SPAI: The Future of Drone Technology - A Conversation with Sparc AI’s CEO, Anoosh Manzoori
[Music] hey everybody welcome back to equity Guru this is Chris Parry your founder president uh Chief bottle washer uh wrestling promoter uh and commentator with today we have an manouri all the way from Melbourne Australia's second best city uh talking about spark AI SPC do a well not AI but AI um the ticker symbol is Spa I and um an you're in the the AI space when people hear this do they automatically like give you the fared eyebrow and and wonder like what your scam is well you know I think um I think like with any new industry that that's emerging there's always uh you know actors that um you know not necessarily playing playing well but um look we've got a technology that actually has been in development for about 12 years um well before the term AI was even invented um we've got a series of patents around the world and um you know we've now in a position where we've got something that can actually be put into a into a drone and actually taken to Market and commercialized so we're pretty excited with where we are at the moment yeah my first look at your technology had all my alarm Bells around uh autonomous driving going off about how it might be useful considering there's been so much uh AI driving clutter about liar and radar and Laser and GPS and uh but in talking to you you're more focused on uh the Drone side of things what do you see as being the primary uh initial at least use case for your technology yeah so the initial use case really is around this concept of Target acquisition system um so you know if you're if you're standing out in the field somewhere and looking at something that say a kilometer away um there's no real way that you can use GPS to work out the location of that object um so whether you've got GPS or where you're sing is actually irrelevant you might have GPS and that's fine but you can't you can't work out the location of something that's that's you know a kilometer away from you um so the technologies that exist today to solve that problem um are things like liar radar or laser finder um some people try to use image recognition if they can recognize that image um but it's it's you know it's quite a challenge all these technologies have their have their limitations um so the defense have a you know range of products they use and they they called um Target acquisition systems and they combine these three or four Technologies together on the one unit so they typically sit on a tripod they're quite expensive they're quite heavy um there's one group for example that actually has a 50 m cord attached to it so the operator could be away from the unit because obviously you're looking at the enemy but the enemy could be looking at you these things aren't covert um and so you know they're they're expensive things right um and then if you want to take this technology and put it on a drone um well first of all it's too heavy um it's very expensive um it's just not going to work on a on a line um you know a few thousand drone it might work on a a drone the size of a small plane that cost tens of millions of dollars um but you can see there are some limitations in in in terms of trying to to do this so what spark does is is effectively a software based solution um so it it it's got a series of algorithms that can determine the location of any distant object without any any supporting Hardware um the only we actually use the camera that's really for our benefit as a user so we can see what we're looking at in fact I did a a demonstration where I blindfolded a camera and we can still determine the location of where we we're pointing at um so there's a series of algorithms we've got it we've got a patent around it um we've developed a whole Suite of software around this technology that obviously we haven't disclosed you know as part of a patent it's this proprietary that allows us to basically Target acquisition an object that that could be tens of kilometers away from us and then we've extended this further by actually saying well if we can determine location of objects in front of us we can actually build an autonomous um vehicle whereby we can navigate autonomously without GPS whilst using these algorithms to to help us navigate that's a lot of tech that is just wiped off the board um because legitimately right now in the in the battlefield you know you've got drones that have Laser Technology to acquire a Target at the moment you're tossing a l laser out on the field you're Spott right if you're sending radar signals out you're Spott and you know there's it's it's a very obvious need for stealth being you know the presence of stealth drones um you can't just be sending pings out all over the battlefield that an enemy can glom onto and wipe out technology so one question I do have for you though is if you're covering the C the camera how exactly is the system registering anything yeah so we we know the terrain Chris of of obviously planet Earth right um and so we can we can take the terrain and actually put it onto onto the device on the camera or on the Drone directly um and then we can use a set of algorithms based on the the tilt if you like the XYZ tilt of the camera in terms of where it's pointing at um that intersect of where it's looking at where it hits the ground that Delta between the height and the and the height of the location um we can work out the location of of effectively what we're pointing at um so as long as we've got the terrain um which we do um we can we can work out the location where we're pointing at that's crazy so hesitate to call it a a manual Sol an analog solution but uh considering the amount of high-tech expensive gear that this undoes the necessity of uh I I I hesitate to find another another phrase that would better describe it yeah we're sort of displacing um some of the incumbents right um and we've we're relying on software and algorithm um to actually make those calculations because at the end of the day from where you're standing looking at an object a kilometer away it's just mathematics right um so we've got the mathematics actually be able to calculate that that that that distance between the two objects right so the tough bit on anything to do with military or government is obviously getting into the system uh getting you know past all Security checks getting things tested out uh having them figure out first of all they want to do business with you then and that they can do business have you at this point started that process do you have a weigh in yeah we have we have started that process um I think we're probably about maybe two to three months away from being at a point where we've got a really full commercial product to bring to Market we're currently working on version 2.0 we also have got this crazy drone that we're building that can basically autonomously fly by itself without any any sort of connect to to a radio or to base um so I think within next sort of two or 3 months we'll be in a position where we've got a product that's that's commercially ready that's been extensively tested um typically to go to market um we have to go through a distribution channel so commercially one opportunity for us is to have this available embedded into drone operating software or drone flight management software where it just becomes part of the Drone and then you can activate this you know on a SAS model whether you're trying to determ a location of you know something to do in in the crop for farming or for sign inspections for commercial or property um that market is I guess a much lower hanging fruit for us to to get to revenue relatively quickly because there's a very clear distribution path there for us to integrate into drones into cdks into you know other Distributors um so par is a good example we we we chose the parro Drone um which was is a drone designed for the US military it's used by First Responders around the world and they've got an indirect sales model so they don't sell directly to the military they don't sell directly to you know to to the police force they go through these authorized resellers which is about a hundred of them around the world there's only one in Australia for example um and so you can partner up with these sorts of companies they not only provide the Drone provide the flight training but they also bring in technology companies that value to the Drone to solve their particular use cases so again for the military we've got a fairly clear distribution path um to get ourselves in front of uh the right decision makers you know for us to go and do it ourselves um we we'll take us years we may or may not be successful but if we go through those companies that are already supplying products and services to the military it's going to be a lot easier for us to to be part of that mix no doubt it's a it's a a good quick way in uh to not have to go through that ordeal uh the parrot obviously is is a a drone that's been used for a long time and for not just military but for First Responders um I think you see this as being something really valuable to First Responders in in the short term is is that correct yeah absolutely I mean you know it's something that um obviously First Responders there's there's situational awareness and surveillance there's also the need to capture evidence right so again being able to Target acquisition record the location of events happening um at a very precise level is going to be very useful also with the ability to have autonomous flight so for example we've already got the ability now that we've demonstrated where we can Target something say a kilometer away and then the Drone can automatically fly to that location and investigate and then fly back um this is done in a completely GPS denied sort of environment um so I think there's there's a lot of use cases that we can sort of fit into with this sort of technology for sure so geolocation is is an increasing issue with drones right you where you can fly is as important as being able to fly uh does your software take that into consideration where there's no fly zones or height restrictions or Municipal rules for that matter yeah so it doesn't it doesn't do that on its own um but having said that if you take a drone in a particular environment that um you know there's restrictions in terms of its height or in terms of ability to fly particular um you know zones um we can certainly do that we can even create a an orchestrated um flight path purely using the spark coordinates as opposed to using traditional GPS um coordinates as well um so we can you know create a bespoke use case if you like for for every uh every type of client if the if the object moves does spark recogn nize that and and consider it no it doesn't it doesn't recognize it um it doesn't it doesn't follow it um but it's a good question because there's no there's nothing stopping us from using spark with other Technologies right so there is technology already available that that can follow um a particular object um we can certainly combine that using using spark so it we touched on earlier on the the general AI industry it seems to have blown up really large and of course come back to earth a little bit uh once the bubbles deflated do you find people are are looking for good AI ideas or is is the AI terminology a little bit soiled right now yeah I think um you know from where we're sitting people people are looking for AR solutions that are really solving real world problems um you know it's very easy I think to get caught up in the hype of ai ai is not going to solve everything but I think if you actually have real AI in terms of real technology not just processing third party data and solving a real problem I think you you're got to be around for a long time it's it's kind of reminds me of blockchain you know where a whole bunch of companies were just saying we're a blockchain company and it was like yeah but blockchain and what like what are you doing with blockchain it's one thing to be an AI company but what is it that you're actually doing with AI that makes sense in the real world and I think quite frankly with the way the world is right now with the number of conflicts that are out there but even outside of conflicts uh just search and rescue or being able to to spot a bad guy hiding in the bushes um there's a lot of use cases for what you're developing here and billions of dollars being spent on the highest of high-tech solutions that may be sort of overthinking the problem yeah I mean I think you can you know if you look at the defense spending there's there's a there's a real shift towards high-tech Investments um as opposed to traditional you know artillery and weapons um there's also you know a lot of a lot of focus actually on on emerging Technologies um so we're seeing a lot more startups that are starting to develop you know technologies that have been taken up by the military as well at a much faster Pace than than than previously so I think it's a very interesting time to to be in this space so the big question I guess for for retail investors who might be new to the story is um spark nice Tech good ideas got patents that's great and might be really important when when uh the tech is proven out but how long will it take to get you to a place where it's monetizable where where things appr proven to the point where potential customers are coming to you yeah I think we're we're at a point is where you know definitely the next sort of you know two to three months we'll be at a point where we've got a commercial viable product um you know we had to prove that this technology you know works at a very simple level on a on a mobile phone um so you know we did some some you know some crazy tests you know everything from driving at 100ks an hour and testing it taking on top of buildings um doing it nighttime um so we really tried to to break the tech technology at a very simple level and then we've managed to get it onto three different microchips um so we can now got a portable Hardware device that we could integrate and sell directly to to customers and now we've also successfully installed it on a on a military grade drone um that that works perfectly well as well so we're at a point now where we're just refining it we're really doing a lot of situation type testing to work out what the scope of this technology is and we're also working on version to which will have a lot more capability and features uh for this product so we're getting to a point where we'll be able to go to market and sign up Distributors and then hopefully get ourselves in front of customers we also launched recently a service offering where we can actually provide some services around integration work or Consulting work if a particular customer has a particular use case that wants to say combine it with you know Vision recognition or some other sort of technology that we can then provide that add-on service for those customers uh the the the monetizing plan uh is it a a oneoff sale or is it an ongoing contract that you're looking at down the road yeah we hav't finalized of pricing um but you know if we were to do some sort of peer-to-peer sort of um comparables of um sort of add-on software that that's sort of Enterprise level or military grade that's sold onto um existing drones you know it's anywhere between sort of2 and a half to $5,000 a year per drone as a as a subscription um so that's typically what we're looking at also depends on the the complexity and the the cost of the Drone um obviously it's the if it's if Ukraine for example wants you know 10,000 of these on low low grade you know cheap drones it'll be a different price per drone versus you know a much more high-end um drone that we sell to the United States but all we sold is a description model per per drone per anim do we have your commitment that good guys only yes absolutely good good good that's always a big question for me when things touch on Military it's like if the US says no we don't go to Vlad do we no absolutely we're we're a Canadian listed company so only good guys only good guys all right man well look I think that uh it's really interesting looking at your share chart right now because it is a story that that has been under told so far you've been in development you've been in in uh developing and testing technology rather than the the growth phase um but your share charts held up I I know you've been uh buying in on this for for not months but years when did you first uh find the project and uh and get Dragged In yeah so it would have been probably probably close to 5 years ago okay um so I was introduced to sumon l he was the the founder of of the company he was a mathematician um by trade and he spent at that time about five six years um developing the technology um received a lot of government support as well um and funding and um so I was I was really the first investor um that that invested in the company and um you know there's been subsequent rounds of funding and I've I've participated in every single round and I I joined the board I think it's just over 18 months ago uh where Simon was sort of struggling with the capital markets um he's not really probably the um the best guy to to front a commercial strategy although he did sell it to Telstra which is Australia's largest Telco they were using for bushire detection on fixed cameras um he sold it to Big Rose which is uh the largest um uh sort of uh manager of our um a Freeway highway system in Victoria so these were sitting on Emergency Management vehicles um so the technology certainly worked on a fixed camera um uh type model but the way he structured the the software he had to basically go into every customer and basically do a whole development work every single time the customer wanted to use it so it it's a very clunky way to to commercialize very expensive um and it was you know just quite limited into has a scalability so you couldn't scale it um so when we stepped in we basically rewrote the whole software from scratch our first sort of task was to put on our microchip so we can then distribute the chip have that easily integrated and give it scale and then we've got it to a point now where we've got a software kit that we can easily install onto any device um and we've also managed to actually get it onto a drone which has some additional complexity you know with the Drone moving with shift with wind curvature of the earth there's a whole range of additional physics problems that we have to solve um in doing that got it um on the the business model side of things uh you know 2 to 5,000 bucks per per drone as a license how much of that is is margin uh be probably about 95% plus margin we will wholesale that out to um the stor distributor will obviously add their margin they'll add other things you know to it and wrap it up as part of their offering but um for us it's you know it's no 95% margin so once the dev work is done and the sales work is done it's money for Jam yeah it's it's a big investment up front um but we'll definitely recruit that once we get into the subscription of this this this software is there anything that you worry about that could get in the way of this software getting picked up or have you der red it to the point where you're happy um we're definitely over the hump in terms of the technical risk um so you know I did have I did have a number of shareholders call me and say you know you you plan to put this on a $25,000 drone um obviously you put it on a cheap drone first haven't you um we hadn't um but you know we tested it um so extensively on on a phone on a fixed camera plus we knew the history of the company um we were pretty confident the mathematics would work on a on a drone so we're at a point where we we're pretty comfortable where the technology is um we don't have any sort of technical risk our risk really now comes down to uh finding Distributors and getting this in front of customers I guess the good thing is mathematics rarely change over time that's true yes um if uh if there was one extra thing that you'd like investors to know about spark AI is there anything left that we have in mentioned uh well look you know from a management point of view we're we're we're very committed um the Insiders own you know just under 40% of this company um I've been bankrolling this company since I I took over about 18 months ago um we haven't done a you know any sort of large financing um since then and we'll continue to support um this company financially until we get to a point of commercialization well I see through your news releases that it's a steady drum beat of testing and and proving out the the technology so uh congrats on on what you've achieved so far is a lot of startups out there that are in the kind of space that you're in that you know raise hundreds of millions billions to achieve less yeah well we've uh been pretty fish in the way we've uh managed the uh managed the company we' all got a lot of experience in terms of developing Tech and bringing Tech to commercialization and and exit so um not our first radio but we're we're definitely committed to to the success of the business so everybody get out there and get this on your watch list spai is the ticket symbol it's on the csse spark AI is the name of the company and Spark ai. is the website and N manuri thank you very much for for chatting with us and look forward to uh to hearing more about the company in the weeks and months ahead thanks Bruce thanks for your time
2024-09-18 15:59