SkyDrive Is Designing a New Quadcopter That Fails Only Once Every One Billion Rides

This interview has been edited for length and clarity.

Interesting Engineering: Flying cars have been a fixture of science fiction for decades, but safety remains a major concern. What’s SkyDrive’s approach to keeping passengers safe?

Mark Blackwell: The question really is ‘Do we need to meet conventional aerospace failure rates, which is 10^-9?’ The answer, of course, is yes. We're taking the standards and processes from conventional aerospace and applying them to these types of vehicles. Our current model is a tech demonstrator that requires special flight permits and a special test pilot.

IE: How are you incorporating that dedication to safety into the design process?

MB: The safety assessments are happening at the aircraft level, the system level, and the component level. We’re looking at the life of the vehicle. We’re looking at what kind of testing, verification, and qualification is needed. We’re also looking at what configurations are needed at the aircraft level to meet regulator requirements for things like bird strike, motor failures, battery failures, certain weather conditions. We do all that by modeling simulation, by analysis, and then finally by testing the physical hardware. All that is baked into the product that we will take to market.

IE: When do you anticipate bringing SkyDrive to market?

MB: We’re aiming to achieve certification by the end of 2024 and launch the vehicle in 2025 at the World Expo in Osaka. We'll have the public actually use the vehicle during the expo.

IE: What does it take to build a team that can make something that has never been made before, assuming you see this as something that's never been made before?

MB: Well, it's certainly novel in the sense that the technology has never been applied to a flying vehicle of this size. Of course, the drone industry is very mature and developed, but they can’t fly beyond the visual line of sight. I would agree our vehicle is novel.

Building the team is actually one of the big challenges because it's a collision of lots of different industries, and there are different schools of thought. You look at the automotive industry. They are excellent at mass manufactur[ing], and their work with materials is what we need to produce vehicles at scale. But the safety standards and the design processes are different to what we would see in the aerospace world. 

Then we've got guys in the drone industry who very much have a build-test-fly-learn type of attitude, which is great for getting things working fast. But to get things to certification, that process sometimes isn't the most optimum. Balancing all those different approaches is for sure one of the biggest challenges, but it's also a strength because we get different perspectives on problems and different ways of doing things. So making sure those approaches are integrated is one of the key things.

IE: What are some engineering challenges you’ve faced so far?

There have been a lot. Figuring out how the battery is used has been a challenge because the design isn’t fixed. We haven't validated how this vehicle will be used. Can we swap the batteries, or do we recharge upon landing? What kind of battery life should we expect from this type of vehicle? Determining how to optimize battery usage is a challenge.

Aerodynamically it's a challenge. Especially with this tech demonstrator, the counter-rotating propellers cause a lot of aerodynamic interaction from propeller to propeller — a lot of wake effects — and onto the body, especially when we add things like crosswind. That's very challenging aerodynamically.

The stability required in the algorithm that controls the speed of these motors can be difficult, then linking that to the structural design. Because obviously, the main aim is to reduce the weight because the less weight we have the further range. The design of the crossbars is very key in allowing enough rigidity to not get any vibrations from the motor because that can then create instabilities and runaways that can cause problems. So there's a lot of tuning that’s happening. Not so much of the design is fixed yet. Everything's been tuned at the same time. 

IE: Was the quadcopter design an obvious choice?

I think it's the simplest choice for a tech demonstrator. The high-level objective is to keep it in a small [physical] footprint. We looked at octocopter designs, but obviously, as we're adding more propellers, we're increasing the footprint. When we move to the certification product, then maybe we won’t see this exact configuration. To deal with things like bird strikes, you really have to look at the layout in a different way.