[00:00:05]
>> Well how did you 1st you know that us. Here. He will play. Well and serve right. Laughs To be a triple check back with masters and Ph D.'s for tough. Rhetoric the new center for the region want their mobility I've got a right to them You did their brand no one will probably know as for tax cuts morning this is where you want to this is Larry Centers we're starting to see more and more overlap and opportunities for engagement where we're very mobility people the only energy all intersect at.

[00:00:48]
The previous session they can issue will follow the whole hour rather than actually let you know whether you'll take questions throughout the Old and right and we leave all the time until 1 30 pm and different people do it differently you know you want to start you know 1015 minutes before the end and then to questions if you did that the incident where they are available afterwards if you want to talk to the thing itself up you know you get one.

[00:01:24]
Thank you Jim. And so I'm really excited to be here get to know him and supply chain the logistics Institute just a few months ago in the process of building this center for urban regional air mobility I'll tell you a little bit about that at the end of the talk but I want to 1st bring you up to speed on why this is exciting at least in my opinion right now.

[00:01:47]
Short story we're sort of poised in the aerospace community to have a transformation happen and it has to do with reek Eric arising what it means to fly simple as that and you know we think of going to the airport and flying 500 miles right what if you took an aircraft and flew 25 miles.

[00:02:10]
You know that's that's sort of interesting flying 25 miles instead of 500 and what would that look like and why and you can just imagine that's entirely new market for aviation for people and for cargo and so. You'll recognize that there's a lot of o.r. and logistics problems in this space as we start to redefine aviation So that's the overall context and I'm going to talk a little bit about why this is happening from an aircraft technology standpoint too and one of the key reasons is electrification of transportation it's happening in the automobile and it's now happening in the aircraft so that's why we're at the point we are so let's start with a sort of brief history of electric flight then we know when the really the 1st electric airplane flew it's not on this chart is earlier than this you know it was in the seventy's someone did it is kind of a stunt it was a lead acid battery it was like a washing machine motor right and they made something fly and it listed it here I need to update my slide that that that happened but really the history of modern electric fly again because of these things right again because of what was happening in consumer electronics so you look at the Apple Newton there from 1993 when we were looking at night had batteries replacing lead acid and ultimately nickel metal hydride batteries came along this was pretty lithium ion right the things that we have in our lithium ion so we had to these P.T.A.'s which were really pushing the technology for high specific energy to give enough talk time on your device.

[00:04:01]
Now people been flying little radio controlled airplanes for many years and it's some point that battery technology combined with motor technology and what was going on in controls engineering the ability to make a brushless d.c. motor very reliable very effective those 2 technologies intersected and suddenly all the hobbyists are flying around these little fixed wing airplanes now you guys probably know that that has extended to go beyond a little fixed wing airplanes into the quad copters you might own one you might fly it around campus so all this is happening really starting heavily in early 2000 military picked it up and then came 2007 the Apple i Phone Apple put a big investment in lithium ion technology and really we've been we've been moving forward in that direction 2008 Tesla Tesla said hey we can make an all battery electric car right not like the Prius which was hybrid electric and had an engine and a small battery system that was nickel metal hydride but rather we could have a fully electric car with a battery system this lithium ion that was a big deal 2011 NASA sponsored something called the Green Flight Challenge which was basically to ask how how much efficiency equivalent efficiency could you get out of an airplane so a company called Pipa stroll in Slovenia modified one of their gliders with an electric propulsion system and blew away the competition with some amazing numbers like 400 miles per gallon equivalent right which is pretty pretty striking.

[00:05:48]
The drones I mentioned you see 2013 we really started seeing quad copters everywhere 2015 that same company that made the airplane the glider airplane was position to actually market a product so they built a flight trainer that they called the put this to Alpha electrode you can buy one of these now you can fly it in the European Union and now very recently we're getting authorization to fly those in the us for electric power train lithium ion batteries NASA says looking at this technology now very heavily and one of the neat things about electric is suddenly having this big internal combustion motor that had to be in one place and turn a propeller we don't have a problem anymore we can distribute propulsion around the airframe because these motors are very highly powered is and they maintain that power density as we decrease their size so suddenly I can fragment the propulsion system put thrust where there is drag and get favorable aerodynamic interactions Ok so this is all happening a lot of neat stuff you might ask why electric well in the context of the electrification transportation you probably aware of the reasons the same things for airplanes is for for other vehicles one is that they're quote unquote green are right we take the problem of c o 2 emissions and other dangerous emissions off of the vehicle.

[00:07:15]
Now that doesn't mean that we don't have those problems anymore they just simply transfers the problems to the terrestrial grid right so now Ok And you know it frankly that's Ok for me because I know we can do things that scale there and more efficiently thermodynamically Ok so there's some good reason to do that but by the way we need to be a little bit cautious because when we look at the lifecycle carbon we need to consider them in your faction of the aircraft we need to consider that batteries don't last very long and they have to be replaced and they have rare earth metals in them Ok so this this stuff is just now now coming up.

[00:07:52]
I mentioned distributed propulsion with Nasa's introduced it's mechanically simpler Erin the Amec advantages so that's exciting The other thing is low noise if you've ever been to a general aviation airport where you have small airplanes you'll have in your mind that distinctive character of sound of a piston aviation engine I love the sound but it could be annoying to a lot of people and it's a problem for airports that operate a lot of airplanes right so lower noise because of electric and most importantly most profoundly lower cost and pretty much all aspects lower cost of acquisition the capital cost of buying the aircraft lower cost of operations because now I sourced power from the electrical grid at say $0.11 per kilowatt hour as opposed to $7.00 a gallon for unleaded aviation gasoline yes it still has a lead in it right now the big jets but the little airplanes Ok and salute just a lot of cost advantages the other ones maintenance cost.

[00:08:57]
Electric motor has very few moving parts really just the sort of really single moving part and so it's a lot simpler it vibrates less I means it does less damage to other parts of the airplane and the maintenance costs come way down. Ok so I showed you hey this is happening there's a lot of sort of small efforts things that don't look like they'll change the world out there right from that previous slide but look what else is happening there's a rise of something that's called Evie tall as the acronym e.-v. tall that stands for the lecture Rick vertical takeoff and landing the tall vertical takeoff and landing is an old aerospace term for things like helicopters right but also crazy things that they try to the 1950 s. having jet engines mounted up and down that would directly lift you off the ground things that would tilt right lots of configurations that give you a vertical takeoff capability turns out all those things they tried were immensely mechanical comp mechanically complex and failure prone and effectively none of those were successful other than the helicopter except for anyone know which one we use now a vertical takeoff and landing aircraft has something that tilts f. $35.00 is fair I stand corrected 2 things if there would be 3 who that Harrier stand corrected I was thinking of the $22.00 so there's about 3 of them the v. $22.00 Osprey a the Harrier jump jet and the f. $35.00 right so there's a few but now here comes electric and the entire game is changed nobody knows how to do it and so what you see on this slide is a few companies that are trying different things here's a 6 images of different prototype aircraft that are flying the top left is a company called Volvo copter in Germany.

[00:10:51]
And it's basically a giant multi copter right you see a quad copter x. a copper flying around campus just imagine you put 18 of those rotors out there and you've got a void topper. The one in the middle is a company founded by Larry Page actually Larry Page being the co-founder of Google he's wanted to do this stuff for a long time it's called the 0 core us and you see it's a multi copter it's got a whole lot of those little rotors that it's got a wing in the middle and the wing gives you for flight efficiency a lot of rotors gives you however efficiency but if you want to go far and go fast you need a wing and so that's why they're looking at that configuration air bus behind a top right tilt wing both of those the front wing in the back lean tilt and that gives you some air in Amec advantages the job e s 4 on the lower left is a tilt rotor it has in this in this rendition I think it's 12 rotors a new version has 6 each one of those tilts independently of the way Aurora flight sciences as a company that was actually acquired by Boeing for their work in this area and they have a configuration it's now becoming called Lift plus screws you have a whole series of lifting rotors on those 2 pods on the side and then you have a push a propeller on the back of the airplane and lastly you have this company called Lilium out of Germany in the bottom right which is the entire depth it's a ducted system and the ducts above the trailing edge of the wings tilt Ok now for some context here 6 examples.

[00:12:27]
The one of the professional societies in aerospace is called the American helicopter society I should say it was called the American helicopter society they've now rebranded themselves as the vertical flight society largely because of what's happening in this space and they're tracking over 200 prototype aircraft of even taller craft in development worldwide now some of these admittedly are just people in their garage but many of them are now increasingly large aerospace companies well funded start ups etc And I'm showing some of the levels of investment here so this is happening and it's really exciting now you might say why why would people be interested in an aircraft like this well they're targeting a market that's becoming known as urban air mobility which is given the acronym you am so urban immobility the name says it all right it's flight in cities it's an air service that operates in cities so think about an aerial taxi that might carry you as a passenger from somewhere in Atlanta down to the airport that would be a mission in urban area ability and what really launched this field is that uber said they're interested in 2017 and they started initiative called Uber elevate and these 2 images are from that elevate white paper and it really sums it up so if you look at the image on the right you see a parking deck in the city the rooftop of the parking deck there's a lot of these aircraft and parking stalls there's a couple of excrete that have an h. listed on them that's the landing area for the for the aircraft and then you see the aircraft it's one of these novel e.-v. talls with a lot of different rotors.

[00:14:12]
The other thing going on is that's in a city it's in a place where we wouldn't normally have aviation so you can think about the challenges from a policy standpoint from a land use standpoint regulatory standpoint there's a lot of complexities embodied in that image the image on the left is what you would see as a as a rider as a consumer of the service and that's your cell phone app your trooper cell phone app so uber wants you to be able to book and Guber air trip just like you can book a nuber x. trip river pool or maybe better yet multi-modal service where nuber car takes you to avert a port as these things are being called you get on the aircraft you fly to another location and then you do last mile in Newark are again right so a full journey so is a pretty exciting kind of thing and I think for those of you working nobility you recognized in any challenges of doing that on demand managing the fleet managing the crew the pilots there's a lot of difficulty and opportunity there and so what's different about this you am compared to other forms of aviation and you might have detected already that the answer's basically everything's different the user experience is different we could have passenger air taxis and by the way I didn't really talk about the cargo side but a lot of these aircraft are being Invision for carrying crates carrying cargo packages that have certain urgent delivery needs to them I wish I would have put the slide in here that showed some of those specific cargo aircraft because this is the supply chain logistics Institute after all that there's there's a quite a few novel configurations specifically tailored for cargo same same overall technology though.

[00:15:59]
The aircraft technology is different there's the electrification aspect there's also the flight control aspect which is really hard and then there's the aspects of autonomy going even farther than flight control of could we get rid of the pilot that's what people want to do in the long run to really further bring down the operating costs operations and infrastructure how do we manage the infrastructure the land use issues the public policy pieces the airspace management if we imagine how the f.a.a. manages air space in the u.s. now one controller one flight controller typically handles about 20 airplanes we now imagine a world where there are tens of thousands of flights above any given major city around the world at a time we're not going to hire an extra $505000.00 flight controllers right we're going to automatically manager cities and handle the dispatch problems Ok so this is another piece of research that's being worked.

[00:16:59]
So you know kind of channeling those things you know wise you am difficult the electrical technology battery technology it's still not quite there so people are working on this metric that is specific energy the amount of energy per unit mass of the box that is the battery pushing that number higher pushing the volumetric density higher as well.

[00:17:23]
Simplifying flight we talked about autonomy things like that the other aspect is even if we can't go to pull up Tani right even if we still have pilots there's a massive pilot shortage right now we can't find enough pilots to fly commercial aircraft around the world so companies are scrambling to train them into getting prepared so one idea that has been put forward is what's called simplified vehicle operations the idea that maybe we could make the aircraft simpler to fly we automate to the inner loops of flight control and we automate increasing numbers of those but we always maintain the most difficult part of situational awareness and perception and a sign that the human Ok that's an entirely different paradigm of flying aircraft and we don't know how to do that yet and we especially don't know how to get the f.a.a. and other regulatory agencies to approve anything we put forward base that's a bigger area that's getting worked manufacturing Let's say we want to build tens of thousands of these aircraft per year that's actually sort of happening in people's planning processes now.

[00:18:36]
To meet demand Well tens of thousands of aircraft for years a very different scale than aviation if you look at Boeing Boeing build hundreds of aircraft per year and it's a different scale than automotive where they build hundreds of thousands to millions of units per year somewhere in the middle right composite airframes that are difficult to manufacture we need to build $10000.00 a year nobody knows how to do it the aerospace manufacturers they will scale up an interesting way the automotive manufacturers are entering the market and saying we'll scale down our rate of production and we'll be able to address the market in that way.

[00:19:12]
Anybody follow the news of the Consumer Electronics Show you have a look in this time of January to see what kind of exciting new electronic gizmos are going to show up in Las Vegas that just happened about a week ago and a big new Splash announcement is Honda I the car company has founded an urban air mobility division and they're going to build an aircraft prove or they showed a picture of it I wish I had it's got rotors everywhere but they are going to leverage their manufacturing strengths to this market.

[00:19:44]
Noise is a big problem even though these things are expected to be quieter than helicopters helicopters have a lot of noise challenges to get public acceptance there's a lot of detailed work on there are the a mix in acoustics to really allow us to design these are crafted for low noise Ok So where might we operate these aircraft because initially they'll be very short range maybe less than 60 miles maximum range that's not very far you're not flying between cities so you're probably going to end in a 5 markets where there's a lot of population density and where you can be very competitive against ground transportation modes in an economic sense right and so I looked at a map of the u.s. from the Census Bureau that shows urbanized areas and the urbanized areas are indicated by the purple dots here and I just circled of you know some of the biggest urbanized areas in the u.s. with 50 mile range.

[00:20:46]
Diameter circles Ok And you can start to see that paid maybe I start to start to get somewhere there I can I can start to address a lot of urban areas in the u.s. with a 50 mile range. But as you're very well aware every city is different its shape is difference morphology of the you know where people live is different the population the spatial population distributions the geography the terrain if there's mountains if there's rivers if there's a big way all these things are different and now imagine where I would place that infrastructure which is going to be largely new in order to serve the potential demand and even to generate the demand for the service right remembering now that these aircraft have a limited range and they need time to recharge between flights batteries are not instantaneously recharged it takes minutes maybe hours to recharge batteries so if you want to get operational tempo up you're very concerned about that so there's a lot of challenges about the city and which of these aircraft is better for which types of cities and one of the reasons and by the way which type of service to what's your vision of how this service operate right Will you allow a stop in the middle where you can pick up more people will you not is it a scheduled service is an on demand service is it somewhere in between all these are open questions and depending on your answers to those questions because batteries are so in are jet Occleve fragile right now you get a different aircraft and that's why we saw in that earlier slide we don't have a dominant aircraft configuration yet.

[00:22:34]
So the key points here that I would make right now we don't know is an aerospace person I'm an airplane design person by training I want somebody to tell me the requirements for the aircraft right and that's old school military language of here's what the aircraft must do for us to buy We love that as engineers because then we can just check the boxes but nobody has done that yet and so there's a big effort to understand and discover the requirements even at the highest level payload range speed those are the main things we need to know and those things are going to vary city for City and service type to service type Correspondingly we don't know that aircraft configuration as I said earlier because that's going to depend on those requirements and lastly another important part is we don't even really know the physics how to model the physics for these new aircraft properly especially in the early phases where we're making trade off decisions based on these requirements right we're trying to get in the right ballpark of then surely I can hire 80 engineers or 100 engineers and tell him design a vehicle work on it in great detail and we'll get a good solution but to even figure out what's the right vehicle to work on is big challenge because of issues like the noise the ergonomics and the battery performance.

[00:24:00]
Ok so that's a that's a background on kind of what's happening in this field I want to introduce you to a little bit about what I've been doing in my group in some very rich collaboration's that we have going on here Georgia Tech and as part of this I also wanted to make you aware of it and let you know that in this new center we formed we're trying to build momentum and interest in georgia tech to put together teams to go after these problems so that that's kind of my alter your motive in giving this talk by the way.

[00:24:30]
But some of the things I want to tell you about is the work that I have done on looking at Evy tall performance modeling and this is done with Hoover and funded by NASA subsequently a little bit of work on lithium ion battery modeling at the right level of abstraction for the tradeoffs with requirements that we're working on now and lastly just to touch on some operations research type problems that I've been doing in my group and in collaboration with others to really Invision the operation and by doing that to envision what to determine what the requirements would be and this last category of the operations research element is one area in particular that I think we need a lot more emphasis.

[00:25:13]
Ok so I did a study in 2017. Which was looking at the basis to ask the following question What is the level of battery specific energy Ok that we need to make these things work for our trial market which is going to be the Dallas Fort Worth area that was there that was their question and they gave us some requirements and said Go for it.

[00:25:40]
I did this work in collaboration with Rob McDonald who was at the time a professor at Cal Poly and has since decided you know what I kind of don't want to be an academic anymore I'll go to work with who are doing this stuff so Rob's doing some neat work there's a video of this that's of like an hour long if you if you wanted to go see our An entire presentation it was at a summit that they held called the 1st day summit basically what we did is we started in all aircraft studies start with a mission right what's the purpose of the aircraft it's almost always go from point a to point b. but there's a lot of detail in between for a vertical takeoff American landing aircraft you have to hover for a while then you climb and then you cruise like you know forward flight and you reverse that to land.

[00:26:29]
It's important when you consider how much battery you need on the aircraft that you also Invision the possibility that maybe something goes wrong and I actually knew a little bit of extra energy right I need a reserve of energy so there's a reserve requirement which you see here is of additional hover an aborted landing and we take off again and we divert to another vertical we're 2 nautical miles away and we land again with a hover Ok so that's a 1st sizing and planning purposes you have to make sure the aircraft can do this from a safety standpoint but you know that in nominal operations the aircraft will almost never use that much energy the aircraft will almost always hover for a shorter amount of time fly for a shorter distance etc but you have to make sure you can do it and so this defines the notion of what we would call a sizing range or a maximum distance for the aircraft and the nominal mission range and what you're seeing here there's a lot of details on the side of the takeaway is super good analysis of their own data and in the Dallas area and they have the benefit of getting of having a lot of data because they carry people around Dallas anyway and so they know where people want to go origins and destinations of their trips and they determine from that that a 55 mile maximum distance would be acceptable but the nominal trip would be 24 nautical miles or I'm sorry this is actually statute miles 24 statute miles Ok and the on the on the pull on the right you see a plot that is the distribution of flight distances in a network that they were envisioning at the time.

[00:28:09]
Yes that's kind of our starting point and we modeled a whole series of aircraft configurations these novel e.-v. tolls you see a few of them here everything from a single main rotor helicopter coaxial helicopter to a whole lot of hybrid configurations and here I don't need energy hybrids of what I mean is propulsion hybrid some that have tilting wings some have tilting rotors right different configurations for each one of those configurations we did some analysis we did some literature search to establish ranges of potential more fundamental aerodynamic or propulsive performance characteristics So for example.

[00:28:57]
How much lift to drag ratio could we get out of these aircraft turns out that's the predominant impact on efficiency and forward flight the ratio of lift to drag. Typically a Boeing 747 might have a lift to drag ratio on the order of 18 Ok 18 that's that's pretty good sailplanes could have a lift to drag ratio 50.

[00:29:22]
A multi copter if you fly a quad copter has a lift to drag ratio of your lucky of like 2. So that gives you a sense of what we're talking about in terms of the efficiency and for flight now there are some people who say we can get an analogy of 5 out of a quad copter So we'll see.

[00:29:41]
But we presumed ranges of these values because knowing the different engineers who designed solutions that are novel and could could have different tradeoffs of these parameters the other with disclosing disclosing is the measure of efficiency and hover and you know it's basically simple idea it's the amount of weight carried per unit area swept by rotating blades Ok The higher that number the higher the disk loading the less efficient the aircraft is and hover Ok so for each one of these we have a series of these values and from that we can immediately derive the requirements for power and hover and power in cruise power is just one step removed from energy because if I take power multiply it by the time it's in cruise of the time it's in hover now I have the cumulative energy requirements for that mission so that's why we can we can look at it this way.

[00:30:38]
And each one of these configurations I can look at power required in cruise versus that to drag ratio and power required to hover versus disclosing each one has a particular place and I could cope plot those if I wanted to as you see here where crews powers the x. factor hover power is the is the y. axis so it.

[00:31:01]
And you know each one of these little bubbles colored bubbles indicates the sort of range of variability we might expect for each one of those canonical concepts based on the designer's approach to synthesize in the aircraft right so that those little bubbles become a kind of basin of typical capability for those aircraft.

[00:31:23]
And what do we want we want aircraft that have really low hover power and low cruise power Ok ultimately So we want to be as far to the bottom left of this plot as possible this is kind of a parade a frontier by the way if you recognize that term and multi objective optimization problem if I minimize both of those things that's in general good but when I didn't specify a particular mission a particular amount of time in hover versus a particular distance or time in cruise then one of these is probably going to be a winner and so you know with data like that and knowing our mission we can actually overlay you'll see these contour lines a battery specific energy so we're looking at you know 100 watt hours per kilogram versus $400.00 in various places here it's to see if I can bring up of laser pointer and show you right so here's a 200 watt hour per kilogram 40600 line by the way batteries right now our battery packs for aviation are almost $200.00 per kilogram and we really need like we want more we want $300.00 we always want more.

[00:32:34]
In So with that kind of information we did a sizing study and basically the concept of sizing an aircraft to introduce a term to you is a good friend of mine industry says the question is how big he weight how weighing the wing area how thrust how much how much engine do you need and basically this was mass sizing to so figure out to carry 4 people Ok.

[00:33:00]
And go a certain distance how much battery do I need and how much aircraft gross weight do I need and so I'm going to skip the details of this plot but other than to say we presumed a baseline of 300 watt hours per kilogram specific energy and we presume to baseline $5000.00 pound vehicle which gave us a nominal range for each one of these aircraft to sort of said at the near term levels of technology what can we get out of a multi copter not a gyro a helicopter etc and then now let's look at sensitivities where we say All right let's change the weight or let's look at the variability in its lift to drag ratio or it's disclosing and sort of see what the implied changes to specific energy are and what we learned from this study is that these aircraft want to be heavier than a traditional internal combustion engine aircraft we have to have a lower so-called payload fraction less of the vehicle is payload more the vehicle is battery right and that's how you get the longer range but that means the aircraft is being heavier is going to have a higher capital cost potentially to start to see the character of some of these of these trade offs.

[00:34:15]
the other thing we we did is is a very careful accounting of how much energy each mission segment required and a key bid of this is that for safety reasons we were almost would were just never going to access a certain fraction of the batteries capacity and the reasons are reserve one we only fly reserve been an emergency situation but also there's margin that we apply to protect the battery if you take a lithium ion battery and you start discharging it and you get to the very in the ever looked at your cell phone when you get below about 20 percent it just falls off like a rock harder laptop same kind of thing can happen to airplanes because what's happening is the voltages dropping i have to hold power and i've got to increase the current flow and that's a faster draw it out of the capacity of the charge so this really strongly coupled effect means that i really need to make sure i never used the last 20 percent of the battery and i may not even want to use the 1st 10 percent because have you noticed when you charge your phone that it charges pretty quickly except for the last few percent right the last 5 percent takes forever to charge it's because the battery had to go from a constant current to a constant voltage charge mode and it charges slower now imagine you're operating a service were you want to have high rate of operations here not going to tolerate charging that last 10 percent perhaps ok so that's the news dinner gee that's just deadweight we carry around of us inaccessible energy so it's only a portion of that success a bull and that portion changes with distance rights as i fly the aircraft a different distance Different aircraft have different characters of these types of plots showing how the energy is used one of which I'll highlight to you here is look at this thing called a tilt duct Ok this one has an enormous amount of unused energy because it's hover efficiency is so poor Ok so the result is that I can have very notice that this stacked bar chart doesn't even go past 10 miles range because that's all we can get out of it with this battery type considering the need for all these reserves and similarly look how much the energy is of the energy is consumed by those reserving reserves and particularly in hover Ok so we see the different character of these concepts coming out and yet if I change those requirements I get different answers.

[00:36:45]
We can look at now battery performance of recharging let's presume I recharge after each landing so these zigzag plots are plots of the battery state of charge you can think of it is how much fuel is in the tank right this is a fuel gauge plot verses time time of day in operation Ok so you start out at the beginning here maybe this is I forgot where we normalise that this must be 6 am right so the x. equals 0 time equals your location let's say 6 am you fly a flight and the battery discharges So you have that negative slope then you land and you recharge it to dash line moving upward and then we we didn't fully recharge it look we don't have to do that uber said Hey just charge for 5 minutes but charge some ridiculous power like $300.00 kilowatts there's some new technology that might let him do that so you charge for 5 minutes put a little bit of energy in you fly again you deplete that energy charge again cetera now eventually what's interesting is let's say you're flying this nominal mission 24 miles let's see if I want to fly that repeatably just repeated back to back 24 mile missions what we'd find out is a venture used up that buffer at the beginning and then I'd have to charge longer than 5 minutes in order to safely dispatch the next mission right because I have to protect my reserve and all that so there's some interesting character of Hama charging time you need based on sort of sprints that you might fly during the day during rush hour I might be Ok with flying with a net decrease in battery charge knowing that that initial buffer would let me get in more flights in a fixed amount of time so there's you can just see all the operations research questions that are kind of emerging here.

[00:38:35]
Yeah. The potential for swapping bears is it's a good question and I'm in a going to have one slide on that at the end of the talk where we've we've modeled that as a job shop scheduling problem where you schedule charge batteries into chargers and you have an inventory of spare batteries and you have an inventory of chargers and you try to bring down the cumulative cost considering full lifecycle net present value type cost including capital cost I think it's really interesting idea we interviewed aviation mechanics about this and other people and there's a lot of practical concerns for one these things are immensely heavy So imagine of a 5000 pound aircraft and you're taking out a 3000 pound battery pack that's an enormous undertaking specialized equipment it's not people removing them and it turns out that the connectors are very fragile and fatigue easily so just very fundamental stuff about the plugs is a little bit of an issue here but I think it's an idea that people talked about and there's a need for more work again.

[00:39:45]
Now I want to talk a little bit about how we model batteries and this is change a little bit change in direction I I'm going to go quickly through this but I think it's important to understand you know I come from a background in aviation fuel burning aircraft which is a pretty simple idea I have so and so amount of fuel in the tank and I can get that energy out up to whatever I can access with a pump that's in the very bottom of the tank but the battery the interesting thing is the amount of cumulative interview that I get out of it depends upon the rate at which I ask for it right so the profile of power delivered over time affects the cumulative energy we get out and if I really draw down the increase the power a lot and draw all that energy with high power the batteries going to heat up and it's going to dissipate a lot of the energy it contains is heat so there's a lot of losses associated with pulling a lot of power.

[00:40:42]
That are so there's a tradeoff here and we need to be able to model that in a way that lets us understand the impact on the aircraft performance. And turns out the battery models of there's all sorts of battery models there's the electro chemists rights of who work on constituent of models and ordinary differential equation actually partial differential equation models based on detailed descriptions of the battery design the anode the cathode the separator all the details there those are challenging they run rather slowly and you need to know a lot of information to deal with them.

[00:41:21]
There's equivalent circuit models if you're controls engineer you probably work with equivalent circuit models where you make everything look like capacitors and resistors and inductors that change is a function of state of charge Ok and they build models that are very useful and they're actually used to develop battery control systems Ok so there's that category models but what we're interested in now is something more abstract a higher abstraction model where you don't need to know as much information and that can be used in early conceptual studies in operations research studies so we've leveraged a series of empirical discharge models based on a paper by Sheppard in 1965 that have been modified over the years and that's what we used in the super study and we sort of developed our own way of modeling these and.

[00:42:10]
It's based on these and these curves these empirical curves and this is what if you buy a battery you get a datasheet and the datasheet tells you the cell voltage is a function of the capacity discharge right and in c.c. it drops as I discharge the battery the amount of gold to just providing is going down and as I increase the current I ask for out of the battery the voltage goes down that's the different curves on this plot different currents higher currents have lower voltage delivered to the load Ok so there's there's a lot of interesting things going on here and this is data for this cell that we used in that NASA electric airplane that they're actually going to fly out and NASA Armstrong later this year so this is data from that cell and if we apply some best practices like we talked about earlier like maybe I don't fully charge the battery because I don't have time to and I certainly don't let the battery discharge below the last few percent for safety I trim that data and I say my operational range of state of charge of the batteries and the Arrow or I look at these curves I say wow those are pretty close to linear it's like a multilinear kind of thing and in Indeed you look at multiple lithium ion cells most of them have that character so we developed a really simple model discharge model empirical that has some physical basis to it so I might call it simian periodical where you model the voltage you is equal to some open circuit voltage you 0 minus some k. times the the discharge to pass to the that's the x. axis values Q Ok minus the internal resistance are times I the current the internal resistance r. is the amount of heat that we're going to dissipate ultimately as I draw down the battery at higher powers.

[00:44:08]
And then there's interaction term right those the slopes of those lines change as a function of current and and the capacity so that simple model does a pretty good job and here you see a. Curve fit of that now for those of you who are you know work in stats or or would rather make a more complicated emulator model you can do better than this simple in your model but I love the physical insides i love we can extract the a measure of the internal resistance and calibrate that that data is usually provided by the battery manufacturer the internal resistance so there's a lot of physical insight to it and we get a pretty good overall fit model and so we've used models like that remember this is a constant current model we can turn that into a constant power discharge model and all skip the details but a very good model can be can be developed in that approach by simply a little bit of machination a a.

[00:45:09]
A little one additional linearization which has small effect and the result is we get it we get a nice model into Skippy equations let me tell you what we could do with it we can here's the canonical problems of airplane energetic usage right one you might ask the question if I have a basically a certain amount of energy.

[00:45:33]
Let me see if I can get which problem is which I make sure I see it right the 1st is find the amount of energy that can be delivered when the battery goes from a charged state of charge level Q one to a state of charge level cute too.

[00:45:48]
Here's an example for a car something like that when the fuel gauge reads 3 quarters of a tank 2 in the fuel gauge reads a quarter of a tank how much range could I get in that distance of the fuel gauge that's the top problem the 2nd problem is let's say have a final capacity I want to find the final capacity after delivering amount of energy at an initial given initial capacity at a specified power.

[00:46:17]
That's basically I started at 3 quarters of a tank I want to drive 20 miles now what's the tank read what's what's the gauge read write maybe it's half tank right that's basically it and the last one is kind of the inverse problem this is really useful in the airplane design process which is fine to find the initial charge required to fly with a given amount of required energy to given power and a given final capacity right so I want to land with a tank not empty I want to land there's still some fuel in the tank and I want to fly this far what did the tank gauge do the gauge need to read before I took off this is the dispatch problem we don't dispatch the aircraft unless that happens Ok so with this model we can do a lot of cool stuff including this is just showing the cumulative energy delivered by the battery as a function of the ending state of charge where we started the same initial state of charge and each of these curves correspond to different levels of power and what we see is that the highest power delivery curve results in the least Teemu live in or g. delivered right so it's dissipated more heat and we've basically lost of that energy to use in the vehicle so this is a really rich problem that if you think about it from how that connects to operations it's it's a 1st order effect it's not a 2nd or fact.

[00:47:50]
So I change gears a little bit now and talk a bit about more about the operations research type work that we've done in my group in collaboration with others at Georgia Tech and I'm a little bit you know shy Barris to present our our work in this area to this crowd because that's what you guys probably have a lot of strength in so please don't laugh me out of the room but I'm excited by it and I think there's a lot of neat things that can be done now we wanted to work on finding requirements right so what is the range requirement for the vehicle how far do you really need to fly these things to serve a particular business model and so NASA asked us to do the study and they wanted us to model several cities Los Angeles and San Francisco in particular and basically they said hey is there anything you could do with census data look at where people live look at where people were look at their income levels perhaps to determine their willingness to pay and their willingness to fly these aircraft versus to drive a car based on time savings and then figure out where you put the infrastructure on the ground the new verdict boards And then figure out what the distances are between those reports Ok so that's the path that we started and I'll say that we're just kind of well show you some of our initial work in that area one of the problems we solve is the problem so-called placement where would you put these take off and landing pads and you know this is a map here where the you see a lot of.

[00:49:19]
The segmentation of the map here is that is census tracts for this so those of you've worked with G.I.'s systems the census tracks are what we're looking at in this is San Francisco Bay area. And we brought in an a.p.i. to like being maps or Google Maps a.p.i. so you get turn by turn driving directions and you get drive times as function time a day in traffic right so now we can look at Ok 1st mile last mile let's presume people drove to the verge of port in the drive from the verge of port for their 1st and last mile so we can say let's imagine a point I want to put on the map and it's that magenta dot Ok let's now draw a 5 minute catchment zone around that magenta dot and that's the drive time catching sort of people who are willing to drive for 5 minutes how much of the map could you cover to get into that port and sets the green area you see surrounding this magenta dot and so the problem we solved was and then a certain portion of the people Ok that live in or work in the area shown by the magenta by the green plot have an income at a level which might make them willing to fly these aircraft Ok so there's a sort of threshold income and we just at this point presume the market share estimate Ok of those people to say maybe 5 percent of those people fly like that there's a lot more work that can be done in that area as you can imagine so basically than the problem is an integer programming plot problem because each one of these tracks can either have averted port or not have a port you move around you say I want to place 10 Virt of ports you simultaneously move around 10 of those magenta dots until knowing what the origins and destinations of the people are in terms of home and work thinking about commuting by placing those 10 verge of ports we generate the maximum number of trips in k. that's the problem that we solved and.

[00:51:24]
The this is kind of how it looks this is now l.a. area and you see the top top 10 ver ports than 20 than 40 I want to call your attention to what happened in the bird ports themselves by the way the tiny little blue dots on the left figures and the purple lines are the lines connecting those further ports and what you probably can't tell it but the thickness of those lines are arc is connected to the is proportional to the implied demand for that o. d. pair Ok So basically what's interesting in l.a. is traffic is so terrible in l.a. that if you could place 10 ver ports Your best thing to do is place it right around l.a.x. and ride around downtown and you'd still generate a lot of demand now as I increase the number of reports I'm going to place we open up that network and we see some more distant routes begin and then we go to 40 verdict ports even more so Ok So that was kind of interesting and then we here's we're coming back to what I'm interested in my primary research area yeah.

[00:52:28]
We did not we did not look at capacity this is unconstrained by capacity this huge question So when you think about airspace management everybody realizes that verged port capacities can be the bottleneck on this whole thing if you can fly tens of thousands of flights per day that's really hard to schedule these things in the for ports and from air space standpoint yeah.

[00:52:49]
But what we derived from that was distributions of trip distances Ok So remember in the study I showed you a c.d.f. of those distances same thing here we wanted to be able to do that looking at a different number of reports in different cities and we would compute you know the the mean the 50th percentile and then we would also compute sort of something close to the max the 95th percentile trip distance and for 10 minutes really short 11.8 miles mean distance and 15.8 miles.

[00:53:22]
You know. Maximum And what's interesting is that is considering competition with cars right that's considering competition with cars and a penalty for wait time at the verge of pores Ok So that is telling you the potential about really was telling us how bad traffic is no way as that's the big thing it's telling you.

[00:53:43]
The other piece this puzzle is you know doing something other than just picking up market share out of thin air really doing some work on demand and fundamentally might ask the the I'm collaborating in this topic with Professor Laurie Garo in civil engineering and she's an expert in transportation demand modeling specific focus in aviation so she's done a lot of work in aviation and she's really one of the leaders in getting into this or better mobility space in terms of demand and she titled a paper you'll see that sort of 2nd paper on the right here 1st paper on the right titled If you fly it will commuters come predicting demand for easy tall earlier trips and that's really the question we're asking at what value what cost to a ticket price and at what level of time savings would you prefer to fly as opposed to drive so to get to the bottom of that because this is a new market we don't have existing data to look at Professor bureau has set up big nationwide surveys nationwide at least 5 cities in the us 5 major cities where we've had maybe 2500 respondents in each of these surveys the 1st one looking at just kind of current day where you compare and contrast with driving a personally owned automobile were uber right here and the 2nd bringing in autonomous ground transportation right if you're going to predict the future about flying What about predict the future about driving too to make it a more fair comparison so you know these surveys are basically structured as choice models if you know how that works and here's an example on the right you know would you prefer to take this car if the cost is $5.00 and the travel time is 40 minutes would you prefer to take this aircraft where the cost is $10.00 flight times 15 minutes there's an interesting time of 20 minutes and.

[00:55:38]
You know you have a guaranteed ride home so there's a lot of effort to construct these surveys. In a way that you get meaningful data and that's really pretty challenging to get people to say what they're going to do but that's been very useful so we've been collaborating on that to build multi normally a logic models to then feed into these these other types of operations balls or work on.

[00:56:04]
And just another example of work that we're doing in this is the battery swap example that was asked about and we just put in a paper on this where we're look we've looked at now regional aircraft not this aircraft flying in cities but slightly longer range fixed wing aircraft more traditional looking airplanes.

[00:56:25]
The example we looked at was actually Mocha Laili airlines in Hawaii they fly a lot of really really short routes between the islands and Hawaii and the potential for them to save money with electrification is enormous so we were studying that problem and we asked the question well what if you could swap the batteries and you have extra batteries and you have chargers so what you do is schedule batteries into Chargers and the key ideas as follows It turns out that you pay for electricity not just based on the cumulative amount of energy you get out of the wall when you plug it in but the again the rate at which you ask for it how big is the plug right if it's a $120.00 volt plug that's one thing if it's a $240.00 volt you're going to pay more from the power company if it's a $1.00 kilovolt you're going to pay an enormous bit because they had to plumb the whole system they had to get the cable into you and you're going to pay per month for that so what if you could do peak shaving such that you could basically bring down your cost of electricity on average maybe then it would justify the logistics problem of managing these extra batteries buying extra Chargers and charge him slowly overnight so we looked at that problem and modeled as job shop scheduling and we should some pretty considerable benefits in terms of net present value that paper is under review I hope it's going to get accepted soon so when it does I'll be happy to share that around anybody's interested.

[00:57:57]
And just a few more slides the last one or 2 things I want to mention is that. We love to build their airplanes if you're in the aerospace Department our students love to build airplanes it's a lot of fun this is an exciting time to be an undergraduate student in aerospace or a graduate student for that matter you want to go work for one of these men Evy told companies so we're doing some of that where we build a subscale airplane we fly we work on the control algorithms and things like that so we built a subscale tool rotor aircraft you see that those back to rotors tilt the come propellers and forward fly we just did it because we could it was fun in doing things like that the students here on the left on my favorite photos here they're there they want to competition based on their work on that we're also designing and building novel short takeoff and landing aircraft this is a aircraft a 10 foot wingspan 6 propellers 3 on each wing semi span so it's a blown wing configuration by the blowing the propeller slipstream over the wing we augment the lift of the wing at low speed so we can get incredible short takeoff and landing performance where this aircraft we haven't yet flown it but it's going to happen in February.

[00:59:17]
Can probably take off and to body lengths of the aircraft Ok so it's really pretty exciting and so you know these are some of things you can do with electric repulsion reimagine the airplane basically. So in this big context of here's Electric Slide Here's your mobility one thing that I realized earlier on early on is that.

[00:59:43]
This field is a lot bigger than my own little research area you know I'm an airplane design person I'm interested in performance modeling things like that but I've gotten kind of drawn into this or our side because as for the funding has been things like that but it's really occurred to me the problem is bigger than me and we really need a lot of other people engaged on different parts of problem and we need to go after big proposals and try to do fun stuff together in teams so we formed the Georgia Tech Center for Urban and Regional air mobility we've been socializing this idea across campus connecting with people like Tim and been on this topic to see what we can do together to really solve some big problems were doing things like working on the occult technology Ok here on the left you see aircraft technology research that's a bundle of work where you can just imagine Georgia Tech and grid engaging across the board everything from the people in electrochemistry and battery design to people working in mechanical engineering and high powered in city Motors to people working in electrical engineering and high powered into the electric drives right aerospace people doing integration on vehicles and thinking about are an Amex that's all really cool stuff but then there's the other whole half of the problem that is operations it's the public policy piece of it it's the economics modeling piece of it all these things are of huge interest because companies are pouring billions of dollars into this field and frankly nobody knows anything about the answers they're on the right hand side we there's so much uncertainty on how we build the airplane the left hand side of their craft technology but I think that pales in comparison to the questions on the right hand side of this plot case we form the center to work in this direction.

[01:01:23]
We had a kickoff meeting in January of last year and we sort of framed it as a local problem too we said hey if you're going to do research in urban airmail bill. You need a lab that is an urban area right so we're situated in Atlanta Why not engage the local community to see what we can do maybe we could have some test ranges maybe we could gather data from Marta and other places rights a lot of potential for that so we we had a co-sponsorship between the Metro Atlanta Chamber of Commerce the Georgia Chamber of Commerce and the Georgia centers that invasion so that was a really nice event we had about 100 attendees we had a national panel of speakers from her and other places talk about you am the opportunities we're currently working on a lot of proposals and thinking about how we can go after federal funding on this particularly for multiple investigators for us so if you're a faculty member you're interested in this area I love to talk to you about it figure out how we can write something to create s.f. that kind of thing we have an interesting collaboration going on now with University of Georgia and the Marine University on rural health care applications of drones and not just drones but bigger aircraft what can we do and can we can we do something kind of cool in Georgia to compete with what other states like North Carolina are doing and if you may have heard of you we hear of Carolina's work drone delivery in North Carolina is a big thing now they're taking a leadership position so I think the question is can we do that here in Georgia and lastly we're collaborating we're exploring collaboration with Hartsfield Jackson International Airport in this in this topic area Jim and I attended a meeting with looking at who we do something with them so this is just a quick summary of where we are and that's my last slide I've said a lot so I think we have a little more time I'll be happy to open it up to any questions thanks a lot.

[01:03:21]
This. At end of life after do after they're. So this is a great question you know batteries when I sort of got into this field only about 5 years ago I didn't realize how what a complex system a battery cell is right but it's it's got chemistry it's got mechanics going on right you actually have basically flow of of atoms across these one dimensional layers and there's fatigue and cracking and degradation that happen so they wear out short story and when they wear out they often get potentially unsafe there's a lot of things that happen.

[01:04:04]
So people have proposed one of the things that implies is that after maybe $3000.00 cycles that's kind of about where we are now these things may be unusable maybe Dave degraded such that 20 percent of their the initial capacities gone so you need to replace them there's a lot of open questions about even how and when to replace them do you do it as part of a nominal aircraft inspection cycle or is it a special event.

[01:04:31]
And then what happens to the batteries Afterward I think that's what you're asking one idea that people talked about is hey well if we're going to have a lot of these aircraft flying around to these reports and recharging Well what if I took 2 used batteries and put them on the ground and use them as a power bank on the ground to continue there's not as critical the to continue to be recharged into then peak shave peak shaving from the grid and then charge the aircraft out of that battery bank that's one idea otherwise I think it's just the same character a problem we see in automotive and other places where there's a lot of clever ideas but at the end of the day nobody has any great answers and we kind of just hopefully figure out how to recycle them so we could at least get the Cobalt out of.

[01:05:19]
The open question yet. Ok. Good question What was he getting. 51 and so 55 on his around 50 on the recovery yeah getting well you know what. I would. Describe question so 1st let me address your comment about your friend of the beginning I thought that was really neat cheaper for copper than on ground.

[01:06:17]
That's probably because they're taking a loss on purpose to test the market that's probably what that is but you know just to give you a sense of the numbers that people are talking about let's say 5 dollars to $78.00 per passenger mile right now is what people are thinking about k. that's that's a lot of money by the way.

[01:06:40]
But that's to help calibrated that's that's the target for once we get these e.-v. tolls in the fleet and we get them operating in the near term that's might might be where we are now the 2nd question you had about what's happening about airspace management of the verb ports and you don't have air traffic controllers that kind of thing.

[01:06:58]
Also open question NASA has a project that they've established at their headquarters level to study this it's called a.t.m. x. and they're looking at several different paradigms one of which is based on doing what what they're in visioning to do for drones right so you have the package delivery drones and you imagine those operating while the scale of operations there is much higher than what we're talking about for carrying people right it's maybe tens of thousands hundreds of thousands of flights per day NASA has a program called u.t.m. unmanned air system traffic management so it's a compound acronym u.t.m. and they're using an automated dispatch approach so basically it's a sort of a federated approach where you could have different service providers that interface with the database that manages the airspace to avoid conflicts but each one of the service providers can kind of sell a slot and it's a 1st come 1st serve thing on cell in that slot and what is that slot that's a 4 dimensional trajectory from origin to destination with some associated buffering and guarantees so there's a lot of work though to figure that out particularly when something goes wrong and when you have to catch up if you're not operating the way should so.

[01:08:15]
Open question but it's intended to be autonomous as it as an air traffic control system as well. that for the aircraft it would probably fleet owned right so the been some uncertainty and what had business model would look lie earlier it was unclear if think but goober getting in this if nuber would service the airline and by the aircraft or or just the connection to a consumer and somebody else on the aircraft mean the recent announcement by honda i about their entry into this market c.e.'s or actually it was joe be aviation joe bees another company joe be signed an agreement with hooper were they would only aircraft and they would operate the aircraft's so joe b. e. would really be the airline and lube or would work on the verge a port infrastructure the air traffic management and the connection of consumer with your app good question mostly the verb ports or place is just to pick him bark in disembark passengers right the aircraft's will probably be landed in stored it some other central facility Or distributed facilities yes or.

[01:09:48]
So other types of energy sources there's a lot of people who are interested in hybrid they just don't believe batteries have what it takes and so you know can you take a gas turban that's you know downsized a little bit now operates very efficiently for the cruise condition and we augment its power with batteries that's a type of hybrid scheme that's been in vision there's a few companies working with fuel cells that they believe can actually provide the power density you need solar is interesting generally what people have found on these these kind of applications is solar doesn't buy its way on if you will in other words the amount of incipient solar radiation you get per unit surface area on the airplane is not enough and then you have a cloudy day issues you have the maintenance issues or now I have this other widget on there that has to work but I can't really count on it and so solar is not then preferred.

[01:10:42]
Yeah. You know. It's a great point cyber security is a big deal there's a lot of talk about that in this space and I don't really it's not my area I don't really know where it stands but I know people are I mean even on battery management systems right if you could mess around with a controller that is the battery management systems you could take one of these cells and enter it into thermal runaway in which case you could blow up the battery so it's a big deal and so people are talking about voting systems and you know all these data fusion approaches to the controls people would know about to detect hey that sensor is probably bad maybe that thing has been hacked.

[01:11:30]
To death. Basically. Scary right so. It's good research topic this or I mean there's hope. That you could. Actually play like you're. Actually. It's. As a fact that. Yeah so these prototype aircraft are flying but you know did to fly a prototype I mean you sort of companies are in that might ask why they'd do that well it's part of the development cycle where they're trying to build a bigger they are trying to build the aircraft that will meet the actual requirements.

[01:12:21]
But that it's also for press and to generate. Funding for them and things like that and so these aircraft the prototypes don't fly the range we would need they're much shorter range right now they generally speaking are probably not capable being certified by the f.a.a. which means the safety threshold that needs to be met is not there there's not this is not there but it's that certification is a process by which you prove that it's there that certification that said the aero prototyping has been going on for about 3 years now this company jobi is rumored to be far ahead and they signed a contract with Goober to initiate operations in Dallas in 2023 so I don't know if that means they'll have a fully certificate aircraft in 2023 or if they'll have an experimental class aircraft that will fly some trial flights it's a little bit unclear the certification process is a potentially 1000000000 dollar problem per aircraft or more.

[01:13:20]
Potentially say 5 years per aircraft. So. That. John. Yeah great great question initially companies are talking about helicopter pilots so in the very early days you bring in a trained helicopter pilot cross train them into these aircraft now helicopters are very unstable and you have to have a lot of piloting skill even the earliest versions of these aircraft will have some kind of an automatic control system to take a lot of the pilot workload down but they're still in visioning trained pilots as you go forward there's this idea of simplified vehicle operations were increasingly you automate systems and that if you can convince the f.a.a. you can then reduce the pilot training requirements so instead of spending hundreds of hours and thousands of dollars in flight training maybe someone close more closely to coming just off the street you know with a limited amount of training could fly in the maybe eventually you get to where you could be in the aircraft and provide some high level with some basic training safety assurances and then ultimately fully autonomous it's very speculative as we get farther out as to what will happen when.

[01:15:00]
Right. Yeah I'm happiest about another 5 minutes and I have to get to the airport in a little while but I'm happy to stanch out a little more if anyone's interested.