Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ


Choose your language

InfoQ Homepage Presentations The Future of Transportation

The Future of Transportation



Anita Sengupta discusses the future of transportation with an eye towards how machine learning and AI will help shape the future.


Anita Sengupta is an aerospace engineer, rocket scientist, and veteran of the space program. She worked for NASA for 16 years where her engineering projects included her PhD research on developing the ion propulsion system for the Dawn Mission, the supersonic parachute that landed the Curiosity rover on Mars, and the Cold Atom Laboratory an atomic physics facility now on board the ISS.

About the conference is a practical AI and machine learning conference bringing together software teams working on all aspects of AI and machine learning.


Sengupta: Thank you so much for having me today. I'm really excited to be in San Francisco. I don't get to come here that often, which is strange because I live in Los Angeles, but I do like to come whenever I can. For my talk today, I'm going to talk about the future of transportation, specifically on the things that I worked on that I think are kind of the up and coming thing, the thing that I'm working on now and what's going to happen in the future. I am pretty active on social media, so if you want to engage with me on Twitter, please do so @Doctor_Astro

Engineering the Future

I think part of my career has always been about just doing fun and exciting new things and all my degrees are in aerospace engineering, ever since I was a little kid, I loved science fiction. Show of hands, people here who like science fiction? That's at least 80% category, green, in terms of passing that test. I actually am a Star Trek person versus a Star Wars person, but I knew since I was a little kid that I wanted to be involved in the space program, so that's why I decided to go the aerospace engineering route and I wanted to build technology. I got my Ph.D. in plasma propulsion systems. Has anyone heard of the mission called Dawn that's out in the main asteroid belt? My Ph.D. research actually was developing the ion engine technology for that mission. It actually flew and got it to a pretty cool place out in the main asteroid belt looking at Vesta and Ceres.

I did that for about five years and then I kind of felt like I had done everything I could possibly do on that front, from a research perspective. My management asked me if I wanted to work on the next mission to Mars. There's very few engineers in the space program who'd be like, “No, I'm just not interested in that.” ,so I was like, “Yes, what do you want me to do?” And they're like, “We want you to do the supersonic parachute for it.” I'm like, “Awesome. I know nothing about parachutes, but let's do this.” The good news is nobody knows anything about parachutes, so when you take on a job like that, you kind of have to learn on the job and work with other people who do it as a job in the field, in the private sector and learn all about that.

I did that for about five years, worked a little bit on the sky crane, which is the landing system for that. Then my management is like, “So we have a new challenge for you, which is something related to atomic physics and creating this facility on the International Space Station to make Bose-Einstein condensates.” By a show of hand, does anyone know what that is? Far fewer people, but still a few nerds in the audience. That's another state of matter that was theorized by Bose and Einstein, which is basically something which is just above absolute zero. It turns out you can do this kind of experiment in microgravity and get too much, much colder temperatures. I led that for five years, it's called the Cold Atom Laboratory mission, you can actually read about its scientific findings online.

Then I decided after 16 years at NASA it's time for me to do something in the private sector, so I got recruited to join a hyperloop technology company, and I'll talk a lot about the hyperloop today, but it's a new form of transport, which is actually very similar to space travel. I went off and did that for about a year and a half, and then I decided to do the full-on private sector thing and start my own company, I'll talk about that a little bit today. I think kind of the purpose behind why I'm saying that is that life as an engineer, as a computer scientist, as a technologist, you have the opportunity to innovate and do many different things in your career, and the more you challenge yourself, the more complicated the problem, I think the more exciting your life will be, and what you produce as a result of your career will be.

That being said, I do think it's important to have a purpose behind what you do, and for me, understanding how I can help society is important, and although the space program is awesome, I felt like I wasn't doing enough to help people right here at home. The reason why I show this graphic is because hopefully everyone knows what the planet in the middle is, that's Earth, where we live. The planet on the left-hand side is Venus, which is one planet in from us, and the planet on the right-hand side is Mars, which is one planet out from us. And what planetary scientists believe is that at the start of the solar system, these three planets, called the terrestrial planets, were actually very similar to each other in terms of having an atmosphere, water on the surface, but over the course of the past four and a half billion years, the planets have actually evolved quite differently.

Venus has a surface temperature of 475 degrees centigrade, which is the temperature that you set your oven at if you want it to self-clean itself. It has a surface pressure of 100 atmospheres, so basically 100 times what the surface pressure is here, so something happened that caused the oceans to boil on Venus which creates this sort of hellish environment that you see today. On the other hand, Mars has a surface temperature in the subzero range at best, and has a surface pressure of only 1% of what we have here on Earth, so what that shows us is literally that climate change is a planetary phenomenon, and we can view it by actually measuring the planets that are closest to us. There is an argument behind why we do planetary science exploration, to understand how our own planet is going to evolve. I think it is really important that we do what we can today to solve and combat things related to climate change so that we can protect our home world.

You can go back even further in time and you can look at any ancient society. One of the major focuses of those ancient societies is astronomical observations, the picture on the left is Newgrange, which was an astronomical observatory. The picture on the right was from ancient Egypt and they were also focused on what was going on in the heavens above and the stars. Astronomy has actually served as the basis for the development of fundamental physics, sort of like since the start of society, so I think that's actually kind of exciting.

Now, although I am not a scientist, I am an engineer and there's a Star Trek quote related to that, which is what I think Bones says, “Damn it, Jim. I'm not an engineer, I'm a doctor”, or something like that. One of my favorite quotes actually is from the father of modern aerodynamics, Theodore von Karman, which is that science is the study of what is and engineering builds what will be. That's pretty powerful because what it means is that you can take the tools of math and science to create a technology that doesn't exist yet, and hopefully make the world a better place, I think most of us can do that in our daily lives as engineers and as computer scientists. What I will say, however, is that especially when you're the private sector, when you're trying to do something new and you're trying to drive a business case, you can find people who will be kind of negative and not support your views or it may be difficult to raise funds.

I'm going through that now with my own company and I can give you a wonderful, inspirational example, which is from a very famous scientist, Lord Kelvin, who invented the Kelvin scale, basically said that heavier than air flying machines are impossible. Just a few years later, the Wright Brothers actually did their first demonstration of flight lifting body aerodynamics. You can see how it really is by tackling these challenging problems that you can actually create a brand new society and tremendous economic growth. That ties in really nicely to science fiction because if you take a look at a lot of the technologies that we use today, they actually came about in science fiction first. Star Trek, my favorite TV science fiction, basically invented the Motorola flip phone before it ever came out in the original series and Spock and Kirk talking to each other. Star Trek: Next Generation, a couple of decades later, people were working on their tablets, which is kind of what our iPads or our Samsung tablets are today.

Technology Has Changed the World

Science fiction actually can serve as a motivation for young people to create these brand new technologies. I think that is exciting and it probably motivated a lot of us in this room, it certainly motivated me to become an engineer. Ultimately, we really can build the future that we imagine, it's if we take this to looking at a particular element of technology, specifically transportation technology, you can also go back in time and see how transportation technology has essentially revolutionized our society. Go back to 3000 BC when the wheel was invented. Imagine how that changed societies for human beings the ability to move around goods and services and people in ways that had never been done before, with a level of efficiency which allowed people to use less calories probably, and able to grow their families and their societies.

Move forward a couple of thousand years, the Industrial Revolution was actually enabled by the development of the railway networks. You can really see how transportation has the ability to revolutionize society in a way that has never been done before. Move forward a couple of decades, now we're on this exponential Moore's law curve almost with regards to transportation, personal mobility was enabled by the development of the internal combustion engine and the car, which we still use to the present day, which allowed us to have expansion of cities and to suburbs, so really allowing human society to grow like never before.

A few decades more, the Wright Brothers, the invention of traditional aircraft, commercial aircraft, and engines. Now we have a society which is globally connected where you can travel at fast speeds over land masses and also transpacific, transatlantic which connected society in a way which also has never been seen before. You can really see how transportation technology has revolutionized our world.

As with all things, eventually, you start to hit a limit and transportation networks have a problem now. I think San Francisco, you deal with this, in LA I deal with this, which is congestion. Because we're kind of at peak capacity in our existing transportation networks, we're starting to feel pain in our personal lives, because it's horrible to have to commute an hour to two hours every day. There's pollution which gets coupled to that because of congestion in the air from cars that are on the road. Then even from a business perspective, if you have goods and services which are trapped in ports, trapped at airports, you actually can't transfer those goods as much as Amazon would want you to, for example, so you do need to do something about this.

If you take a look at what all cities and countries and municipalities are looking at, they're proposing things which are kind of more of the same, bigger airports, adding a runway, more highway lanes on the freeway which we know we don't need. We're not really seeing a revolution in the way that we can solve this problem, I think that's unfortunate, but that also means that it's ripe for disruption. By a show of hands, how many people here have seen the movie "The Martian" or read the book, "The Martian"? I think even though it doesn't say engineering, he says "In face of overwhelming odds, you science the shit out of it," That usually does work relatively well, it worked well for Mark Watney, obviously, so ultimately what we want to do is reinvent transportation.

Customer Requirements

I am a systems engineer also by training, I did that for several years, what you do is you start with your customer requirements. In the case of transportation, your customer is both the end user, the person who actually sits in Uber or in the airplane, and your customer is also the government who pays for you to develop the system. What we're talking about here is a way to reduce travel time, it doesn't require warp drive. Hopefully, that happens at some point in the future, but there are ways to increase speed by using new technology that I'll talk about during the course of this talk. Also passenger experience is important, I think most people say that the reason why Uber did so well is because it really provided a really good passenger experience in terms of being on demand and affordable.

Another thing which is important, which is the quantitative way of defining the engineering requirements is capacity. How many people do you need to move around per hour, per direction between point A and point B and point A and point C? A really important one that a lot of you may already be involved in if you're working on autonomous systems is safety. Ultimately, you can't create a new transportation technology which is less safe than the existing modes, what you do from the engineering perspective actually has to improve safety.

Cost, of course, is key, so nobody's going to implement your system if it ends up costing too much money. We can see that with the challenges of different rail networks which have been proposed in the United States, specifically in the northwest. Something which is important to me is that if you're going to propose a new transportation technology, it has to be at a cost point that everybody can utilize it, not just the wealthy elite, that's not fair. Mass transportation only works when it's available to the masses, so that's important when we think about the global society that we live in.

A really important one, and this is important because it couples to climate change, which is environmental impact. Whatever engineering technologies that we create for transportation, we should be doing something to reduce our carbon footprint. That means you can actually do something to help the planet as well as making people's quality of life better, by proposing a new option if you think about it in that way. Then with mass transportation, ultimately you want the system to be scalable. It may start in a particular city, go to the suburbs, go statewide, go nationwide, so you have to take into account the fact that this network can grow over the course of time.

What is in the Near Future?

What is in the near future? By a show of hands, who has heard of the hyperloop? When people think of the hyperloop, this is usually the image that comes to mind, which is Elon Musk and his evil Mr. Kitty there, there'll be more of those during the course of this presentation. It is true that in 2013, Elon Musk released a white paper, which talked about how a hyperloop could be a solution to California's need for mass transportation. He was doing this actually to contrast with the high-speed rail that was put on the ballot as a referendum in sort of the mid-2000s. His argument was, why are we going to put in this old technology, which is decades old and not very efficient, when we can do something new like the hyperloop, which is more energy efficient and a better way to actually move people around from point A to point B?

That was a really powerful message that he sent to the community, it actually put the concept of the hyperloop in the public domain and the public discourse. Interestingly, the hyperloop was already proposed and it was proposed in the realm of science fiction for many decades. Some of us watched "The Jetsons" reruns when we were kids, which had people moving around in tubes and you can find several examples of a concept of a hyperloop in science fiction in the 1960s, 1970s, and somebody told me there was actually a book from the 1800s which also mentioned something similar. It's science fiction serving as this catalyst for change in the transportation sector because of people who have these imaginative ideas.

The hyperloop is an amalgamation of existing technology. What a hyperloop is, is a maglev train or a maglev passenger vehicle, so this is something which carries people in it. You can choose the number of people that it carries, it uses magnetic levitation to get it off the track, and it uses electric propulsion to move it through. What's different about a hyperloop is that a magnetically levitating passenger vehicle is put inside of a vacuum tube. The reason for that is simple, when you get to faster and faster speeds, your primary energy consumption or energy loss mechanism is aerodynamic drag. If you can eliminate the air, you eliminate that primary energy loss mechanism, so by putting a maglev train inside of a vacuum tube, you're able to get to very fast speeds with low energy consumption. You only pay for it once when you pump down the system initially and we can talk about the community later about how one actually implements that.

This technology is something which uses existing vacuum technology, can utilize existing maglev train technology, but you put it together and you have this more efficient way of operating a mass transit system with higher capacities, with lower energy consumption at much faster speeds, basically being able to target supersonic speeds. That's pretty neat, it seems like a great idea. Then the question becomes is why are we not building hyperloops everywhere? There's a simple answer for that: it's expensive. We can even see with the California high-speed rail analogy, I think they're up to $60 billion, $80 billion. They're blowing through their budget, they're over schedule and nothing is being done. What's inherent to infrastructure projects is that they're very expensive and they typically tend to run over their initial estimates and they take a lot longer.

How Do You Innovate?

What that means is that if you're going to implement a new transportation infrastructure project, you want to come up ways to innovate to actually make it cost effective, so how do we innovate? This is just something we should drive home for everyone in this room, is that you use the entrepreneurial mindset, you use the startup analogy as a means of creating disruptive technologies and doing things differently than the dinosaurs, which exist in the infrastructure and transportation space as they do, unfortunately, I can say that from experience. Of course, I jest by showing you images of TV characters, but this is ultimately the environment that a lot of us work in now in the entrepreneurial space.

What couples nicely to the software sector is agile methodology in terms of the way that you would do your design practice where you prototype and you test and you prototype and you test. There's absolutely no reason that you can't use that same philosophy when developing hardware as well. When I worked in the hyperloop sector, and there are several hyperloop companies in the United States, in Europe, in Asia, the one thing that you do is you prototype and you test as quickly as you can, so that you can validate the technology and improve upon that design to get something that will work.

These are some examples of hyperloop technologies that have been developed in the startup space. At the company that I worked at, we actually built a half kilometer hyperloop that works in the deserts outside of Las Vegas, not to carry people but to demonstrate the technology and under vacuum going to fast speeds. There are other companies who are also based in actually Southern California who are developing a test track out in France actually, in a similar way. By doing this, you can actually demonstrate that it is feasible and then hopefully you can continue to get your investment that comes in.

How Do You Lower the Cost?

What is key to lowering the cost of any mass transportation system is to increase ridership because that's how you generate revenues. If you type in this into online for cat hyperloop, this actually comes up. I do like cats, so this is the second of many images you'll be seeing cats in my presentation. The way you actually lower cost is by increasing ridership, because you generate revenues and that's basically how you do the public-private partnership model. Oftentimes people like me desperately want to take public transportation, but it ends up being so slow that you end up not taking it and therefore your ridership isn't there.

How do you lower cost? That's actually using the concepts of AI and machine learning. If you make a system on demand, if you optimize the system where it's running at capacity when it's needed during rush hour, when it's at a lower rate, when it's not needed, you can actually make the system more energy efficient and more efficient in terms of reducing wait times. When we were developing hyperloop technologies, we were doing it specifically in mind, we're doing an on-demand system based off of what people needed because they were requesting rides on their smartphones to make that system work.

Another element which is really key here is the fact that when you're moving at these super fast speeds, you can't possibly have human beings operate these vehicles. Autonomy is also required to be able to ensure that the system is safe, because you can't rely on human reaction times for a vehicle which is moving 1000 kilometers per hour. What is nice is that the autonomy space is being obviously developed quite a bit in the automotive sector for self-driving cars. That happens to be a very challenging problem in comparison to a hyperloop, because you are out on the road, you have bicycles, you have cars, you have animals, you have other traffic that you have to account for, which makes it a very difficult thing to do from an algorithm perspective. From a hyperloops perspective, you're inside of a vacuum tube so there's nobody in that tube except for you and the vehicle in front of you and the vehicle behind you. You can actually find synergies behind what's being developed from a diagnostics perspective and from a software perspective in the automotive autonomous sector and actually leverage that immediately over to a hyperloop technology.

I actually spent a few hours this morning down at Waymo visiting my friend who works there, and they're making a lot of progress on this front, so this system is pretty much, in my opinion, from an engineering perspective, ready to go. There's also the regulatory aspect that has to couple to it so that will take some time, but the technologies that are being developed in the automotive autonomous space are directly applicable to this mass transportation space for something like a hyperloop. By being able to find these synergies in different areas, you end up having to not pay for that development cost yourself and leverage it as part of a larger ecosystem, which I think is a really powerful and important thing.

The hyperloop highway is coming to you very soon and what really stands in the way is raising more capital to be able to finish these developments, and then hopefully we'll see one implemented in the next 10 years or so, somewhere in the world and in other places we'll carry on from there. The one final thing I'll say about the hyperloop, which is exciting, is that it's very different from a train system, which stops and starts. What it is, is you're on a highway where every one of the main tube in each direction is going at maximum speed and then they exit the highway to go to the station that you happen to be at, which means that you don't have to pay for that energy cost of accelerating and decelerating. Coupling that to the on-demand architecture allows you to make sure that everybody in your vehicle is going to where you want to go and it does allow you to be more efficient in that sense.

The Future of Air Travel

The future of air travel is what I'd like to talk about next. Once again, I can use my science fiction analogy to "The Jetsons" and flying cars. There isn't any version of the future of society where there aren't flying cars, I don't like the term “flying car” because it implies that you have to be able to use the vehicle both on the ground and in the air, so I prefer “air taxis”. My new company is actually based on developing a technology for that so I'll show you a video which gives you an idea of what a system like that looks like when implemented in a city.

What we're proposing is an electric vehicle, which has the capability to take off vertically and land vertically, but because it has a tiltwing geometry, it does its cruise, its traditional flight, using lifting body aerodynamics on the wing. It's very different from a rotorcraft, because a rotorcraft basically is in this veto space all the time and it happens to be very energy inefficient. It also happens to be very noisy. The only way that we can get around the energy efficiency required to support using batteries only as your power source and the noise abatement requirements that would be required for a high density urban aerial network, is to use an alternative technology, which is either a tiltwing or a tiltrotor and we're proposing a tiltwing.

Now, the other part of making a case for this is that infrastructure happens to be very expensive, to build a new airport is very expensive. You can utilize existing infrastructure, which for example, general aviation airports- you may not know this if you don't fly for fun like I do- but general aviation airports are all over the place. There's usually three to five general aviation airports in the vicinity of any major commercial airport but nobody uses them because jets can't land at GA airports, because the runways are too short. By going with a geometry which allows you to take off and land vertically, it allows you to use these existing aviation spaces which are already in the areas that you care about in terms of urban, suburban, city type areas.

The other thing which is important is that if you use it only for urban transport, your range has to be a couple of journeys between charging, which means that you can use purely battery power, which helps us to address the climate change challenge. The good news is that batteries are being developed for automotive sector, for electric vehicles, which means that their energy density is going up because their use is going up. The cost per unit is going up because their use is going up, so we can couple that technology to aviation, to the aerospace sector at a relatively efficient cost to drive the cost down of these vehicles.

In terms of what our vehicle looks like from a design perspective, it looks kind of like a big Cessna. It has a wing which has electric motors with propellers on it, which has the ability to tilt to take off and land vertically. In terms of tip to tail, we're looking at about 38 feet in length, and the wing to tip would take about 26 feet in length, so it's kind of a small footprint from that perspective. It's nothing like a jet, which means that even if you're not landing in a general aviation airport, you can use existing parking lots, for example, to land these aircrafts or you can use tops of buildings as long as they can handle the load of the vehicle.

What I do think is really important, and I've actually learned a lot about from working with the software community, is using the agile mindset. What we're doing is building and testing subscale vehicles with our funding to make sure that we could demonstrate all the modes of flight. This gives you an example of one of our flight test, this was our very first test where we fully demonstrated vertical take-off and landing, tilt wing capability for cruise that you see here, and then eventually tilting back again to get us to land. We're located in Detroit, which is interesting, most people don't think about Detroit for aerospace, but Detroit has a huge automotive industrial design, mass production infrastructure. We utilize that and we're located right at the field in the city of Detroit.

You can see here now, the wing that's tilting, you're seeing Detroit in the distance there and we're coming in to land vertically at this towered airport, which we would propose to use for people coming into the city of Detroit, as opposed to the traditional international airport. By doing testing at the subscale, you can actually demonstrate the aerodynamics, the software control, in a much more cost-effective environment with a pretty good cycle time between design and test, design and test and redesign, and get it to work, so that's an idea of what we have done so far.

This gives you an example of our largest vehicle that we built. We've done fifth scale vehicles that I just showed you, and then we've also done a third scale vehicle and we were demonstrating at the Detroit international auto show. As we go up and scale, we go up in complexity, and we go up and showing that we can actually certify an aircraft like this in the future.

Why Take an Air Taxi?

The question becomes why take an air taxi? There are several reasons why. The first one that I would give you is to describe to you the current mass transit experience that I experienced in my way out here, as well as my cats, which is sit on an aircraft. This is the one I flew out right now. There's no room, you're stuffed into the person next to you and there's no leg room. Just as I have a fat cat and I have a skinny cat and that's the way it works, then my personal opinion of the current passenger experience, that's me. The reason why you would take urban air taxis is pretty simple. One, it's five times faster in terms of an end-to-end speed from door-to-door than if you were going in an Uber or a taxi in a congested urban environment. If you can reduce the price point of the vehicle by using autonomy coupled to automotive technology, you can also reduce the cost to the end user.

What we're looking at is a network of air taxis, which will be basically like taking an Uber ride or like an Uber Premier, if you're taking the Uber black car. We're trying to get down to that price point actually. Then, of course, you have a beautiful view if you're above a city as opposed to sitting in traffic, so I think from a passenger experience perspective, it's really a good way to travel to urban environments.

To give you another idea of what something like this would look like on the interior of the vehicle full scale, we've come up with this view of what the interior looks like, it's a pretty spacious cabin. We're targeting to carry a total of four passengers with one safety pilot, we are designing for autonomy, but you still have to have a safety pilot just because you have to go through the operational certification perspective. With four passengers, it's like an Uber ride, where you have two couples or a family. You could take one person if you wanted to, but then, of course, that would increase the price of your ride.

You can use air taxis for passenger transit, which makes a lot of sense, but that's not the only potential use case. You could also use it for cargo transport, humanitarian missions, reconnaissance missions. Here you can see how the wing tilts to enable that vertical take-off aspect and then eventually it will come back over to the cruise position so that you can do your cruise transit in a more energy efficient way, as compared to a rotorcraft. Because batteries do only have a limit to their energy density, you can't do a full range regional to regional trip, but you could actually modify the design to have a hybrid electric perspective to give you a longer range, but our MVP will be battery-only for short range missions to begin with.

Handling Air Traffic

What is really important though, and everybody asks this question is like, “Well, that's great, but can we really handle that amount of traffic in the skies as it exists today?” There is an air traffic control network that exists today that all commercial flights fly off of, that I fly off as a pilot, but it only handles a certain level of density. To support a higher density level of traffic, which we would anticipate on one of these platforms such as Uber Elevate, for example, we do need a solution for traffic management and collision avoidance. There's a nice analog to the automotive sector, which is the V2X connected vehicle technology.

If each one of these vehicles has its own telemetry package, you can come up with ways that you could regulate where the aircrafts are relative to each other to create basically highways in the sky to do traffic management and collision avoidance. That couples really nicely to AI because ultimately if you want to have a really safe system, you want to regulate the traffic that way. When you're regulating the traffic that way, you're also creating more efficient use of the airspace, because you can couple more rides in per unit time. That is key to making a system like this work downstream.

The initial flights, of course, could be traditional pilotage, traditional air traffic control but anything that you would need 5 to 10 years down the road, you would have to implement one of these systems. The government obviously has to be working along with that to make that happen. It's something which inherently has a software fix to make the system operate more efficiently, so I think that's kind of exciting if you think about it that way.

Then another thing to understand is that if you're using electrical energy to power your transportation system, whether that's a hyperloop, whether that's a metro, or a battery powered aircraft, you also want to get your power off of a smart grid that comes from renewables. What's so nice about a system like a hyperloop is that because they're designed to be above ground, you can use the actual infrastructure component of it, which is the tube to mount solar panels to generate your own energy, which can be used to power both the hyperloop as well as the neighboring cities nearby.

When we think about all of these infrastructure solutions, we want to think about the smart grid and we want to think about how that production of energy from that existing infrastructure could be used to power both the aircraft elements, as well as the transportation on ground structure elements. This is something which if you're not in the business of producing large quantity of things you may not think about, but the traditional model that most companies have operated under, whether they're making cars or they're making jeans like the Gap, is this linear economic model, where you take your raw materials, you go through production, you generate a lot of waste and you just throw it out. What companies are trying to shift to now and pretty much all sectors is the circular economy, which is that you try and reuse as much material as you can in the next time around in your cycle, or you can find other industries that need your material, that they can use in their own economic models.

That's very difficult to do when you're living in a vacuum, but if there was a way to combine that, either in a municipality, in a state, or countrywide, we can make that work. A great example I can give you is that, because I'm a big fan of craft beer, is that a lot of craft breweries, at least in Southern California, they actually take their basically hops, not hops, but barley, which is no longer used and they give that to local farms to feed the animals there. That's a way of thinking about how what you're doing and how it can be coupled to that larger economic model so that people can share and reduce basically the amount of waste which is produced.

What is becoming a really important, huge portion of our economy is mobility as a service. Mobility as a service obviously refers to on demand, but it refers to the entire end to end journey, so the big cruise in the middle, as well as the first mile and the last mile. That is ultimately going to be coupled to software applications, which can give you your scooter, to your Uber, to your train, to your airplane, back to your final destination. That's the only way that we're going to be able to support the density of traffic that we expect to see in our global society, which really does couple nicely to where we're headed.

Commercial Space Travel

What comes next in the future of transportation is actually very exciting and it ties into my background, which is commercial space travel. These are two examples, one just north of here, one just south of here. Commercial space travel is already happening. These companies are funded by private capital, the company on the left, Blue Origin, has Jeff Bezos as its head and obviously Jeff Bezos has Amazon, so he has a lot of his own money and he was funding development of launch vehicles to take people on suborbital trips, which could just be joy rides or they could be used for supersonic travel, or even hypersonic travel between destinations on Earth. Then eventually of course, they're working on a launch vehicle to take people off of Earth.

On the right-hand side is Virgin Galactic which is located in Los Angeles, they had demonstration flights recently where they're taking people up so that they can become astronauts or commercial astronauts, but it also couples to a potential for travel which enables much, much faster travel in an aircraft sense over land masses. They too will be shifting over to a model which enables exploration of space.

In terms of space travel to low earth orbit and space travel beyond low earth orbit, those are also under development as we speak. Typically, government space programs are the domain of deep space travel, the image on the left here is actually the Orion vehicle, which couples to the SLS vehicle, which NASA is developing. Orion vehicle's entire purpose is either to send people to the moon and then bring them back for an Earth re-entry, or to Mars and bring them back for an Earth re-entry. There is also a commercial aspect of this which is SpaceX, they obviously have their Dragon capsule, which has the ability to go up and back and forth to the international space station, but they too would like to support lunar missions, as well as Martian missions. In my opinion and I think in many people's opinion, commercial space is kind of the new sector of the economy which really hasn't been tapped into yet.

I was just at Blue Origin last week, the Blue Origin business model is basically to facilitate infrastructure in space so that people can have companies in space. It's hard to say what those companies would be or what those companies would do- there's mining, there's power, there's communications, but it's kind of up to all society to be able to come up with what that use case is for the commercialization of space. There are companies now who are starting to provide that transportation infrastructure to make that happen because you can't have a business model or an economy in space unless you can actually get there at an affordable cost. I think that's really exciting to think about.

Self-Driving Vehicles

This comes back really nicely to self-driving vehicles. I have never been too fascinated by self-driving vehicles because every single vehicle that is designed to land on the surface of the moon or the surface of Mars, for example, has to operate autonomously because the time delay is too long between Earth and the final destination, but self-driving vehicles are already being used on the surface of Mars. I think there's a lot from a software perspective and from a diagnostics perspective for hazard avoidance that we can couple to Earth-based technologies and there already have been.

Another area which is key for the future of human exploration of space is actually to use artificial intelligence coupled to robots to do a lot of the menial tasks, the difficult tasks and the risky tasks, whether it's repairs on orbit, whether it’s assembly of space stations and vehicles on orbit, or even to do menial tasks on the surface of the moon or the surface of Mars. In this area, there's a lot of robotics development at NASA to create something called robonaut, which is intended to do all of these menial tasks in space. One of them is actually already on the International Space Station, but it wouldn't make any sense to have human beings, for example, assembling facilities on the surface of the moon or the surface of Mars, because the radiation environment is so bad that you'd only want to minimize their exposure.

Even this is something which is under development for this particular use case in the space program, but I think there's a lot of technology which is being developed in AI for use in private sector applications, which directly couples over to what would be needed for a future human colony on the surface of the moon or the surface of Mars for example. Even though this doesn't couple to transportation, when you do set up those first colonies, you really have to think about all of the infrastructure that's required, you also have to think about the human condition of people existing in a distant location for many, many months without being able to return home. Then right after you have the first human colonies on the surface of Mars, presumably you would have the first cat colonies on the surface of Mars. I actually found this on Twitter, somebody tweeted this to me because they know I like cats, and I thought that was adorable.

I think it really should leave you with the question of, would you yourself want to live on the surface of Mars? Even though it might sound a little bit pie in the sky, I'm pretty sure the first people will get to Mars in the next 20 years. I'm pretty sure people will go back to the moon within the next 10 years. If all of these models actually happen, we're able to access space in a lower cost fashion, we start to actually set up permanent habitats in these locations on the moon and then eventually to Mars. This actually can become part of your reality, which is interesting and it all couples back to the development of the transportation system to get you there. Think about that in the context of how you can impact the space program if that was something you're interested in.

To wrap it up, and I'll take questions, I think that transportation has largely focused on the hardware side of things because that's typically been the domain of automotive and aerospace engineers, but nowadays, because mobility as a service is the next big thing, it completely couples to software development, it completely couples to the need for everything to be on demand and being more efficient and optimizing things actually will lower cost and reduce everyone's carbon footprint.

Questions and Answers

Moderator: I'll ask the first one, when will we see vertical take-off, landing air taxis?

Sengupta: Our subscale aircraft fly, so our plan is in the next 18 months. As we're raising our next round of capital, we'll build a full-scale aircraft, that will be for technology demonstration purposes, not to carry people. There's a whole FAA certification processes that you have to go through, you can get your experimental certification about a year, but I would say that we would actually start to see aircraft that carry people for commercial purposes in the next five to eight years or so. Everyone could take a helicopter now if they want to, but it's so expensive, this is going to be a lot cheaper.

Participant 1: You were talking about the density of these aircrafts in the air and V2X being used as the platform for making sure no one hits each other or something of that sort. I do know that right now that there is the current one for jets that actually looks at trying to warn pilots about imminent collision. Why are you not looking at using that system and augmenting it with V2X, then going directly to something that's brand new?

Sengupta: ADS-B is going to be required by the FAA by 2020 in all aircraft, right now it's optional. All commercial jets, all expensive aircrafts have ADS-B which uses GPS. Because the density is going to be relatively high, that probably won't give you the level of precision that you need to be able to support that level, that density of traffic. There are already highways in the sky, which are called Victor airways, but when you're on an instrument flight plan, they have to create an airspace bubble around you, which is a couple of nautical miles. To support the density of traffic that you would need to see for one of these things, that has to be shrunk quite a bit, so you don't want to have a person in the loop controlling that location necessarily, and you need to have something which is more precise.

It's probably the case that using an onboard lidar system coupled to the output telemetry coupled to the software architecture on top, that would be a much more precise way to support an even higher density of traffic. We are already using ADS-B and it does tell you “traffic, traffic,” is what it says to you as the thing gets too close. Even your GPS will tell you if you're too close to terrain, it's like terrain warning, you're getting too close if you're flying a little airplane like I do. It's the first step and it's already being mandated by the FAA, which is great because it actually makes things so much safer, because literally when you're flying around, I mean you can have another airplane which is like 500 feet below you. I mean that's how close you actually have to fly.

When I do like instrument approaches in Long Beach in my little single-engine propeller airplane, there's JetBlue right in front of me. It’s almost antiquated and archaic that it's taken us till now to get to that point. But I think this couples so nicely to making it even more precise and it just makes a lot of sense and right now air traffic control is run with a human being in the loop, and those poor people are under so much stress during the government shutdown, and I experienced this firsthand during the government shutdown, the poor air traffic control workers weren't being paid and you could just tell that were cranky, they were snappy and they actually made mistakes because of that. Being able to take that human being out of the loop and have them as a backup, but not as the primary, I think could make things a lot safer as a result.

Participant 2: I think a lot of these that you're discussing are really cool. I like "The Jetsons" and Star Trek. I'm just wondering with the larger goals of mobility and the environment and trying to help people, is there research or thought being given to providing mobility for people in areas of the world where they don't have a lot of infrastructure, and ways of maybe aircrafts that are very inexpensive, don't use a lot of energy?

Sengupta: I think that the only way to reduce CO2 footprint is to be able to go off of grid-based power. For those situations, having a hyperloop system, which is all grid-based and goes over long distances, is kind of key, but I do agree with you though that there also needs to be a change in our mindset. I would hate to see any of these systems being used to facilitate people who want to go between LA and Vegas and see a show and come back the next day, because that doesn't do anything to address reducing their individual carbon footprint, so there can easily be a misuse of this technology to just facilitate people who want to live that kind of jet setting lifestyle. For people who live in remote locations, the right way to do it would be hyperloops on the ground.

LA is a great example of a decentralized city where everybody lives in different neighborhoods, only some people actually come into Downtown Los Angeles to work, most people actually work elsewhere. I also think we should also challenge our idea and I think everyone does this in the software community, of do we even need to go into the urban centers? Can't we just work remotely? Can't we just be distributed teams as a means of reducing our carbon footprint? There are so many things that we should think about as a means of reducing our carbon footprint related to transportation, which coupled to VR for example, let's have virtual meetings and real VR like "Ready Player One," instead of driving into work, I'd be fine with that. I would say for your particular question though, ground-based travel actually makes sense to use a hyperloop over long distances.

Participant 3: Hi. Several times during your talk, you mentioned human backup to software. One of the things we do know about humans is they're very bad at situations that require nothing happens, important, nothing happens, important. We know from nuclear reactors, from even the Tesla accident. How do you cope with that problem and deal with that issue?

Sengupta: The human backup is actually mandated by the regulatory bodies right now, because as you go through your operational certification, you'd have to have that human being in the loop. That doesn't mean that that's a good solution, it isn't a good solution, but that actually comes from the requirements from the rail administration, the agreements that they've made to do self-driving car certification. Then obviously even when you're flying right now, commercial, you're pretty much flying on autopilot for everything except for the take-off and the final approach, so that actually comes from our regulatory body.

I'm not saying that that's the right solution, but one could also make the argument that, “Uh-oh, we had this big problem with the 787 MAX and the autopilot getting out of control,” and it was the only human beings who could have gotten out of that situation. There may have been a new problem where they couldn't get it, I don't know the full details, so that you can come with an example there where that is important, but in general, I think that isn't the right solution to make things safer, probably more rigorous testing is the way to make things safer. I only say that because we have a requirement to do that from a legal perspective as a company when you're doing these things.


See more presentations with transcripts


Recorded at:

May 22, 2019