Q&A with Dave Vos, Head of Google’s Unmanned Delivery Vehicle Program

Dave Vos heads up Project Wing, Google's unmanned delivery vehicle program. Originally from Capetown, South Africa, he came to the United States at age 26 in order to do graduate work at MIT. While there, he earned his master's and PhD degrees. He has been involved in creating automated flying machines for over 20 years.

Q: It seems like everybody’s talking about developing delivery drones lately. Why the big fuss all of a sudden?
A: Many of the same technologies that have put smartphones in our pockets—smart software and small, inexpensive sensors like GPS and accelerometers—can be used to fly small vehicles on pre-planned routes. It’s become a lot easier for companies around the world to develop relatively inexpensive platforms for amateur and commercial users alike.

Q: Why is Google working on them?
A: Think about the congestion, pollution, and noise created by delivery trucks double-parked all over our cities, or the fact that we send a two-ton vehicle across town to deliver a two-pound package. On the other hand, a self-flying vehicle that can cover about a mile a minute would guarantee speed, accuracy, and on-time delivery. They could open up entirely new approaches to transporting and delivering goods—they’d be cheaper, faster, less wasteful, and more environmentally friendly than ground transportation. They also have the potential to help in crisis situations, like delivering medicine and batteries to cut-off areas after a natural disaster, or helping firefighters improve communication and visibility near a wildfire.  

Initially, we thought that defibrillator delivery would be a natural way to implement our vehicles—when a person needs a defibrillator, every second counts, and drones don’t have to deal with traffic. Ultimately, we had to put that ambition on hold because we realized there are many challenges with integrating into the emergency medical system that are outside our control. But we certainly hope we can try again someday.

Q: Where is Project Wing right now?
A: Last August, we successfully tested real-world deliveries in Australia using our prototype vehicles. But our goals require more than us building our own operational aircraft—we aren’t going to be the only game in town, and we need to ensure that everyone can operate their own unmanned aerial vehicles (UAVs) safely. Today, no system currently exists to manage the layer of airspace under 500 feet. So we’re also working on a traffic management system that could support a scalable, safe, and reliable commercial aerial delivery service, alongside others. As with any such project, we need to gather feedback, so we’ve been talking to regulators and aviation experts to develop a common approach from the very beginning.

Q: How’s your relationship with the FAA [national aviation authority for the US]?
A: It’s positive and collaborative—we often meet with them and other regulators to explain how our technology works. We recently held a seminar where we invited the FAA, members of the Small UAV Coalition, and other members of the aviation community to gather feedback on what sorts of technologies might enable safe flights at low altitude. Meetings like this help inform our own product development.

Q: How do you feel about the FAA’s proposed regulations, which allow for limited, low-risk operations, but effectively rule out an aerial delivery service like Project Wing?
A: While we don’t necessarily agree with everything in these proposed regulations, we’re supportive of the FAA’s goals of integrating UAS into the national airspace. We recently submitted comments to say that the FAA should be able to approve more advanced operations as operators demonstrate greater safety and reliability.

More generally, we’re committed to working with governments around the world, as well as the broader aviation industry, to safely integrate small UAVs into the airspace.

Q: How do UAVs know where to gois there someone sitting behind a screen controlling them?
A: We’re still working this out—but the short answer is, while we’ll need to have an operator overseeing the vehicles, we’re designing our systems to be highly automated.

Q: How big are they?
A: The vehicle we’ve been testing is about one-and-a-half meters from wing tip to wing tip, and about one meter long (from nose to tail). But we’re looking at lots of different design options because different vehicles are good for different things. It’s too early to know what our final design, or designs, will look like.

Q: When will I see a self-flying vehicle delivering packages to my door?
A: There are a lot of technical and practical issues that still need to be resolved—for example, people’s concerns about safety, privacy, noise, or air congestion. Should self-flying vehicles be allowed to operate at all times of day? What’s the best way to let people know who’s flying vehicles above their property? We’d need to have answers to these kinds of questions before starting a full cargo delivery service. But we’re getting there—we’ve been testing people's responses to the design of the vehicle, its noise, and the drop-delivery experience—and will be listening carefully as we develop our technology further. We expect we’ll hit our safety and reliability targets in a matter of years, not decades.