening everyone and uh a very warm welcome to you all um I'm very honored to to be hosting this session with the the topic of enhancing mining efficiency through the combination of drone autonomy and AI um a really great uh um privilege to be here with and a big thanks to the 10 and the team for hosting this next gen session uh thanks Natasha next slide please I thought I would um kind of start off with the the sort of presentation with the with the poll a little bit of a controversial question um and I'm very very keen to to get a lot of your uh a lot of your feedback to know who's in the room today but really to to get us kick-started I'm very interested to know how do many of you think that the implementation of autonomous systems and AI in general really will impact the workforce uh with specifically in the mining industry you know will it a create new jobs with more specifically in fields like

**[1:14]** data analytics and and maintenance uh career paths will it be displaced workers where autonomous technology will will really sort of uh replace the people on the ground because technology just can do it a lot faster and quick and safer than that when humans can or see will it have no significant impact on the work for Force at all if you click on the right that you'll see next to there's three little options chat polls and QA if you could just put it into the polls there and we'll we'll swing around back to this later thanks Natasha so looking at the next slide a little bit about us um and who we are so my name is Chris Clark I'm the CEO of Delta joint International which is a global drone mining services provider to mining companies across across the world we started the business in in South Africa and grewed up through Africa and into to Ghana and we're

**[2:14]** now currently based out of uh out of Australia with offices in Perth and Adelaide and we're listed on the Australian stock exchange and really this business has been going for for 10 years built on the the so the precipice of being able to to do everything for the customers really from from the regulations to the maintenance we're really sort of full position ourselves as full full integrators uh in how we deploy these solutions to the mining industry thanks Natasha so a little bit of brief overview of the mining industry for those who may be unfamiliar well mining may not really be a sexy subject and uh and generally a first-time career choice um we we really got to sort of consider mining as really one of those sort of Baseline sort of Industries uh it's a really important contributor to not only our local but the global economy at large while many of us may think like mining doesn't sound so sexy you know if you think about how we're going to create the the

**[3:17]** batteries the lithium batteries these the electric electric vehicles of the future this all comes from very highly intensive Mining and how we disturb the environment and how we remove the topsoil to get to those all bodies um in in a particular way we're becoming a lot more focused in in the mining industry on efficiency and being better custodians of the land and the resources and the environments in which we operate so you're finding a lot more focus on these soft and then hard factors such as ESG which tends to put a lot of these sort of pressures on on the mining industry so it's a very interesting space to be in mining where we need to get product out of the ground we need to create new products we need growth we need economies to grow but how do we we do this in a sustainable manner of taking into account a lot of these these other factors and how mining impacts people and our economies and the environment around us thanks Natasha

**[4:21]** so one of the very interesting things when we started about 11 years ago was we really underestimated the in the the um so the immediate uh impact that drones would have in the mining industry um you know when we um I sort of remember the first day when we sort of introduced this technology I remember uh walking into a surveying office and remember seeing this broken drone on the floor lying in a thousand pieces all over the place and I and I remember saying to the chiefs of it at the time I said you know Jacques you know what what happened here and he said well you know we we bought this drone it was a really big it was like a hexacopter and um he said it was costing me like 50 000 us at the time I think it was the days before you know I mean I think when DJI still had the sort of those m600s and very developmental platforms and uh he said we bought this drone they said it was the easiest drone in the world to fly and I remember saying to Jacques I said you know but if it's so easy to fly why is it lying all over the floor in a million pieces

**[5:24]** and he said look it was was very easy to fly just wasn't so easy to land and I remember at that sort of time that light bulb moment saying well you know drones and for many of us here on um involved in the Drone industry you know there's this sort of uh perception that you can just really sometimes take these things off the shelf and start flying immediately but many of us really kind of sort of failed to recognize that sometimes the customer's not necessarily just looking for a drone they're looking for what that actual drone produces which is that data it's that photo it's that that output and how that has come across within the industry and we realized this quite early on that you know it wasn't always just necessary about the Drone and the Drone Tech and the hardware itself but it's about the actual outputs in which it gets delivered and Mining has been one of those those Industries which is really grappled onto to drones so quickly because of its immediate safety benefits that it's uh that it provided um if you take in context you know if you wanted to calculate um the volume of a

**[6:27]** stockpile and how that essentially how much your mind is digging on the ground you would have to send men who would manually walk these 30 40 meter high stockpiles sometimes these could be coal stockpiles really um which sort of have the propensity to spontaneously combust so your shoes would be melting even as you're walking over these stockpiles and drones immediately being able to remove these um these sort of surveyors off of these stockpiles to give that immediate sort of safety value and we really were quite excited in how drones have just become part of the the day-to-day Norm in the mining industry and are only are not only growing thanks Natasha so when we're talking about drones and drone autonomy now this is the really exciting um sort of aspect of drones is you know when we first sort of uh started playing with them you know there became a very manual um if not sort of a toy based sort of product um you know we we would have you know sort of our fingers and manual sort of

**[7:27]** inputs in there and these slowly progressed into to being a lot more assisted um and sort of systems right so when a lot of if you think about in legal terms we talk about an RPA or a remote pilot aircraft uh or in our Paz which is a new pilot aircraft system and the reason why you're hearing maybe a lot of that sort of legal jargon is because they realize that drones don't just operate on their own but they they operate within an ecosystem of sensors and software which helps the drones provide a lot of assistance to the pilot in the way in which it operates and that is brought on this real Advent with the with the um with the Improvement of these sensors and the software as well as flight planning software as well that drones have now become have moved from this manual and assisted and semi-autonomous sort of fashion really to what we're seeing today is an autonomous and Sunni a fully autonomous system and when we're referring to a fully autonomous system referring to

**[8:28]** drones that can that can operate on schedule they've got um they're fully integrated with a detect and avoid systems so they can operate safely within the airspace within your predefined sort of parameters and be able to carry out missions without any kind of human intervention at all now there's a lot of Regulation which we'll chat about now that's involved in getting into this level again you know when you're looking in our RPA or a drone you can look at it the whole system which involves the software the hardware and the regulations as well within that and uh but we're seeing a lot more Regulators being a lot more open to these autonomous and fully autonomous systems which really makes it quite exciting and this autonomy is just one single ingredient in of how what you integrate into your overall offering to to the customer thanks Natasha so another agreement which you want to consider is again coming back to the original

**[9:30]** story that customers don't necessarily always want to just buy the hardware it's about the data it's the problem that you're trying to solve at the end of the day and and what we are um same as everyone I hope everyone can hear me okay on the side there um so what what again would you want to look at is what is the actual problem that you're trying to solve and that really is around folks around the data the data elements the the question the questions that you need to answer for the customers and so when we started our original business we were sort of really processing in a in a very singular sort of fashion we'd go out collect take a lot of photos come back process and output 3D models or 3D Point clouds and uh and so it was a really sort of a one two step sort of process but what's very interesting about the mining environment is that it's a very complex uh and very dangerous kind of environment it's constantly changing the road that may be here today may not be

**[10:32]** there tomorrow so so so everyone who's trying to operate or even to drive on a mine has to be very keemy and aware of this changing environment um and what may be happening around them and what this this challenging and very fast moving environment that creates is an opportunity for very very frequent data and when you start flying your drones every single day in a mining environment it becomes it becomes a bit of a sense of information overload because well you make it a very nice picture in a visualized sense of what the Mind may look like at the end of the day the Mind needs to make a decision they need to calculate the difference between what happened yesterday and where the mine is today and what what is the trend and where that is going and this is already unique parts of but where artificial intelligence or AI now comes into play because ai ai is able to take all these models all these data sets that you're capturing on a daily basis and be able to to synthesize information between each of those models to be able to give you a

**[11:37]** recommendation or at least a red flag AI is not perfect I mean I don't believe it's with the um at the level of Terminator yet where we think robots will be taking over the world just yet but AI is definitely giving humans the ability to to work a lot faster and make and process a lot of information in a shorter space of time and and this is where we're seeing it very exciting coming into into the into the mining environment thanks Natasha so if you take these two ingredients right so it's the Drone the Drone autonomy and the AI and now you combine them together we have a very special um uh sort of sort of recipe that now starts to form because what you're starting to do is if you take AR models um like let's say for example in this particular image which is detecting people and vehicles and you're um and you're applying them in a um to a thermal to onto a thermal sensor as well you're able to start using applications within security so

**[12:39]** what what this means is that you can deploy boxes and autonomous solutions to a mine site or to many other let's say even kind of dangerous areas as well be able to fly these drones on a free frequent basis and let them fly and stream and only stop and Trigger once they've actually detected a vehicle or a person or a particular situation and this is what creates a very um a very interesting um sort of opportunity for for drones and drone autonomy and AI combined together thanks Natasha so combining this sort of all together and thinking about a lot of these opportunities that um that we that we're sort of seeing where you you're able to to deploy Jones on a very frequent sort of basis but as well as overlay these AI layers what would you think would be the most promising applications of this autonomous Technologies in air in

**[13:43]** the mining industry do you really see uh there's just the primary focus is going to be around the safety and reducing accidents as the main driver for for autonomous Technologies and AI in the mining industry or B do you think it could be in the productivity and efficiency another interesting sort of angle to look at is is the decreasing environmental impact increasing sustainability because the more information that we have the more that we know about how much we're impacting the environment the quicker we can we can take to put in those short-term interval controls or the the ability to rectify it before the problem gets too big enough because sometimes we don't know and and drones get that really great aerial view to understand how we are impacting on uh within our environment so so taking I guess a bit of a look at the um I'm not sure about the polls at the moment how they're sort of tracking um at the moment Natasha but um it would be very

**[14:44]** interesting to to sort of note where everyone sort of stands um in um I think within the first poll but um I think we'll come around we'll probably come around to that one thanks Natasha we'll go to the next slide so this all sounds really well and great and very sexy um and uh but there's definitely going to be some challenges in implementing this because again while a lot of a lot of these products that you look at look very very sexy and very easy to use and and typically it's they're being sold as off the shelf I think for for many of us in Industry no this is not generally always the case there's a lot of integration involved there's a lot of regulatory hurdles um you know you're going to hear terms like Beyond visual line of sight um implementing your remote operating centers some of us operating jurisdictions where this is is not a concern but for many of us you know flying drones um you know I mean remotely without a pilot without someone to be there to check the

**[15:47]** weather to check the airspace it really frightens a lot of a lot of regulatory authorities and the ability to prove this puts a lot more pressure on you as the organization because you have to prove you you've got to be able to simulate relate a pilot as if they're on sites and how we operate these autonomous systems um remotely as well again on the second point the the key thing is we're not we're not here to sell drones we really care about selling the data we're here to solve a problem detect gas leaks uh detect Intruders we're here to to really understand the problem is not only just Gathering the data but also making sense of it IBM coined a very interesting term where we don't just have something called AI but we have something called xai which in other words responds to explainable AI so AI that not only can um can just operate a result but actually a result that you're able to track back

**[16:49]** step by step you're able to validate the ai's methodology and how it got to that particular answer because sometimes people think about AI as this big black box you throw a lot of problems in and you get an answer out and thus the machine must be right but there's a lot of customers and Engineers out there that want to understand the steps to how did the the AI get that answer and it's really important that we all understand the same way how we we understand the fundamentals of batteries going to drones and how lithium batteries should be handled and taken care of and you know charging in 40 degree heat will lead to a certain situation of probably the battery degrading and over time it's the same with with computers systems and programming is we've got to be able to break those apart and into their into their basic fundamentals and to be able to help the customer um deliver them a consistent onset day after day when you're deploying a lot of these during the box systems um you're also going to consider the infrastructure it's not just about the Box themselves we're seeing a

**[17:53]** lot of a lot of requirements around weather systems around connectivity and especially even power and Off the Grid depending on a lot of the the countries that you go into and the safety of these systems and the wind impact of this as well it's not it what's quite interesting is to then see the sort of the cost base of you know either taking a pilot to single pilot on a drone and going out to site versus deploying the box which requires a lot more capital investment to to set up these permanent um sort of installations but they're not only recovered over time so you're going to find that there's going to be a particular niche of applications which are going to be well suited to this which is going to speak to the high frequency of their data capture and that's where AI comes in and why it's uh it's uh it's such a valuable proposition because if you're deploying a lot of these nested Solutions AI is going to to help you just process their data a lot a lot quicker

**[18:54]** what's going to be very interesting is that skills in that knowledge Gap you know when we sort of chat about um the poll one it was um it really was about what do you do with the current people on site because you you're gonna have these new systems these new electronics deployed you're going to have a lot more information and a lot more data coming through who's going to make the final decision because you're still going to need a human being who's going to supervise a lot of the equipment and and those methods and the last one speaks to integration and interoperability a lot of customers have spent a lot of time setting up a load their geospatial systems their security systems how does your does your offering then speak within within that and integrate into the customers API and present accordingly because remember at the end of the day the customer is not looking for the Drone they're looking they're looking for the data and uh and their particular output thank you Natasha so that sort of leads on to our first case study of of today which is um Rio tinto's

**[19:55]** mind of the future program and uh and really regards to the the recently launched good idea mine uh up in up in the Pulver and Western Australia and what you're seeing is a lot of these big International mining companies you know they're really about finding the sufficiency you know they they want to be paying back dividends um back to to shareholders so it's all about uh finding the efficiency in which you can and how you can extract the product and rehabilitate the land um is uh as cheaply and um and as cost effectively as possible uh really so there's been a very big Drive recently into to to to deploying autonomous drills because and and water cuts and autonomous drills is really about having high Precision GPS's you're removing people out of the drills so you've got listral errors less errors means less wastage um we're then looking at real deploying autonomous Hall tracks and trains

**[20:55]** um so instead of having um people just being truck drivers inside the the inside the put which instinctively is a very dangerous area if you think about if you're working sort of like 400 meters down in a pit you know they have Rock Falls all the time and uh it's it's really been about the sort of how do we um how do we remove people out of the pit but what was quite interesting with especially like sort of in Rio's case um a lot of regional support for autonomous systems was around cost cutting and you're going to actually have it a lot today that the reason why a lot of people say well reason why we want to go autonomous is we want to cut costs and actually what we find quite interestingly especially in Rio sort of case was that was probably the primary driver for for deploying autonomous systems was yes we spent a lot of capital up front now and we and uh and we deploy this ton of systems and we'll save on labor over time but actually what was quite interesting is this taking case in point in autonomous whole trucks if you're driving if you're a truck

**[21:59]** driver on a on a mine driving a truck over time as a human you get to feel their truck you get to understand the vibrations you get to know when things don't feel quite right so even before you know the the sort of smoke coming out of the engines you already know something's not great and you need to take the truck to the workshop but in an autonomous situation you you there is no sensors or initial senses um the truck goes until it breaks down until it's it's done and uh and so what they generally found in in sort of Rio's cases when when they started deploying these autonomous trucks in the beginning they had a lot of breakdowns which were occurring and some which you had to replace all engines or whole whole parts and they started to become a lot more reliant on installing a lot more sensors or vibration sensors and visual sensors it really became you know not just implement seeing an autonomous system but it became implementing more monitoring systems and actually at the end of the day when they actually calculated the number of people that they removed out of

**[23:04]** trucks but versus what they then put into remote operating sensors they actually find out that they're employing five percent more people in managing autonomous systems than they were than they were in in uh truck drivers which they replaced which makes for such a fascinating sort of finding point because here we're thinking that um that autonomy is here's to replace you know everyone's jobs but it actually just uh it really just changed the perspective of where those jobs went uh which became a lot more um really about putting those people into these remote and safer areas and nice air-conditioned offices sitting sitting at a screen and uh and and and driving the trucks remotely rather than on site sensation but some of the thing could be sent with the barracks gold digital transformation this is a a really big Global uh gold mining company as well which has invested in very similar Technologies such as autonomous haulage systems

**[24:07]** um but what they've done is they've taken it one step further and um uh Barry Gold is actually deployed you know it comes in um I sort of call a digital twin but it comes in very many different sort of words and this is all already about well you're collecting all this information you're getting all you have these autonomous systems deployed and I have these Sensors how do we um then decide um how do we deploy the predictive maintenance and get these real-time analytics and it's about how these knowledge systems are all talking to one another and are visualized in a 3D sort of world so so Barrack's really taken that one step further and looked at how we can take efficiency and autonomous systems and make them become a lot more proactive a lot more predictive as well and this also then even gets down to the point about you can't fix it unless you measure it so so we're even getting to the point where we've got safety devices um you know a lot of us are really aware smart watches but on a mining environment

**[25:10]** we're now getting these devices which are helping us to take fatigue uh detect dehydration as well as if you're entering unsafe areas so it's uh we're really um sort of being supported by a lot of these autonomous systems which are all working as one in a single Collective thanks Natasha so what are the benefits of combining the Drone autonomy and AI in in the mining environment well we always talk about the 4ds it's it's really about removing people from the dull the dirty the dangerous and the idea the deer the deer refers to um you know things that just got to be done like you know we've got to go monitoring we've got to someone's got to take a torch or a thermal sensor and got to go out there and monitor solar panels or gonna go walk a stockpile or around a perimeter fence it's things that are important that you need to be done but it's uh but where autonomous systems are really coming in and where Jones play this is be able to

**[26:13]** remove people from these very dangerous situations and things that actually become very repetitive and boring after some time as well so we're seeing this really great benefit of automating all these asset inspections the surveys is monitoring the Securities into without replacing the current workflows just making them a whole lot better and making them a whole lot more more real time thanks Natasha so where do we think the future of Mining and drone autonomy will go combined with AI so as sort of mentioned in the previous slide everything is going to become a lot more real time you know we we human beings might only be doing the monitoring or the test or the survey once once a month or you know maybe every twice a month we're now getting to know where you can do this every day and a lot of customers are going well what's really the benefit of measuring every single day you know we only need our data once a month or maybe twice a month but the same thing could be said about

**[27:13]** the three original days of when mobile networks sort of transition from Edge to to 2G to 3G now we're on 5G you know in those very early days when we all had very basic you know sort of blackberries or Nokia phones Natasha look you look quite young I don't know if you ever had a 2g or 3G sort of phone a Blackberry but um so maybe the Sasha doesn't know what I'm talking about but you know in those sort of days of 2G you know when you heard 3G was coming out you said well all I really do in my blackberries I just send text messages I'm really not you know too worried about um you know any more speed I'm very happy with just getting my text messages but the Advent of having 3G brought on brought on new interesting inventions like the App Store and apps and so having that extra bandwidth or that capacity for information creates more Downstream applications and that's what we're going to start seeing with drone data being available on a daily basis is that its interoperability with other applications and empoweredness of situational awareness

**[28:17]** will only improve safety as well as the information that can be obtained from on-site and how this will all eventually play out is really into these sort of mini sort of swarms or drone fleets that are going to be be able to swap between data delivery as well as then physical delivery in logistics as well within the mining environment because for a lot of you those who've been on a mining environment it's not always the same sexiest place to be and not always fun so if you if you're able to order your favorite fast food or able to order snacks you're the more entertaining you can make to work on a mindset the more attractive it becomes thanks Natasha lost slide them so in conclusion um it's um it's really great to to see how these different ingredients are all coming together to bake the perfect cake uh around to achieve the efficiency that the mining industry is looking for because information is only

**[29:20]** required to become a lot faster to make these decisions um a lot more impactful as well as to improve the overall safety it's also really interesting to note through through to this as well is that autonomy doesn't mean that people are going to necessarily lose all their jobs it just requires that we're all going to become programmers in one way or the other and the same way how we all have to learn office and macros at one stage is now we're just going to learn a totally different tool set and we're all going to be Engineers to to some degree level so I think with that um Natasha doesn't go to the last slide and I'd like to just say thank you very much to um to everyone and the team there as well very happy we'd love to get your feedback if you would like to to scan the QR code there as well please feel free to rate US and contact us as well and we really enjoyed our time with you today and thank you very much for your time thank you Natasha thank you thank you Chris for this insightful session um I really hope that everyone's learned a lot from this

**[30:24]** um so without further Ado let's head over to the Q a section A lot of people have put in their votes and a few questions so hmm I was talking here so you might probably have to if there's the quick Q as um okay there's a question can you hear me Chris put in the chat in the chat for me if you like yep can you see the question in the Q a yep there's a question so what while Natasha is working that it's um it's quite interesting to see the

**[31:25]** results so um people saying what do you think of the most promising applications of autonomous applications Mining and 51 of you voted for increasing productivity and efficiency about 40 of you instead improving safety and reducing accidents and nine percent in decreasing Environmental questions I don't see it in the chat session you want to put in the WhatsApp there for me Natasha let's see

**[32:30]** perfect let's have a look all right please provide a specific challenge within a mining environment that result by using drones and AI specific use cases please they're saying there's actually quite a lot um you can look for in a mining industry perspective I think always the the scenario with Jones is what can't you do with drones right so you're combining you've got this Hardware base layer and so let's say in the net in the nest sort of solution so you've got this autonomous solution that can be operated remotely so you're really about wanting to deploy these Solutions in in situations where you need very frequent um data results right that then brings on the Advent of the AI which just helps you do um detections it helps you sort of take a lot of information and be able to to make decisions a lot quicker so one of the applications which we're using this technology in is is within the security sector so flying the the drones at night along

**[33:35]** perimeter fences essentially replacing security guards detecting Vehicles detecting people uh within within a you know within a mining sort of environment so so really performing security controls we also have other applications where these drones are flying every day and calculating stockpiles and stockpile volumes really you you can you can even program the AI modules as well to um to detect hard hats if you really if safeties are concerned as well so the opportunities are really are are Limitless it really depends on the customer how you want to customize that last uh that last 10 as well foreign do you think multicult is preferable mining Industries and what is the risks with multi-copters um I think it's quite interesting multicopters are definitely the biggest um end of the market it's the largest sector um especially with with the DJI and

**[34:39]** the and you're building very such reliable pieces of equipment multicopters are also a lot more wind resistant so you're finding for environments where you you need a high wind tolerance as well multi-copters can exist but when you're getting too much larger environments again depending on on how soon that you need the data you're never going to beat the physics of a a long-range fixed Wing drone or a veto piece of equipment so it's really about the kind of environment that you're deploying to there's not going to ever be a a one-size-fits-all approach um it's just really about understanding the hardware and that data capture element and understanding the environment and then and then the next steps would be about processing that data so so multicopters probably will work maybe about 60 to 70 percent of situations but there will be some scenarios where you're going to need a longer range pieces of equipment

**[35:40]** from Simon peer 2p um speaking of ESG I'm curious about the uses of drones for monitoring Eco remediation during Mining and post mining periods so you're probably thinking of Rehabilitation there um as well some Pierre so yes drones in the rehabilitation environmental have got some of the most interesting applications especially around autonomy what we're seeing in Australia at the moment is that even before you start mining you you can fly a drone using really high resolution sometimes these 100 megapixel cameras and you can actually use AI to then map the ecology so you can actually then understand the species of plants that really exist even before you start Mining and the reason that this is important is that you're then able to quantify and then once you actually rehabilitating the mind so you put back the topsoil before you start planting the seeds you can actually

**[36:41]** replace the exact ecology that existed before you even started mining as well and what you're seeing is a lot of seeding drones coming to play in this so so you're able to then essentially put the GPS coordinate per seed of where that seed spray should should deploy and so you're returning the environment back to its original state that it was before um before you even started mining which is really great so a lot of advancements coming through I curse I'm based in Adelaide as well been doing inspection work on container cranes in the ports on Australia building digital twins how does autonomy and air work in structures like this that have a lot of intricate pieces so I think what's quite interesting on on the AI component is um you could if you get one well let's start off first if you if you think about a on the

**[37:42]** visual sense so so training AI detectors or models to to find something so you've got to probably start up with about 20 20 30 photos to just train up a basic detector and you say look this is a watch or this is a pen or this is a a crane um you know or bolt you say this is what is normal this is the Baseline because what you then want to do is with their eyes that every time it detects something that's not what you train in the original 20 or 30 photos it's then going to give you this exception report but what you find is that you know you fly a drone from a different angle or you look at a bolt to different a different kind of sunlight all of a sudden it throws on an exception report and the short answer is that a lot of people have built whole businesses just on a single algorithm for looking for a particular asset within within an environment so let's take for example in crane so what you're really thinking about is if you're thinking about sort of ideas um you know the autonomy can work you can put in ring geofences uh

**[38:48]** um and lots of other kind of sensors now to to to be able to operate in these highly complex environments but but implementing their AI component is going to be a little bit tricky because you've got to build them train the model and that can take sometimes between six and eight months so really good I started for whether the big problems the problems that are very frequent and uh and how do you how do you solve those and start off that at a very basic level but uh but very happy to have a cup of coffee with you since you're in Adelaide as well so anytime that you need thanks for your question all right I think we'll finish up with this one as well and the question is what are the solutions to the challenges of implementing drone autonomy and Ai and Mining um well we we sort of uh put it out on the slides there it's you've got to take it step by step because you've you've got to solve them all from being it from regulations to the communications to the integration

**[39:51]** um challenges are there but they're not impossible that's why because I guess we call them Challengers is that it's just we we gotta know it's more than just the Box in front of you it's more than just the software um it's a lot of these Basics that you've got to put into place to to be able to realize a complete solution and answer to a customer and and I guess all my real sort of piece of advice is is to find a particular Niche that can be um you know that can be solved and that needs to be solved on a daily basis um so the original founders of Google Larry Page and Sergey always said they invested in businesses that um they were like toothpaste you know anything you needed to use more than twice a day they felt that there was the kind of the business that you needed to invest in and I think that's where we're going to go with joint autonomy and AI in general is to find these very high frequency use cases that uh the career will power the business

**[40:53]** so I know so Natasha I just want to say thank you for my side you've been great no we can't hear each other but um but I really appreciate all your help today and thank you to everyone um as well for for your time right thank you thank you for this informative session Chris uh I put it in the chat we also have a rocket mind booth in the Expo section so you can all head there and ask your questions the team they'll be more than happy to help you out so thank you thank you so much Chris yeah we should end thank you yeah okay thank you