Drones, drones everywhere - Zipline’s New Delivery Drone - Is 10 minute delivery worth the cost?
Welcome to the Robotics Roundtable
This is our fourth newsletter. In it, we take an article that piques our interests and discuss it from our unique perspectives. Sevy - the robotics hardware engineer, Connie - the robotics software engineer, and Sean - the social entrepreneur and marketing + finance dork.
This week we chose a video from famed YouTuber and NASA Engineer Mark Rober about Zipline’s new delivery drone. You can watch the video here: Amazing Invention- This Drone Will Change Everything. For us, the video acts as a launchpad for a larger discussion on how drones should fit into our lives.
If you enjoy our work, please consider sharing, visiting our blog, and subscribing to our newsletter where you can read more about our thinking, get updated on our latest events, hear about breaking robotics news coming out of Los Angeles, and follow our journey as we work to unite the Los Angeles robotics ecosystem.
Video Summary
The YouTube video titled "Amazing Invention- This Drone Will Change Everything" presents a revolutionary approach to package delivery by a company called Zipline.
Zipline has pioneered the use of autonomous drones to deliver critical medical supplies to remote Rwandan villages, demonstrating the potential of drone technology to transform logistics and healthcare. The drones, launched from a catapult, are capable of a 150-mile round trip, operating 24/7, and have successfully made half a million deliveries in six years.
Remarkably, their new drones are designed to be whisper-quiet, with the main drone flying at 400 feet to minimize noise disruption.
The video also explores the broader potential of drone technology for doorstep deliveries, suggesting that drones could replace traditional delivery vehicles, thereby reducing traffic, emissions, and increasing delivery speed. Furthermore, it speculates on the future use of larger drones for human transport, potentially serving as airborne ambulances.
The impact of Zipline's drone delivery system in Rwanda, where it has significantly reduced in-hospital maternal mortality, underscores the transformative potential of this technology. The video concludes by highlighting the inspirational effect of drones on the next generation, fostering a new wave of problem solvers intrigued by the possibilities of drone technology.
SEVy’s Corner
As a mechanical engineer and someone interested in designing product features, Zipline is beautiful AND scary. Last-mile delivery is a problem people are itching to solve because it is a GIANT market, currently inefficient with energy as Mark Rober shows, and the most expensive step of distribution by far. Amazon has presented two illustrative solutions.
First, their Scout robot is a sidewalk delivery that drives to people’s doors. They shut this down recently due to complexity and lack of traction. There are many other companies still trying this solution.
Second, their Prime Air was another solution where a giant drone carries stuff from a facility to someone’s yard where it lands and drops off the package. The technical problems were numerous and given some regulatory concerns from the FAA, the flights were never reliable enough even in initial testing to gain traction. Also, the drone sounded like a flying lawnmower. Not what you want to see landing in your apartment street or front yard.
Enter Zipline, a company with traction delivering supplies by drone in Rwanda since 2016. Instead of spending millions trying to force a solution in the US, smartly they have been working with the FAA to figure out how to achieve the regulatory concerns. Chief among them is the drone being aware of all other flying objects in the sky. Planes have radar but to fit that on a small drone is difficult. Zipline also has to do it cheaply if they want hundreds of drones delivering packages around a city. They have recently found an ingenious solution of using multiple microphones to detect where objects are. Just like your Alexa Echo speaker knows where in the room you speak from. But how does it do the last-mile delivery?
So now for the fun part, the mechanics. The first problem is noise. This needs to be solved to get buy-in from users. Zipline solves this in two ways. One is a quiet propeller design. The other is the lowering of the droid that holds the package from high up on the drone. This allows the noise of the powerful drone to be far away from the drop-off zone. It also solves the problem of scary blades getting close to any people. It does create more problems though.
Having a swinging droid on a line is dangerous. It only takes one droid cable to get caught in power lines to cause a catastrophe. Imagine the drone or electricity hitting a person. Therefore the lowered droid has fans that operate in two directions. That way if a 20mph wing hits the lowered droid it just blows 20mph in the opposite direction. It is like balancing a broom on your hand and keeping it perfectly straight up and down. Computers are great at this, even better than humans. This pendulum being accurate is great because Zipline can place a package right where they want it more safely.
It is genius if not a bit complicated from a mechanical standpoint. It also allows them to systematize the process where a drone can land at a store or warehouse and lower the droid to people (or robots ;) that can load it. This will be great for scale.
I talked about the beautiful and now the scary. Some examples include people grabbing the lowered droid when it lands, a dog jumping onto the droid, a clothesline blowing around the dropped cable, and someone walking out their front door into the droid. If a problem’s likelihood is 0.01% that means something will happen one out of 10,000 times. That means 100 drones operating 20 flights a day in a city will have that problem occur once every 5 days on average.
If I went to Zipline I would want to know what perception sensors they are thinking about on the droid and drone and how they still will detect these kinds of rare events. Also, if all goes wrong where does the drone try to land? The real world is dynamic and uncertain but just as technology can make the last mile efficient it can also solve the problems the new solution creates.
This basic idea is the core of science and engineering. If you look at any product, technology, or experiment it always solves AND creates new problems. It is a fun/hard game of whack-a-mole where the solution is sometimes hardware, sometimes design, sometimes features, sometimes UI, sometimes software, and many times multiple things. But if you keep working you never get all the moles but you do get to a point where the net benefit is strong and the world is better off and more exciting. Zipline’s delivery drone is the start of that exciting future. From my far-away vantage point, they are creatively whacking moles at the last-mile delivery more effectively than any other company yet.
Sean’s Corner
Drone delivery has been a long standing promise for robotics.
However, it’s fraught with challenges from noise, to traffic controlling once drone delivery is at scale, to inevitable job loss, to defense against nefarious drones, to just the aesthetics of wanting a clear view to the sky without a constant swarm of robots blocking the sun.
Zipline seems to have solved the challenge of noise and has certainly proven that they can create at-scale delivery operations - but have they solved the other issues?
For me, the true promise of drone delivery is a reduced environmental foot-print of last-mile delivery. While wheels are far more energy efficient than rotors, with their sharp right turns and zigzags, streets are not efficient means of travel. Drones offer a direct line-of-travel option that can transfer last-mile delivery from non-renewable sources to renewable. The transition is needed and welcomed. However, I worry that this promise will remain unfulfilled if cities and towns, in an attempt to keep our air clear and safe, confine drones to drive over existing streets.
Would we not be better off with automated delivery vehicles? Or a delivery aircraft carrier (hybrid last mile delivery), where an automated truck handles the last mile of transport and the drones handle the last 10-100 meters?
CONNIE’s Corner
I think that of all the drone delivery companies in the world, Zipline is doing it right. They have built a city drone that for the first time, actually has a chance to be successful at urban delivery. Let’s take a comparison with a standard DJI M300 quadcopter, one of the largest consumer drones from one of the largest manufacturers. The Zipline P2 blows the M300 out of the water with four times the payload capacity, a similar range, and a fraction of the noise. This combination rotor/fixed wing design is the way to go for delivery purposes, and addresses the pain points with off-the-shelf quadcopters.
Other companies have noticed combination drone designs. Los Angeles based company A2Z Drone Delivery has build a “Long Range Delivery Drone” called the RDSX Pelican. It boasts a 14 mile delivery radius and a payload of 11lbs. Unlike Zipline, A2Z is selling the drone itself rather than the service. It will be interesting to see if larger companies previously building their own drones will turn to third parties and focus on integration with their own platforms. If so, we can expect much wider adoption.
For the idea of drone delivery as a whole, I think there are still FAA regulations and concerns about overhead traffic. A drone falling with a parachute is still a several pound object falling from the sky into populated areas. There will definitely need to be regulation, perhaps “drone highways” that need to be established before we send hundreds of drones into the sky. In addition, the issue of “where to drop” is still unsolved. Although “dinner-plate accuracy” is great, it doesn’t help those who live in apartments, nor does it confirm the safety of the drop points in dense cities where construction, pets, and people must be detected in real time. Overall, I’m unsure if drone delivery will ever reach regular consumers, but may find continued adoption in the urgent needs and medical sectors that have space for deliveries and need for speed.
Synthesis
We were all impressed by Zipline’s progress and thrilled by its ability to help Rwanda get needed medical aid quickly. Even with their silent delivery, we all believe there are systemic problems with drone delivery in most areas of America.
We felt that, at scale, drone delivery would crowd the sky reducing quality of life for urban and suburban dwellers.
To solve this problem, we discussed the option of restricting drones to flying over streets. However, this puts traffic at risk from inevitable failures. Further, any security measures meant to enhance safety would likely make the drone heavier and bulkier, reducing its range and carrying capacity. Additionally, reduced flying directly to their target, further magnifies these limitations and makes an autonomous rolling land vehicle a better solution.
We also discussed hybrid solutions, a land based vehicle that rolls through a neighborhood with drones that launch to deliver individual packages. We felt this was a good compromise, but wondered if it was worth the complexity.
We did see one use case that obviated all of our concerns: life saving medical transport. Whether it be medication, emergency blood, organs or something else, we all agreed that the benefits of drones far outweighed the downsides.
We hope you enjoyed our 4th newsletter. If so, please share with someone who would also enjoy it.
Thanks,
Sevy, Sean, Connie
We hope you enjoyed this, if so or if not, you can let us know at welcome@lalovesrobotics.com