Raffaello D'Andrea: Meet the dazzling flying machines of the future

Hits: 1431

Raffaello D'Andrea

When you hear the word "drone," you probably think of something either very useful or very scary. But could they have aesthetic value? Autonomous systems expert Raffaello D'Andrea develops flying machines, and his latest projects are pushing the boundaries of autonomous flight — from a flying wing that can hover and recover from disturbance to an eight-propeller craft that's ambivalent to orientation ... to a swarm of tiny coordinated micro-quadcopters. Prepare to be dazzled by a dreamy, swirling array of flying machines as they dance like fireflies above the TED stage.




What started as a platform for hobbyists is poised to become a multibillion-dollar industry. Inspection, environmental monitoring, photography and film and journalism: these are some of the potential applications for commercial drones, and their enablers are the capabilities being developed at research facilities around the world.


For example, before aerial package delivery entered our social consciousness, an autonomous fleet of flying machines built a six-meter-tall tower composed of 1,500 bricks in front of a live audience at the FRAC Centre in France, and several years ago, they started to fly with ropes. By tethering flying machines, they can achieve high speeds and accelerations in very tight spaces. They can also autonomously build tensile structures. Skills learned include how to carry loads, how to cope with disturbances, and in general, how to interact with the physical world.


Today we want to show you some new projects that we've been working on. Their aim is to push the boundary of what can be achieved with autonomous flight.


Now, for a system to function autonomously, it must collectively know the location of its mobile objects in space. Back at our lab at ETH Zurich, we often use external cameras to locate objects, which then allows us to focus our efforts on the rapid development of highly dynamic tasks. For the demos you will see today, however, we will use new localization technology developed by Verity Studios, a spin-off from our lab. There are no external cameras. Each flying machine uses onboard sensors to determine its location in space and onboard computation to determine what its actions should be. The only external commands are high-level ones such as "take off" and "land."


This is a so-called tail-sitter. It's an aircraft that tries to have its cake and eat it. Like other fixed-wing aircraft, it is efficient in forward flight, much more so than helicopters and variations thereof. Unlike most other fixed-wing aircraft, however, it is capable of hovering, which has huge advantages for takeoff, landing and general versatility. There is no free lunch, unfortunately. One of the limitations with tail-sitters is that they're susceptible to disturbances such as wind gusts. We're developing new control architectures and algorithms that address this limitation. The idea is for the aircraft to recover no matter what state it finds itself in, and through practice, improve its performance over time.






When doing research, we often ask ourselves fundamental abstract questions that try to get at the heart of a matter. For example, one such question would be, what is the minimum number of moving parts needed for controlled flight? Now, there are practical reasons why you may want to know the answer to such a question. Helicopters, for example, are affectionately known as machines with a thousand moving parts all conspiring to do you bodily harm. It turns out that decades ago, skilled pilots were able to fly remote-controlled aircraft that had only two moving parts: a propeller and a tail rudder. We recently discovered that it could be done with just one.


This is the monospinner, the world's mechanically simplest controllable flying machine, invented just a few months ago. It has only one moving part, a propeller. It has no flaps, no hinges, no ailerons, no other actuators, no other control surfaces, just a simple propeller. Even though it's mechanically simple, there's a lot going on in its little electronic brain to allow it to fly in a stable fashion and to move anywhere it wants in space. Even so, it doesn't yet have the sophisticated algorithms of the tail-sitter, which means that in order to get it to fly, I have to throw it just right. And because the probability of me throwing it just right is very low, given everybody watching me, what we're going to do instead is show you a video that we shot last night.






If the monospinner is an exercise in frugality, this machine here, the omnicopter, with its eight propellers, is an exercise in excess. What can you do with all this surplus? The thing to notice is that it is highly symmetric. As a result, it is ambivalent to orientation. This gives it an extraordinary capability. It can move anywhere it wants in space irrespective of where it is facing and even of how it is rotating. It has its own complexities, mainly having to do with the interacting flows from its eight propellers. Some of this can be modeled, while the rest can be learned on the fly. Let's take a look.




If flying machines are going to enter part of our daily lives, they will need to become extremely safe and reliable. This machine over here is actually two separate two-propeller flying machines. This one wants to spin clockwise. This other one wants to spin counterclockwise. When you put them together, they behave like one high-performance quadrocopter. If anything goes wrong, however -- a motor fails, a propeller fails, electronics, even a battery pack -- the machine can still fly, albeit in a degraded fashion. We're going to demonstrate this to you now by disabling one of its halves.




This last demonstration is an exploration of synthetic swarms. The large number of autonomous, coordinated entities offers a new palette for aesthetic expression. We've taken commercially available micro quadcopters, each weighing less than a slice of bread, by the way, and outfitted them with our localization technology and custom algorithms. Because each unit knows where it is in space and is self-controlled, there is really no limit to their number.








Hopefully, these demonstrations will motivate you to dream up new revolutionary roles for flying machines. That ultrasafe one over there for example has aspirations to become a flying lampshade on Broadway.




The reality is that it is difficult to predict the impact of nascent technology. And for folks like us, the real reward is the journey and the act of creation. It's a continual reminder of how wonderful and magical the universe we live in is, that it allows creative, clever creatures to sculpt it in such spectacular ways. The fact that this technology has such huge commercial and economic potential is just icing on the cake.


Thank you.