The Insitu Unmanned Airvehicle System: Then, Now and What's Next

Just another WordPress site

The Insitu Unmanned Airvehicle System: Then, Now and What's Next

all right so let’s resume welcome back hope you enjoyed your lunch and had a chance to network and brainstorm and if you do come up with great ideas let us know we would like to hear about them and fund them so keep those groundbreaking ideas coming so our next speaker is a close colleague of mine dr. Roth right Zach and he’s the chief engineer for guidance navigation and control at in-situ and his current R&D role is to pursue of robust autonomy for institute vehicles his product support roles are focused on integration of autopilot functions within the in-situ unmanned autonomous systems and associate GNC software builds over the past 25 years row of GNC contributions have covered three distinct levels of R&D fundamental research in flight dynamics and control evolutionary development in applied flight dynamics and control and innovation for UAV industry he serves as board member on regional u.s. first robotics organization and supports the local stem initiatives prior to institution at the u-dub and as a commercial pilot he obtained his PhD from Georgia Tech and MS from university of delft in netherlands without further adue growth it’s a pleasure to have you here yes I I worked at in Sydney or within city rather since 2001 when when I was still a faculty at new dub and in City was still small enough to qualify for that s in SBIR and working working with a university then would provide in city with with research funding and we also worked with the Washington Technology Center here which at the time was just down the street here I’m a little disoriented but somewhere on this side of the building so let me I’m gonna tell you a little bit about about in situ and of course when I when I put this together I have to admit I didn’t quite realize that its founder dr. Ted my gear was going to be in the audience so you’re welcome to ask me questions either after or during a meeting but but no no Ted okay I didn’t quite maybe do all my fact-checking quite correctly here is was actually a brief note from Ted that was really the original idea for a small error sound right in the idea behind this was that look that the rest of the United States is kind of benefiting from the West Coast releasing weather balloons providing in-situ weather data but whereas we on the west coast we kind of like get only very few little bits and pieces of data in terms of weather that is mainly from satellites or from ships taking those weather balloons out in the Pacific so the idea of having a using GPS at the time to have an arrow songs that could go out over the Pacific let’s say a full day and then be retrieved with that sort of data was a very very interesting idea and it started with a proverbial a garage gosh sorry a start-up right and and there in the middle is Ted and two of his early colleagues and they’re starting to work on this airplane of course a garage shop in in the Silicon Valley that that is about the price of a small luxury apartment in the Columbia Gorge so one of Ted’s friends from Stanford from Stanford days Andy from float Oh convinced him I believe to come to this area and in the background you’ll see a me right here in the background is one of many barns that he had in the middle of this orchard where now to this day we build equipment for in-situ and then Andy and tare together eventually convinced Steve’s Levi yet another Stanford mob member to do the business end of things and you’ll see I’ll show you in a minute how in city of course has come about it’s very impressive and all these all these three folks are very unique impressive individuals to role models in in the stem field here’s a little bit of what happened with the air assault yes you may recognize the u-dub wind tunnel over there on the left the arizona was nicely scaled to fit right into that winter and all that was nice with very little wall effect right that makes a

huge difference because that allows you to get over a lot of data without having to put a lot of money into a scaled model now a lot of work probably can attest to most of the work for over those over those a year since that idea written down on that note went into finding the capital and the backing and getting this possibility communicated and ted then worked together with u-dub and pulled off a great stunt in 1998 in August he had essentially one of the Arizonans make it across the Atlantic autonomously and at that point a lot of a lot of exposure was brought to to the ability and and it’s I think created right like like little over seventy years after Charles Lindbergh did the same thing and created some change how people can conceptualize what these aircraft can do because that’s the hardest I think to communicate I put a little footnote in there because the irony is is that in the meantime Eris and idea that the rights to the Arizona were sold initially to a company in Australian and came AAI and now it’s text ROM and it’s one of our biggest competitors now here’s a little bit at the background of in situ right in so in situ I’ll show you in a little bit what what went into some of the work of making that system as small as possible and initially Ted was interested in both you know getting the attention from Noah also getting attention from the tuna fishing industry right that like relatively small ships in other words but of course the US military got wind of it and Steve sliwa being Steve sleeper I saw a business opportunity and you know the whole thing kind of exploded and that by the time we find ourselves today we have 800 employees in fact until about a week ago I could say we had 800 employees and we just bought a small company on the data processing and so now we’re 850 employees and you see that we’re located mostly in the gorge but our main our headquarters is really centered in Hood River and Benjamin white salmon Washington so a little bit about right drones UAV small autonomous aircraft if you’re not familiar with this industry you get exposed to it maybe in the news and typically the media has some very interesting representations of what what these drones do we don’t even use that word drones UAVs right stands for we use UAV and UAS unmanned autonomous systems or unmanned autonomous aircraft and there’s there’s now a large variety of them you can all go on Amazon now and buy them commercially with all the consequences that go with that of course but there’s many different tiers of these kind of aircraft and you’ll see in the bottom left that’s essentially where in sitting fits in I use the word tier and that there is not necessarily a common definition but in terms of size of aircraft you see that it’s still you can like one or two persons can carry the aircraft and also we do not have a landing gear I’ll show you what our landing system is here in a little bit and we launch not from a from a runway but we launch from a catapult as you can tell and I’ll have a demo of that as well at the moment where we’re kind of poised as a whole industry for the civilian airspace to open up and that is a huge huge market potential of course we have to deal with lots of regulation the FAA is involved the FAA is not known for moving fast so the industry we have talked to many different industries some of these industries have lots and lots of money behind it and they would love to have you a these do this for them the UAV is more or less commercially viable and even though that is still a challenge I have to have to say and I can put that in perspective here in a little bit here are some of the applications that the UAVs are useful for specifically a small UAV system typically is still too expensive but except for the dull dirty or dangerous type of jobs and one of those is is the tuna fishing industry and after the 1998 crossing of the Atlantic Ted’s had gotten the the attention from this industry and they approached him with this particular form what you see here is a tuna fishing boat that can

like this happens to be a very very nice day clearly not all days are like that out in the Pacific or the Atlantic and what you see on top of the roof there is a small helicopter and that helicopter basically goes out the spot tuna they can they can like they either have either that the pilot is trained to be a spotter or it’s the pilot plus a spoiler and these if you do youtube on this it’s like helicopters and tuna fishing you’ll see very quickly how dangerous that is so having a small solution would go a long way and that’s where the initially that was called see scan that is now that what is now known as is gay and eagle in in in 2003-2004 Ted started working on off a small signer Shackleton that’s right here in Admiralty Bay and that was a I spent some interesting sunbird times on that boat too and and so that there is and and this is a really nice problem I’ve presented there’s a us first for the kids with the robotics because they have to solve very strange problems with those robots and I said well how about catching an airplane that flies 60 knots and catch it on a very small boat right so this this always makes for wonderful material so let me talk to you a little bit more about the details of our system and I’ll have some videos which which will give you very quickly and a good idea of how that works so in in the in the system you see there on the left side is the the pneumatic catapult I’ll say a little bit more about that and then the claim to fame of the small so-called small footprint relatively small is the skyhook system it’s essentially a glorified Genie that there’s a vertical line suspended from it and we we Lance the airplane by flying the the wings that have hooks at the end into this is a vertical rope and then I’ll show you the ground control system of course that is behind the scenes how we operate the airplane now to put it in perspective some of the numbers here you’ll see that this this airplane is about 40 pounds and about seven and a half of that is payload right so we can only carry a very small payload that that I see as a common misconception that you know like all the things you could carry with this airplane is actually very little of course relatively speaking is the same for most aircraft but definitely size weight and power on these aircrafts are very important this is one of our other varieties of that of that ScanEagle right the payload is the than in the nose is the money-making end of these aircraft that’s the only thing that the customer really cares about in fact they don’t even care about the fact that it’s an airplane they are just interested in the fact that there’s a camera and a nose here’s the slightly bigger airplane that this the newer aircraft which is not a program of record for the government if the ScanEagle is like a sports car this is a bit more like a pickup truck so it’s designed to carry a little bit more but the the here the the payload to the airplane weight is about 1/6 so it’s a it’s a bit better and here’s an idea of the kind of a kind of video that’s a quality that we that we provide now I’ll I’ll give you this this is sort of modern quality and as you see it’s it’s only this is a typical operation for us where we’re about 3,000 feet out and 3,000 feet up so you could think of it as about a let’s say 4,100 feet slant range and it’s almost the point where you could recognize somebody which of course that military likes of course has also all kinds of implications for for civilian applications here is a little bit of an idea of the ground system right very significantly we have to always communicate with these aircraft so the antenna there that is that is always aiming that’s the aircraft in order to get a large range that is our communication link and that is also how we receive video down and you can you can tell what is going this is also a little bit of the weak link for these systems and that is why the FAA is so concerned because it’s fine if we put these aircraft in airspace out there but what if that linked somehow breaks right because somebody else turned on a big radio which if you’ve ever looked at a picture of a military night like a Navy ship there’s so much stuff on the top of those that sends all kinds of stuff out so that communication

link can get broken relatively easily and so that’s pretty important of course the we also are pretty intensive with the operator like the operator is flying both the aircraft and I have to I’ll explain that in a qualified at in a second but also the the operator is concentrated on the payload when I say fly the aircraft we call it operator not pilots because we we try as much as possible to make it hands-off so flying the aircraft means you provide waypoints you tell it where you want it to go and you can upload that to the aircraft fee that communication link and at that point the airplane flies its mission and now the operator is free to concentrate on the payload itself and then when there’s something of interest either maybe you’re mapping a certain terrain or maybe you’re observing an object and you can orbit about that object that’s all that the operator does with the aircraft so let’s see here is an idea of the launcher so the launcher is a nom attic catapult oh let me let me back up how did I let me if I if I don’t aim just right on that video starting button here then so as you can see we get we have to put a lot of energy into that airplane in a very short distance and time and so the the typical blue forces are 20 to 25 GS and here’s a little bit of that evolution all of the launcher so you see here’s the early days in fact that is tad in his car with one of the C scans on on top I think that in the meantime his teenage daughter may have inherited that car so it doesn’t quite look like this anymore then over time we there’s a lot of lot of energy in this system so the safety of course is is fairly a big concern the amount of energy that that is contained in that clearly gotta carefully be carefully controlled and designed this was a hood tech the Andy from flota was behind most of this design and what you see here also the shape of that of the bar is very specific that is to give us as much of a flat constant acceleration as possible in order to not prevent it like that those peak accelerations because clearly anything you put on that airplane you expose it to more than 20 G’s it’s gonna be have a hard time holding on to that airplane now here is perhaps the more most interesting part of the whole system this is the landing and if if you’re a pilot let’s see if you’re a pilot this is always cringe-worthy but here’s here’s how we land the aircraft and so it’s very abrupt especially on the on the heavier airplane let’s see this is another and there this is the new airplane now and if you’re a mechanical engineer and you’re looking at that Gina you go immediately yes that genius helps really the ideal solution for this sort of thing those genies are made right to take a force up and down not sideways but at the same time this is also the one of the major distinguishing features of our company versus other UAV companies now here is an early idea I’m gonna touch on this a little bit later of how to retrieve the aircraft here’s a little bit of how that that genie idea evolved over time and you see lots of different flavors and varieties and so now in the end we end up with this and I use this for mechanical engineering students sometime to say well okay what would you do to solve this let’s see a few unique design challenges for our system well I just highlighted this this combination and we didn’t even use these tools initially initially it was all you know use sort of like envelopes and simple simulations to compute what’s happening but we have a unique trade-off right if we want to land this aircraft with the leading edge

and then on its wing tip that means now any any kind of like accommodation we make for this kind of landing on the aircraft is gonna have to be carried with it for the entire eight or ten our mission so ideally you want that the solution space to be and in that skyhook of course as soon as we deal with the with the Navy the Navy says no you that skyhook must be portable by two people and must fit in this kind of a box and then pedigree etc etc so pretty soon we’re making trade-offs in so many different directions that it gets complex really fast very interesting design problem one of the things that I really love about that skyhook this was way before my time but it’s of course the Columbia Gorge so there’s lots of wind serving the one of the guys on the boat right there is in fact somebody who invented some of the first kite board and initially the idea was look we’re in the gorse it’s always windy how about we use a parachute and we tow a line and I think there was actually like one line for towing and then another line maybe two to stabilize it in the water behind the boat and then we fly the airplane into that line and then retrieve it roll it back into the boat that’s how things got started and it’s because you in the gorge you have so many creative people that that are used to that when serving industry which is essentially saying they’re a bunch of addicts addicts like they’re so into it like they can’t do anything else and also the carbon fibre layup right that that they have a skill there that where they were able to make these early early equipment in the early aircraft with all those skills so that was really nice that’s how that came about of course didn’t always work and then you just get one of your youngest engineers to volunteer to go and the Columbia is really really cold in the meantime it was very successful we were on many many many different ships of course my favorite is still the Shackleton and I wanted to share this this video with you in fact it’s a video that put together and like about 10 years ago or so and let me see and I love this video because now it’s on my background as guidance navigation in control and as you see we just launched off a ship and here’s a picture from the airplane or taking an airplane ticket and I look how that ship is rocking I’m rolling and we’re gonna land in that line on the left side there you see that the note at the bottom that the crew is out on the back deck getting some air because they’re not feeling so well but now comes a very interesting part right we have a GPS on top of that mast and there’s a GPS on the airplane and look how it’s to get to and now it just flew in the between the line and the mast like here from the airplane it looks even better look at that and notice nobody’s wearing any safety equipment either it’s like if we do that now with Boeing owning us here and it happens again and now look in the background where you see there’s another fishing vessel and we just barely missed that one – it was very interesting and so this actually happens a few more times and it’s just like the odds of that happening like that safely it’s just astonishing and then so eventually they just essentially just sort of fooled the system into thinking that it was a little further outwards than the actual configuration and sure enough and then this time it worked well there we go so and that was in 2004 and I want to just put that in perspective with all these Navy ships and US Coast Guard ships on which we are deployed now in the background there in the orange is Simon Hales he’s our now the leader of our same team still with the company Alex Alex PETA was one of the Argentinian fishermen that approached at originally and their stead doing a hard work himself and then there’s Guillermo right there on the on the very right side also an old Argentinean fisherman let’s see so anyway I just love that video so much that even after all those years that makes a lot of it clarifies so much of the challenges that we have to solve so let’s see mehran you made that introduction of me so long that I’m running a little short on time here so I want to highlight

something about simulation right is very important for our for our work to have simulation and the irony is not guys dealt with the Navy now meaning the government like government engineers looking over my shoulder as we put the GNC on the aircraft and put the whole system together and so many many discussions and meetings I’ve met with them and in the meantime we we created the simulator with them to evaluate how the system works and of course it takes on a life of its own you get these engineers that are focused on a particular system or a particular subject matter expert expertise and in the meantime we have created this string of lies right because everything here is simulated like the atmosphere simulated the airplane is simulated it’s the ship motion and and also the air wake over the ship is simulated we simulate DPS signals and all their characteristics and then we put it all together but of course we make approximations and all of that and then we close that whole loop and then we start flying at a shed like we do it you know let’s say twenty thousand times and then they make conclusion so that looks safe and they sign off on it and I just keep my mouth shut because I go okay good next but the truth is it’s it’s really a really interesting challenge to come up with this sort of system and simulated and make a good assessment of how well does it work really what we can say with a good simulation is we can at least address sensitivities where the modeling is accurate or not clearly it’s a small aircraft there’s no people onboard this aircraft so we can’t really justify the full-blown commercial aviation type of simulation and there’s no need for that let me I’ll skip through this because it gets a little complex but the little more than I want to that I have time to discuss but essentially this is the end result that’s that the the the Navy sort of checkboxes or concerned with they want to see a simulation they want to see sort of what happens if in all these different conditions different sea states different winds different kind of airplanes can they make it to that same skyhook does it land safely is what they’re interested in so like I said like each and every one of those dots has so many approximations in it that who knows I wanna I wanna show you a little bit about where we’re headed and what the challenge is still or and there are still many really interesting things to be resolved one of the trends of course that that’s everybody gets excited about well everybody at their own bein set aside including some of the executives and I always have to sort of be the voice of reason because here here is a block diagram represents the aircraft avionics part of that avionics is what we call the auto pilots traditionally so there’s sensors in there then then you you put a state estimator just our data fusion element in there and then we have of course the control loss that stabilize the aircraft and now of course you can you can like people come out with their telephone like an Android telephone where they can navigate with just using whatever cameras on that on that phone and the few sensors the the accelerometers and gyros in the phone and of course those phones and billions of those are made right the auto pilots we buy or you know more than ten thousand bucks apiece and only few of those are made relatively few so wouldn’t it be nice if we could indeed use the kind of material that goes into a phone and in essence that was really what inspired Ted’s original idea by using GPS and a very simple sensor to make this capability of an air assault of course the counterpart of that is those phones are definitely not providing the same kind of performance that in the meantime our customers pretty sophisticated customers with pretty sophisticated payloads come to expect so that is often an explanation I have to make and now similarly with the image processing in the loop right if you’re a controls person you saw at any time you’ve you closed a feedback loop you start asking these questions like hey what is the effect of closing this feedback loop well the problem is with image processing image processing does amazing of things and you can get open-source software online really really cool lots of lots of work has gone into that and and those tools being so openly available is very nice makes for a very creative work but when you put that in a loop those image processes are still very efficient but not robust right as soon as the angle of the Sun is different or as soon as it’s a little bit hazy the like you can’t rely on it anymore to put that inside a closed loop

for an airplane or you can turn it around you say hey are we gonna rely on this for that airplane to see other aircraft and avoid other aircraft well you know if the Sun is just right maybe it’ll for you but will it also do that if the Sun is somewhere else who knows and so that question is a really significant question and that’s gonna open up a lot of interesting academic work work but my point is mostly for our exact is it’s still academic work and in order for us to bring that to the customer now let me let me and and with this note in terms of autonomy only recently have people started formalizing all these different levels of autonomy and autonomy is a very sexy word and again in the guidance navigation and control world and that’s what we’re after we even refer to it as at tournaments and if you go to the very top level of this of this diagram you you see those statements like negotiating decision making situational awareness that’s a big one understanding of complex and in other words the top level right there is that the level that that’s Stephen Hawking and Elon Musk or warning us about that by the time we’re there the UAVs will take over right well the reality is that red line is where we are that is between the level 0 and level 1 level 0 being remotely controlled aircraft and level 1 is essentially where we operate automatic flight control in other words we’re not really autonomous we’re automatic and we’ve got a long way to go so that should make a lot of students very excited let’s see here I want to close with this remark about the engines the engines is still our biggest challenge the reliability of the engines is really an issue because the engines are so small so we’re really pushing the envelope on that technology right we are our engines are the size of the stuff you find on your lawnmower or moped and here are some of the some of the early pictures like like that one in the bottom left that was our sense of safety was to put a ladder on on both sides of the propeller here I’m not allowed to give you numbers so I thought I’ll do it another way to give you an idea of what the end in the early engines on our ScanEagle were about the price of a Toyota Yaris and then we’ve got more sophisticated because the Navy wanted heavy fuel heavy fuel meaning like diesel type of fuels and debt jet fuels even though we still use a spark plug they want to be able to run those so then the price went up to like a modern-day folks wagon Passat and then the new aircraft actually went up a little more and and this is for one engine so like you can get maybe if you’re lucky you can get about a few hundred hours out of an engine like that and so in other words we have still a long way to go in terms of engine technology to make this affordable for let’s say a farmer to be able to use a system like this it’s still much cheaper to go to your local airport and rent somebody in a Cessna of course in the meantime maybe you’ve seen one of the posters our chief engineer for advanced products d/f Knapp is working with WSU Jake reitman and they’re doing some interesting work that I’m not allowed to say too much about either but well we’ll see where where where this could possibly go but the idea is that you know maybe maybe electric propulsion is something that could solve some of our problems I believe that the evolution of the engine technology as we’ve been working on we’ll be around for a while but this is interesting development that happens in Washington with again under a gigantic funded program and with that I see Mehran is circling around to snatch me off this stage I’ll conclude this presentation we have time for a few questions