We Solve for X: Andreas Raptopoulos on physical transport


[MUSIC PLAYING] ANDREAS RAPTOPOULOS: So this
is a very young idea. We created it just this summer
at Singularity University. But we stumbled upon something
that we believe can be a paradigm shift. So we want to solve for
physical transport. 1 billion people in the world
today have no access to all-season roads. That means that one seventh of
the Earth’s population is disconnected from all social and
economic activity for some part of the year. You cannot get medicine
to them. They cannot get consumer
goods. They cannot get their goods to
market in order to create a sustainable income. If we look in Sub-Saharan Africa
for instance, 85% of roads are usable in
the wet season. Only 12% of roads are paved. The quality of roads
is going down. It’s worse today than
it was 20 years ago. Investments are being made. But at the current rate it’s
estimated it’s going to take Africa 50 years to catch up
on road infrastructure. In the developed part of the
world, we have very elaborate, very complex, expensive
infrastructure. In the US alone we have more
than 4 million miles of road. Now mainstream thinking suggests
that the developing world should go through the same
stages of building road infrastructure we
went through. We want to question that. We’re asking is there a
technology that can help us put a solution today
that can help those countries leap frog? Imagine this scenario. You’re in the maternity
ward in Mali in 2014. And you have a newborn in need
of urgent medication. Today you would place a request
by mobile phone and hope that the medicine
arrives on time. Sometimes it takes hours,
or even days to transverse bad roads. We believe we can deliver it
within hours or even minutes with an electric autonomous
medical supply vehicle such as this. This car has a small payload,
about 2 kilograms, over a short distance, about
10 kilometers. But it’s part of a larger
fleet that forms an ultra-flexible automated
logistics network that spans a bigger region, maybe even
the entire continent. We call this the Matternet. The idea behind the Matternet
is set up a network for transportation of physical goods
that is based on the ideas of the internet. The enabling technology is UAVs.
We want to harness all the great work that has been
happening in academia in the open source community and build
a platform that can allow us to do point-to-point
delivery, decentralized peer-to-peer just like
the internet. The second vital ingredient of
the network is the automated ground stations we use. These are point stations on the
ground that the UAVs fly in and out of in order to swap
batteries and fly further or exchange loads. The third is the OS that runs
the whole network, that optimizes routes, optimizes the
flow of vehicles and goods through the system. It optimizes for weather
conditions and guarantees the security of the system so
we can guarantee to the authorities that it’s not used
for illegal purposes. Let’s look at the technology
in some more detail. The vehicle– that’s a conceptual
design of it– is a hybrid platform. It’s based on multicopter
technologies, such as this copter that we have over here. So this has vertical takeoff
and landing. It is compatible with our
concept of minimal infrastructure. It has a wing architecture that
gives it the lift to drag ratio comparable to that
of a helicopter. We don’t need much. It flies at the speed the
40 kilometers an hour. It can fly in winds up
to headwinds of 30 kilometers an hour. It is steered electrically. So a flies in the rain. And it flies at an altitude
of about 400 feet. It’s optimized for that. It could to fly much higher. But that seems optimal from
a regulatory perspective. And we want it to be
friendly and safe. So we will deploy a
parachute if we sense mechanical failure. In the heart of the vehicles,
we have the interchangeable packet and battery. So when this vehicle lands into
the ground station, the packet is automatically
swapped. We have a 10 liter packet that
can carry 2 kilograms. Let’s look at the
ground station. That’s a conceptual
design of it. The main job of the ground
station is to create a safe harbor for the vehicle to
go in and land safely. The vulnerability of the whole
mission is the docking part where you exchange loads
and batteries. So we want to guarantee a safe
environment for that. This happens in the upper part
here, where it’s a sphere. The vehicle goes in and lands
and docks safely. The ground station acts as a
commercial exchange hub. So someone comes into
the ground station. They insert the packet. It goes up, waits in the dock,
a UAV takes it to the next ground station. And it goes to the
destination. We fly mainly using GPS,
sensor fusion, IM unit pressure sensors, and
magnetometers to guarantee that we can fly even if we
lose GPS for a while. And the idea is that we upload
GPS waypoints at the departure ground station. When we are in the vicinity
of the destination, the destination ground station takes
over with ultra-wideband sensors [? the ?] camera. It brings the vehicle
in the harbor. And then we dock it with
millimeter accuracy. So we think we have
a big idea. And we have the technology
to make it happen. The next big question
is what’s the cost? Well, it turns out that the
cost to run what we call a mission, which is 2 kilograms
over 10 kilometers, is $0.24. Can you believe it? When we saw that, we understood
even deeper in our hearts that we are sitting
on a paradigm shift. Let’s look at the
cost behind it. The cost of the vehicle
is $1000. We can believe we
can achieve it. We have industrialization. We have a few trends
playing in our way, like battery costs. Maintenance is minimal . Because it has only four moving
parts or six moving parts according to the
number of propellers. And it can have a lifetime
of 10 years. Cost of the ground
station, $15,000. Again lifetime of 10 years. This will drop substantially
as we go through more iterations. Now cost per mission, two
kilograms over 10 kilometers for the vehicle works
out to be $0.03. Battery is $0.09. We need to replace it every
600 [? sites. ?] It’s conservative estimate. That cost will go down
significantly. Ground station, $0.10. Energy for flight, the counter-intuitive bit, only $0.02. If we want to run on solar– the plan is to run in places
that don’t have great energy– we need to triple that
component, the last one. So $0.24. What can we do with this? We started looking at the
problem of HIV/AIDS delivery in Lesotho. Lesotho is a very
particular case. It’s a country which is very
mountainous surrounded by South Africa. They have an AIDS epidemic that
is threatening the whole country with extinction. Over half a million people have
the HIV/AIDS virus there. 70% of them live
in rural areas. So we have 400,000 people who
need to be tested twice a year to continue receiving the
medication or make sure that they trace the virus early. They try to organize this
collection of blood samples through 152 clinics, 19
hospitals with labs. They cannot get enough people
to run the samples from the clinics to the labs. And recently we have been
reading about the president’s office initiative to use
ponies to do it. Well, we have pony
UAV to do it. Blood sample from clinic to
hospital is a problem. Results from and medicine
from hospital to clinic. We set up a mini Matternet. And we wanted to get an idea
around cost. We looked at the region of Maseru, 47 clinics
there, six labs, 138 square kilometers. The cost to solve the problem
there is less than $1 million today. $1 million according to the
World Bank is the cost of constructing a 2 kilometer
one lane road. Can you believe it? This is we believe the power
of a new paradigm. Now when we realized what we are
creating, we thought, OK. If it’s really a paradigm
shift, there ought to be more x’s. There ought to be other places
where this can be working on. I mean the internet doesn’t
know boundaries, right? So we focused on the opposite
scale of the problem, not where we don’t have
infrastructure, where we have a lot of it and it doesn’t
quite work. So half a billion of us
live in megacities. China is adding a megacity
the city of New York every two years. It’s going to be something that
this future of people living in the cities is only
going to be more and more evident for us in the
next few years. Let’s look at Sao Paulo,
21 million inhabitants, 6 million cars. It’s typical for people to get
stuck in traffic three to four hours a day. The estimated cost to the
government of congestion is $2.3 billion. The economy needs to
flow of course. In order to move goods around,
they are using 250,000 motoboys courier bikes. There is a huge social
problem around it. People are killed on the
street every day. There is 500 people on average
a year killed in relation to motobike activity, either
pedestrians or bikers themselves. More than that, it’s not
a scalable solution. It’s not a modern solution for
a country and a city that is having explosive growth
like Sao Paulo. We believe we can solve
the problem better. We can plant stations at the
rooftops of buildings or even the sides of office buildings,
you know. We can have an infrastructure
that works 24/7, around the clock, just like the internet. So that’s our vision. Matternet will do for physical
transportation what the internet did for the flow
of information. We have this concept called
the bandwidth which is material transported through the
system per unit of time. And we think that’s going
to go up dramatically. We have a part of the economy we
think that is not currently expressed with current means
of transportation. We think if we add the layer
of transportation that is dematerialized between the
internet and the road, we can create something that
is a new paradigm. Are there any challenges? Of course– legal, business. There’s going to be public
perception issues. It’s called changing
the world. Is it worth it? We think absolutely. Thank you. [MUSIC PLAYING] MALE SPEAKER: Let us define x. x is a solution, a solution to
a seemingly insurmountable problem like climate change
or cancer, one that affects the world. But what if we redefine x as a
challenge, an opportunity for radical thinking, a chance to
light up the world with breakthrough ideas and cutting
edge technology, the stuff of science fiction that just
might fly after all. Solving for x requires wonder
and imagination and a vision to build seemingly impossible
solutions to the world’s biggest problems. Solve For x. Moonshot Thinking.

56 thoughts on “We Solve for X: Andreas Raptopoulos on physical transport

  1. Simply remarkable. But is the 24c/mission still too high a price? In a developing country with limited resources, its not often a choice between having a road or a UAV system, but having a road at all.

  2. In an ideal world, this would likely work well but these little vehicles would be susceptible to birds, rocks thrown by little boys, great targets for toy guns or even hunters with shotguns.. if their cargo is valuable, they could be hijacked by nets, other remote controlled planes, possibly lasers, pirate signals, they would also have to navigate obstacles such as poles, towers, wires, jets.. also the ground stations would be targets for crime.. add the price of security to the system???

  3. Andreas, congratulations my friend! The presentation was awesome and It is admirable the progress you all have made. I am really happy. Keep moving forward!

  4. Place something worth 15000$ in the middle of nowhere in affrica – it won't stay there for long…

    Also 2 kg is not much load – apart of medical usage it will not be usefull a lot.

  5. What proportion of vehicle traffic is due to payloads of 2Kg or less?

    Are there any historical or sociological analyses which cast light on the impetus for transport networks being anything other than heavy-goods distribution? e.g. canals in Europe in the mid to late C19, followed by railways, then highways.

    Interesting enough idea, but suffers from an inherent problem of the "moon-shot" approach: an assumption that challenges faced by societies merely lack a technological solution.

  6. I think it would work well if the remote destination kiosk was simply an inexpensive recharge station or just a landing platform, something hard to steal. Communication can be done by phone. The flying carrier might be configurable to interact with the users directly rather than through a kiosk station. I agree with EffUsh that leaving large investments out in remote places is a good way to loose that investment. We're dealing with humans here, with all their strengths and flaws. 🙂

  7. I don't think it will work… I'd guess some people will try to intercept these drones, maybe shoot them down -> instant drugs and probably some materials to resell….

  8. A crazy sounding vision becomes normality the day it is running. (like the car, phone, internet, …) Go for it @matternet !!!!

  9. SalsaTiger, the same could have been said about the Internet (which is used today by criminals) but it would be hard to image the world today without it. Sure, lots of questions remains about the use of drone (ie. airspace safety) but it has real potential.

  10. i rather like the pony idea. people have raised work animals for thousands of years, are familiar with their health and needs. what if we implant chips in pony-brains to GPS-steer them to places?

  11. The problem is the vehicle is valuable, and hacking takes the form of simple stealing.
    People are great at stealing, and the hacker community will be huge. The market will be persons who want to deliver 2kg of explosive onto their enemies.

  12. To all of those saying we shouldn't do this because it would just be used by criminals:
    UAV systems like this are relatively simple to setup. This isn't so much a product as it is an idea that can be implemented by anyone.

    DIY UAVs like this already exist to an extent, just look around youtube. Everyone is building them, and running them on continually improving open source software.

    This isn't a question of should/shouldn't we. It's an inevitability, all we can do is brace ourselves.

  13. Something i find funny about this concept, is that it could to an extent implement a high lag but high bandwidth replacement for the internet.

    Each of those UAVs could carry 2 KG of HDDs or Flash drives. It would transport multiple terrabytes of data at a time.
    Compared to the internet – the UAV would probably arrive much faster than you could download that amount of data.

    I remember someone doing this with a homing pidgeon, strapped a flash drive to it's leg, compared to download speed.

  14. In short, Matternet costs per ton-km are 100 times greater than current costs in Central Africa (where there are the highest costs).

    Matternet moves 2kg over 10km at a cost of US$0.24. This is equal to US$12 per ton-km. Current costs in the most difficult parts of Central Africa are US$0.11 per ton-km.

    How can Matternet ever be competitive? Even with exponential decreases in costs?

  15. The problem of large remote stations can be solved with the spray on wifi antenna team is presenting in X solve. The tech drastically increases the range reducing the number of stations required. Another tech of flexible printed electronics can be used to increase the range of flight

  16. hook it up with ebay if could shop online in the middle of and receive your package an hour later imagine that

  17. Shooting something like this down would be difficult. At 400 feet and 40kph and I am assuming relatively small, firing a weapon such as a rifle or machinegun would be a very hard target, not impossible, but very difficult. Now, some will say rockets, sure, in some places in the world where these are common you could use those, but in a majority of the locations of the world those are rare. So I think shooting down one of these would be difficult.

  18. The problem is that flying between point A and point B is not like sending a signal along a wire….unfortunately.

  19. 20 years folks were asking "what's internet ? " Is "Matternet" next ?
    As a old African and Northern Canada air type wherein primeval geography still rules, I say give a shot. In my view the big payoff would come when this stuff advances to bigger payloads and greater range. A UAV with , say, a 100km range and 500kg payload would change things greatly ( for the better)

  20. you know countries where sick people are treated have a much slower population growth than those who do. you're probably just trying to be dramatic by saying that but even if you are at least realize you are completly wrong. if you want people to stop having 8 kids, make sure that 6 of them don't die before their parents

  21. In that case, you try again. You can have reliable transport even if each hop is unreliable. It does require you to have some redundancy. This redundancy is probably less expensive that than the cost of protecting the drones.

  22. A charrette in Senegal can cary a few people 10 km for much less than $0.24 /kg. If you are moving crops to market, you cannot afford $0.024 /kg-km for hundreds of kg. This is certainly useful for small valuable products. The fabulous thing about information technology is that there is 'plenty of room at the bottom' – a modern transistor is order of magnitude smaller and faster than a generation ago. But you can't do that with food stuff and other artifacts needed to live.

  23. Death is by its nature the end of a living being, so not taking your own death personally is something of an oxymoron. It is not your role to determine who should be saved. Get off your high horse and just live with your fellow man – down here in the mud. You may actually like some of us.

  24. It is easier to destroy than to build, dying is easier than living. This has been the case since the beginning of life, and yet here we are making progress. Yes, there are way to destroy inexpensive, unarmored drones. So what? Who wants to dedicate their own life to skeet shooting drones?

  25. I tend to agree with your concerns about a 2 kg limit. The best counter argument I can come up with is a swarm of locust. A swarm of robo locust could descent on field of crops and carry the produce to market. This is a couple of generations beyond this matternet, but it seems physically possible. Rather than scaling up the size of the load, you can scale up the number of loads. This may be useful for a wide class of projects. But eventually, you will face the need to lift heavy things.

  26. People who explain that they they are profound is not in fact profound: they are pretentious. You have no idea who I am and yet you call me a worthless person. You are a coward hiding being self-importance and a pseudonym. Go ahead, respond as The Great Noleander and prove my point.

  27. Just like burglars and criminals today.

    But that is not a reason to stop this development and it will continue.

  28. I do not think this is crazy at all only people that are not forward thinking can see the practicality and brilliance of this.

    Quite clearly you see the brilliance in this. But there will always be those that say man will never fly. There will never be billions of communications nodes that we now call the internet.

    Ignorance breeds doubt.

  29. It's not like a plane or helicopter that you hear from far away. You only hear a drone when it's 10 yards from you.

  30. The problem with such a great idea is, the countries they deploy them in. They are, for the most part, isolated, and tech is not going to go with their culture. Someone is going to decide that it's witchcraft, or some warlord will shoot them down, or, the people will just tear down the ground stations because they don't understand it. As much good as this would do, it won't work for more than a few months, that even if they get it fully operational.

  31. Yea, also some drones must have 3d printer, and laser charging. Allowing them to stay in the air constantly and build stuff 😉

  32. With LENR/cold fusion coming out soon this will be right on target to get on the bandwagon of a technological revolution via, LENR. It will happen, if not by this guy someone else. But they seem competent enough to get it done.

  33. If they wanted to get that stuff, they would be robbing couriers right now. Much easier. Besides: there are many countries where getting a gun isn't as easy as in the US..

  34. A great idea – but it relies on there being power for recharging at the landing points that it goes to. Is that realistic? I don't know.

Leave a Reply

Your email address will not be published. Required fields are marked *