Architecture and Hardware

As Air Taxis Arrive, Software Issues Arise

The Volocopter urban air taxi is being prepared for a trial in Singapore late next year.
Within a decade or so, self-driving air taxis will be flying us around cities, sparing us from traffic congestion.

One cold night in late November, 120 transportation experts gathered at the Geovation Hub, a slick London coworking space, to learn about the latest in autonomous transport technology. This was not yet another self-driving car conference, however, but a meeting about a far more tantalizing prospect: self-driving air taxis—also known as passenger drones or personal aerial vehicles—that will, within a decade or so, be flying us around our cities, sparing us from all the traffic congestion below.

Organized by The Aviary Project, a U.K.-based aeronautical advocacy group set up to press for airborne travel options in towns and cities, the meeting counted airplane makers, airport consultants, air traffic controllers, aviation regulators, and city transportation officials among its attendees.

The nascent driverless car was smack in the meeting's crosshairs: "Traffic congestion sucks, and autonomous driving will not solve that problem. All that will do is exchange one type of car with another," said Fabien Nestmann, head of special projects at Intel-backed Volocopter, a Karlsruhe, Germany-based maker of a new breed of electric vertical takeoff and landing (eVTOL) aircraft.

Volocopter, which will trial its urban air taxi (which first flew in 2016) in Singapore in late 2019, is in the vanguard of an astonishing 130 companies, including firms backed by aerospace giants Airbus and Boeing, currently developing and test flying eVTOL designs in an attempt to make what they call urban air mobility (UAM) a reality.

The idea: instead of putting more rubber on the road, use the skies to allow commuters in eVTOL aircraft to fly in multiple low-level airlanes well below those of airliners, yet above those of delivery drones, easing pressure on the roads. Also, if eVTOL batteries are charged by zero carbon grids, lower emissions could result, too.

To scale to mass transit volumes, however, these air taxis will need to fly autonomously, with travellers simply punching in a GPS destination, probably on their phones. That means developing very high reliability software to fly them, and safely managing airspace, is going to be crucial. "They are going to be piloted at first, but automation is on its way. Automated flight is the future; it's just going to take us time to get there," said aviation consultant Darrell Swanson.

What is making eVTOL possible are recent step-changes in the power of lightweight electric motors and high-capacity lithium battery technologies, themselves spinoffs from electric car research and development.

"We have been studying electric propulsion for a very long time, and urban air mobility is probably the most significant application of electric propulsion we have seen," said Nikhil Sachdeva, an analyst with Munich, Germany-based aviation consultancy Roland Berger, which has identified no less than 55 firms developing eVTOL designs in the U.S., and another 58 in Europe.

The ability to use electric power in aviation is leading to a plethora of disparate air taxi designs based on different arrangements of electric-motor-driven rotors, steerable ducted fans, and propellers. For instance, China-based Ehang's UAM vehicle looks like a large quadcopter drone, whilst the Volocopter has an array of 18 differentially controlled rotors atop it. The vehicle from Boeing-backed Aurora Flight Sciences has static rotors for vertical takeoff and a cruise propeller that lends it forward thrust, whilst Airbus's Project Vahana has tiltable wings peppered with rotors, and Lilium of Germany has a vehicle with static wings peppered with tiltable ducted fans.

This rampant innovation means it is too early to say which kind of designs will best succeed. However, the ride-hailing firm Uber is actively encouraging this runaway inventiveness via its Uber Air division, which plans its own urban air taxi services in 2023 as a paying customer of eVTOL taxi makers, because unlike its troubled driverless car venture, it won't be making its own air vehicles.

Whichever designs succeed, Uber Air engineering director Mark Moore (formerly chief electric propulsion researcher at the U.S. National Aeronautics and Space Administration/NASA) is confident we are at a turning point. "With this awesome, new on-demand aviation market, we are at the beginning of a new age. It's like the X planes of the 1950s. We are really privileged to be the generation that will bring vertical flight to the masses," he said in a September lecture on eVTOL's prospects at Rensellaer Polytechnic Institute in Troy, NY.

Uber calls this "on demand" because the idea is that people will simply book a cross-town flight via a smartphone app from one of many hundreds of "vertiports"—roof-top helipad-type platforms. UAM proponents expect these to sprout on rented rooftops in much the same way cellphone antenna masts have sprung up. It's no pipedream: a firm called SkyScape is already scouting for vertiport rooftop space in London.

Yet for all their technological bullishness, the players in this nascent and highly unproven eVTOL space know they cannot afford any safety lapses with either their vehicles or the hi-rise vertiports.

The nightmare scenario that all want to avoid was played out horrifically in an earlier attempt to establish urban aerial mobility in New York City in May 1977, when a New York Airways helicopter running airport hops from the roof of the 59-story Pan American building (now the MetLife building) suffered an undercarriage collapse. A large helicopter tipped over, causing a still-turning rotor blade to separate and boomerang across the roof, tragically killing four passengers waiting on the rooftop, and seconds later one person on the ground.

"We just can't have that happen with UAM," says Sachdeva. "This entire trade will die before it has begun if we have any occurrences of that type." As a result, most air taxi designs will include a "ballistic" parachute, which is rapidly deployed by a small rocket if, say, control is lost or the eVTOL's main batteries fail, lowering the whole vehicle to the ground. However, as this only works at altitudes above 150 feet, and a parachute does not allow for a controlled landing, which is vital in urban situations to protect people on the ground, the chief eVTOL safety measure is likely to be a reserve battery pack.

Sachdeva says there should also be some safety built into the fact that eVTOLs will not have a helicopter's massive high-energy rotating blades, but arrays of distributed electric fans instead. "They also won't be flying a giant fuel tank over everybody's heads in a city, so it should also be safer," he says.

Yet the real eVTOL challenge will be in the software engineering, ensuring the reliability of the safety-critical flight code and its ability to manage the airspace safely and effectively, especially as delivery drones proliferate, both in early piloted fly-by-wire and later autonomous flight modes.

This won't be easy. In independent research into the technical hurdles posed by on-demand aviation at Duke University in Durham, NC, autonomous systems engineer Mary Cummings and her colleagues say using advanced artificial intelligence (AI) algorithms to control the torque and angle of multiple electric motors to achieve vertical takeoff has no precedent, and so will be tough to test.

In a research paper published last year, Cummings and colleagues wrote that while such technological advances hold great promise, "they represent a dramatic shift in propulsion technology with significant hardware and software elements that have never been certified outside experimental operations." The researchers also questioned how operators could provide sufficiently robust cybersecurity to prevent e-hijacking of the vehicles, especially when legitimate remote takeover is also seen as a safety measure.

Such questions are currently being considered by NASA, and by the leading regulators: the U.S. Federal Aviation Adminstration (FAA) and the European Aviation Safety Agency (EASA). Even the eVTOL developers know there is a mountain to climb before the technology is ready for prime time, and that expectations of services starting around 2023 may be optimistic, save for the most regulation-lite regions like Dubai and Singapore.

Despite this, the market is being talked up hugely: investment bank Morgan Stanley has somehow magicked up the figure that eVTOL will be a market worth $1.5 trillion by 2040, for instance. In a safety-critical market, cooler heads might be needed if expectations are not to be pushed too far.

"There is a huge amount of hype at the moment about the potential for eVTOL. The reality is that technology, regulation, and market acceptance are so many steps away before eVTOL can scale," says Jonathan Carrier of Slovakia-based urban air taxi maker AeroMobil.

"It is one thing demonstrating that your technology works. It is another thing entirely demonstrating to regulators that it is safe. They are worlds apart."

Paul Marks is a technology journalist, writer, and editor based in London, U.K.

Join the Discussion (0)

Become a Member or Sign In to Post a Comment

The Latest from CACM

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More