Thought Leadership

Autonomous flying drones could prove to be a lifesaver

Rwanda is located in central east Africa and is known as the “Land of a Thousand Hills” due to its many peaks. The country is ranked 149th in the world in terms of area, but its rugged terrain and limited road infrastructure means it can take hours to reach a rural village by ground transportation, even if it’s not far away.

Zipline International Inc. is a Silicon Valley-based logistics company that designs, manufactures and operates autonomous electric drones to deliver lifesaving medical products on demand. It has partnered with the government of Rwanda to airdrop medical supplies such as blood, and the results are promising.

In his presentation at the recent Digital Twin Summit in Chicago, Illinois, Hersh Reddy, general counsel of Zipline, discussed how his firm is using autonomous flying drones to make it possible and affordable for Rwanda to provide reliable access to essential medicines for their citizens.

Zipline has a three-fold mission:

1) Improve health – Connect patients with what they need when and where they need it
2) Reduce waste – Make the most of scarce resources by centralizing and streamlining inventory
3) Transform logistics – Enable just-in-time delivery and demand visibility down to the last mile

Reddy gave the following example at the summit: A woman’s life is dependent on getting a blood transfusion if she gave birth and suffers from postpartum hemorrhaging. The challenge is getting that blood to her when she actually needs it, because there are many blood types and blood itself has a short shelf life. During the rainy season in Rwanda, it could take up to four hours for a truck to deliver the blood to a remote area. It could take as little as 15 minutes using Zipline’s autonomous flying drones, which could make a huge difference in the patient outcome.

Zipline uses NX software because they consider it a “force magnifier.” Features such as NX Journal enable them to get a total picture of the aircraft so when they’re manufacturing for a use case like Rwanda, the company can get an understanding of the economics over the operating lifetime of the aircraft, a critical consideration. Each kilogram of weight savings increases the range of the autonomous electric aircraft by 5 percent. Using NX Journal enables Zipline to create a custom, bottom-up Bill of Materials that gives a detailed mass buildup, which helps the firm optimize the aircraft. Analysis features like check wall thickness make it fast and efficient to optimize parts.

The ultimate vision is to be able to deliver any essential medical product to one of Rwanda’s 11 million citizens in 15 to 35 minutes. That could transform healthcare because it would mean autonomous drones could be used to deliver life-saving medical products anywhere in the world where ground travel is difficult or impossible.

We hope you enjoy this video.

In upcoming summit talks, you will learn about the synergies that exist between the automotive and aerospace industries and how digital twin technology weaves them together with the digital thread; autonomous/electrification issues from the perspective of an automotive and aerospace customer as well as an industry thought leader, and next-generation/additive manufacturing issues from the perspective of an aerospace and automotive customer as well as an industry thought leader. You will also be able to watch a 45-minute question-and-answer (Q&A) session with the speakers. Keep an eye on this blog in the coming weeks as we provide links to them. 

About the author
Indrakanti “Chaks” Chakravarthy is the Marketing Programs Director working in the Americas zone at Siemens PLM Software, where he is responsible for marketing programs covering the aerospace, defense, federal and marine industries. Chaks has spent the last 30 years working for a number of companies in presales technical, sales and marketing areas of the PLM portfolio in India, Japan and the United States. He has a mechanical engineering degree from Jawaharlal Nehru Technological University in India.

VIDEO TRANSCRIPT

TOM MAURER: I would like to now invite Hersh Reddy over here, general counsel of Zipline, to talk about transformation of another type for you.

HERSH REDDY: Thank you. Can you guys hear me? Great. I’m just getting over a cold, so I apologize if there’s a little coughing halfway through this. Let’s jump right into this.

Zipline is a company that manufactures fully electrical aircraft that are also autonomous, and we use them to do medical deliveries in some of the hardest to reach places in the world.

We basically have a three-fold mission. Using these electrical aircraft essentially allows us to connect patients in a more efficient way to the medications they need. It also reduces waste because introducing an autonomous vehicle into a supply chain actually allows people to keep less medicine on stock. You have less stock-outs. You have less expiration of medicines and inventories. It also allows us to completely transform the way people are thinking about delivering healthcare in that last mile.

This is our current aircraft that we are operating. We’re actually about to release a new one. It’s quite a small aircraft and it only carries about 1.5 kilograms. That doesn’t sound like a lot to a lot of people, but 1.5 kilograms is actually enough payload to deliver a unit of blood, including the packaging that it goes around the medication or the blood.

We fly at about 100 kilometers an hour, so that actually enables us to make the deliveries extremely quickly. We can get an order from a doctor, we get the medication or the blood to the patient, sometimes within 15 minutes, and the range is about 160 kilometers.

So I’m going to actually show you a video, because I think images do a better job of showing the impact in the system.

VIDEO: Rwanda is known as the land of 1,000 hills. It’s a very difficult country in terms of topography, and many areas are hard to reach. Traversing that terrain to get to a rural village can take hours, even though it’s a very short distance.

Today, if a mother is giving birth and suffers from postpartum hemorrhaging, her life is dependent on getting a blood transfusion. There are many different types of blood types. They have a limited shelf life, and the challenge is when you need blood, you need it immediately. But it is unpredictable which geography that will occur in.

Hospitals that are in more remote areas can have a very hard time getting enough reliably supplied blood.


The time to order and get the supplies by truck is just too late. There’s no way that we can overcome the time that it takes to get anywhere during the rainy season. People would say, okay, during the rainy season, we’re just not going to make it.

The best-case scenario would take at least four hours. But with Zipline, we are looking at cutting that four hours to something like 15 minutes. This is a team of Rwandese who came together with Zipline to groom the project and to make it to happen in Rwanda. Their ultimate vision is to put each and every one of those 11 million citizens within a 15 to 35 minute delivery of any essential medical product that they could need, and that’s a revolutionary idea for any country in the world.

The use of drones to deliver life-saving medical products can overcome the lack of road infrastructure. We need to let our imaginations soar. When you have a commodity rare in supply – that might be an unusual blood type, or something like a rabies vaccine – and you need it immediately, the idea that we could use a drone to be able to deliver this to the country is absolutely brilliant. There’s nothing else that could do that in that way.

It is like we no longer go around hills, but we fly over the hills.

We pulled together a team from places like Boeing, SpaceX, Google, Lockheed Martin.

It’s an honor for Zipline to partner with the government of Rwanda.

We place a hub next to an existing medical warehouse, and it instantly enables that warehouse to make hundreds of deliveries per day to any location within range. The doctor sends us their order by text, by phone call, whatever’s convenient. Somebody identifies the need and then prepares the package, puts it on a drone.

Then we take that vehicle, and we put it through pre-flight tests. It’s launched. It flies automatically out to the clinic.


The path has been already calculated. It’s in the computer. 

Then I’ll receive a text message saying Zip is two minutes away; please walk outside. It drops the packet. Somebody goes and collects it. It turns around, it flies home and then we get it ready for the next delivery.

Our service not only makes it possible, but also affordable, for countries to deliver reliable access to essential medicines for all of their citizens. What you are looking forward to see here is saving lives.

This isn’t a small step forward; this is a transformational change to how we provide medical care to people across the world. [End of promotional video]

HERSH REDDY: That’s a summary of the system. I’m going to show you just – so this is actually a map of Rwanda. It’s not a very large country; it’s approximately the size of the California Bay Area.

You can see from one distribution center, we can reach almost 90 percent of the country. We’re about to add a second distribution center. Once we do that, we’ll actually basically, essentially, be able to cover every part of the country.

The main product we’re delivering right now is blood, and we’re working with the Health Ministry to expand that into other types of medication. The reason blood is so hard to maintain, or I should say the reason why blood supply chains are so hard to keep intact in places like this, is because blood requires refrigeration. Blood also requires you to get the material ahead of time. It’s very difficult to essentially have someone with a rare blood type require blood in a remote place, and then to find a donor at that spot. Blood almost is the most difficult of all these different kinds of things you might need to keep on hand as a medical professional.

Solving the blood problem was the most urgent problem, and now we’re going to try to take the same transformational things that we’ve brought to blood logistics into other types of medication.

If you look at – as I mentioned before, the range is about 150 kilometers. If you think about that, the aircraft doesn’t land on the other side; it has to fly and come back again. That gives us an operational radius of about 75 kilometers.

In that 75 kilometers, it doesn’t sound like it’s very much territory at all. But actually, it turns out that that territory is about 22,000 total kilometers. If you do the math, Pi R squared, that’s the area covered by a circle based on the radius of 75 kilometers.

In that 22,000 square kilometers, there’s actually something like 10 million people. Being able to service all those 10 million people with one UAS [unmanned aircraft system] is pretty transformative.

There is some capital investment you have to put up front to bring the launcher and to train the people who operate the system. But considering the service radius, it ends up being very economical; or I should say, it’s actually sometimes more economically efficient than the current ways that people do the same kinds of transportation with trucks, et cetera.

I’m going to show you another very brief video. This just shows end to end what the system looks like on the ground in Rwanda.

[Video plays] That’s a control tower over there that you see, and it’s actually built on top of the toilet facility we have at that base, so it’s kind of a dual-purpose structure, if you want to say.

The white building over there, which we’re now the on the inside of, is actually just a standard shipping container. You can actually see, as the video’s going through the corrugated metal walls, I mean this is just a standard shipping container.

We put a bunch of refrigeration gear in there so we can keep the blood cool.

We have all local employees that actually do the logistics portion of this: the packaging, and the operation of the drones. These are things that we can actually train locals to do.

This is a local, one of our guys, Abdul. He’s actually loading up the launcher. They go through some training. Now he’s launched the aircraft, the aircraft’s on its way now to the hospital.

This is a video at the hospital where it’s doing the delivery to show how what it looks like when the blood’s actually being received.

The autonomy on the aircraft is good enough that we can actually get the package right in front of the front door of the hospital every time. It’s not a big issue.

The bigger issue, actually, is that people get so used to this aircraft going by. Initially, we told people when you see, the aircraft coming please stay outside the drop range. But people see it come exactly on the same spot every day.

They kind of violate the rules that we’ve given them because, you know, it’s funny. Something is science fiction when you first start doing it, then people get so used to it because they see hundreds of deliveries every week. It becomes commonplace. That’s essentially what’s happened.

This is actually the user interface for the operators, and it shows the map. We have a map of the country. It shows where every single drone that we’re operating is currently when they are on flight, because we typically will have four or five in the air at the same time.

[Video ends] So that was just a brief summary of how that works. Let me give you some stats really quick.

Like I said before, we cover about a 20,000 thousand square kilometers service area from that one base. We can do deliveries just on time because the aircraft’s flying so fast, you don’t have to keep the blood on hand; you can actually order it exactly when you need it, which simplifies a lot of things.

It simplifies, for example, maintaining the cold chain. It’s much easier to have one big industrial freezer at a central blood bank than it is to maintain 100 or more smaller freezers at every single clinic that you might require blood at.

We deliver the blood so quickly that it’s received at the clinic side within its temperature range, so you don’t actually have to maintain freezers anymore.

On the other side, let me talk a little bit about how Siemens helps us to achieve this. We’re huge users of NX, and we have actually a pretty small team probably compared to most of their customers. We’re tiny; we’re about 60 people, approximately 40 engineers. The rest of us are in business, lawyers, et cetera.

Having tools like NX essentially is like a force magnifier. We can use tools like NX Journal, et cetera, to get a total picture of the aircraft. When you are manufacturing for a use case like in Rwanda, cost is a very significant factor. Obviously, safety is also. But if you can’t meet the safety requirements in just the bare minimum cost that’s required in that market, then you essentially have a product that’s meaningless, because you can’t actually use it.

If you think about the way the economics of these kinds of things work, you have to be able to amortize the cost of the aircraft across its operating life, the total number of deliveries that it can make. If the cost of the aircraft doubles, you’re essentially going to end up doubling the cost of the deliveries, too.

It’s extremely important that we do things like keep the weight down, because when you keep the weight down, that means you can use a smaller engine. It’s cheaper. It’s an electric vehicle, but people always make this mistake: just because you have an electric vehicle, that there’s no cost associated, that’s consumed, when you when you make a flight.

That’s not the case. Anyone with the laptop or a phone understands that there’s a certain lifetime that you can charge and discharge lithium-ion batteries. The battery technology we use is pretty standard in that way. There’s a certain number of cycles that you can do. We can replace the batteries, but even so, we have to again amortize the cost of those batteries across the number of deliveries we use that battery for.

Keeping the weight down is very key. We can use features like NX Journal to essentially figure out what our budget is for the aircraft. Then, when we violated the budget, drill in specifically into those portions of the aircraft that we think we can then go in and make more optimizations, we can say: okay, you’ve gone way beyond your budget, John, on the engine in this cell; you have to figure out a way to get it down. Then we can use other automated tools to figure out where, once we’ve gone too far on the weight, where can we trim it down and where we’ve gone too far with things like these check wall features.

These are pretty standard in NX, and I think people in different industries have used this forever, but it’s super critical for us because of the engineering team size, and because of the speed as well. We have to finish our whole design process, bring the aircraft out in an extremely compressed timeline for the same reasons I was saying before. We need to bring the hardware costs down, because obviously you’ve got the salaries of the employees. Everybody else working on the team that also has to be amortized against the project length.

Here’s another feature that we actually use a lot as well. .We do stress analysis for the aircraft, so this ties back into what earlier speakers were talking about in terms of digital simulation. We would find something like 90 percent of the faults in the aircraft in two forms of simulation, one which is done directly in NX, things like stress analysis, et cetera.

We can get that first order of magnitude, of approximation of where parts would obviously fail, given the design shortcomings. This illustration just shows that – it’s color coding that particular joint connection between the wing and the fuselage showing where there is too much stress on the material. The material where it’s lighter colored is experiencing more stress in the simulation than the portions where the color is darker. The engineers can go in and say, ok, well it looks like for that particular portion of the aircraft, I’m going to have to bulk it out a little bit more.

That’s one way of getting insights to drive and refine the engineering process digitally without us actually having to build prototypes. But there’s a second way that we do simulation as well, which is, once we’ve had the aircraft designed out, we actually have a hardware test bed, which is a complete replica of the electronics that goes in the aircraft – all the pieces that go into the autonomy. We have the flight computer, we have the GPS, we have the radios.

We built replicas of these aircraft – the brains of the aircraft, I should say – sitting outside of the actual aircraft bodies, but connected into virtual simulations, so these aircraft brains actually think that they’re flying. Then we fly them in the simulated environment, and essentially, what that lets us do is debug the electronics portion and the software portion separate from the physical aircraft. We run thousands of simulations. When we rev the software code, that can be done sort of decoupled from the physical aircraft, and it can be done very quickly.

Actually, the way we do our software development is pretty different from the way places like Boeing, et cetera, do it. We do it very similarly to the way people might do web development or apps development, in that we have engineers constantly updating the code, but that code goes into this test bed and gets many, many thousands of rigorous tests done on it, which gives us the assurance that the code’s actually going to be functional, but, at the same time, gives the engineers the flexibility to make those rapid iterations.

We think this sort of development for autonomy is going to be a key sort of engineering change, engineering process, if you will, which is bringing some of that web development, app development, engineering principles into this more rigorous space. There again, the takeaway is essentially simulation is key to that. Simulation and offline testing: out of the real world, into the simulation space.

This is some analysis that we’re doing for battery optimization, and I think this points the way to how we will be able to – well, hopefully, this is the vision for the future. I was talking about how we do the autonomy simulation purely in the virtual world. Here, we’re doing a simulation to figure out how to do our battery designs, where we are actually doing a simulation of the heat characteristics of the battery in flight digitally.

Previously, the way we would do this kind of thing is we would fly many prototypes of the battery in our prototype electric aircraft and essentially see how they perform. It’s very difficult to do that and to get sort the characteristics of every different environment the aircraft might fly in.

Rwanda is a very different environment, or, I should say, climate situation than the Bay Area. Having successful testing in the Bay Area isn’t necessarily going to give us a lot of assurance that we’re going to have the same performance characteristics in Rwanda. But doing this kind of thermal analysis within simulation allows us to change the environmental factors there to more closely simulate what the actual aircraft operating parameters might be in the real world, without having to actually send engineers there.

Here you can see our lithium-ion battery cells, which are those cylindrical objects. We’re essentially showing how the heat dispersion happens during flight, and you can see that there’s a portion of the cells where they’re getting way too hot. The heat, the internal cells, are so far away from the outside of the aircraft that they’re not getting the cooling that’s required.

Doing this kind of simulation right now – we still do the simulations where these kinds of simulations happen in isolation from other kinds of simulations, like the mechanical stress simulation, the software simulation.

I think the ultimate dream is to have the entire physical simulation of the aircraft, of all the aspects – the thermal, the stress and the software simulation actually flying the aircraft – all happen under kind of the same roof.

It’s a long-term dream, but I think we’ll eventually be able to get there with partners like Siemens. Siemens provides us a lot of the tools to do this the physical side of the simulation. What we’d love to be able to do is also bring in our autonomy piece into this, and then do a complete test of the aircraft inside the computer before we’ve even built the first real-world model of it.

So that’s my wrap right now. If you people have other questions, I’ll be on the panel, you can let me know later. Cheers.

Indrakanti Chakravarthy

Leave a Reply

This article first appeared on the Siemens Digital Industries Software blog at https://blogs.stage.sw.siemens.com/thought-leadership/2017/10/19/autonomous-flying-drones-could-prove-to-be-a-lifesaver/