Humanising autonomy: What do we actually need from self driving cars?

ustwo Auto's design concept for driverless cars: the ustwo Roo

Driverless cars are in development or already on the road thanks to Google, Tesla and Volvo to name a few iconic brands. But emerging autonomous vehicle (AV) designs are capturing everyone's attention without always answering many ethical or social questions.

It won't be long before you start to see these cars roll out on main roads. Development progresses daily – Delphi has bought self driving car company Nutonomy to scale and deliver autonomous vehicles, Blickfeld has raised US$4.25 million in seed funding to let autonomous vehicles ‘see’, and driverless cars are already in 35 cities across the globe right now, as seen in the Global Atlas of Autonomous Vehicles in Cities by Bloomberg Philanthropies.

While it’s an arms race to see who can produce the best foolproof technology, digital production studio ustwo’s Auto branch has released design principles for autonomous vehicles – with people first in mind. ustwo’s Auto team already work for clients such as Jaguar, Landrover and Transport for London. Hear why ustwo Auto published design principles in the video below.

There were more autonomous driving concepts at CES 2017 than there ever has been before, and last year, a massive US$1049 million was invested across 87 auto tech deals, with 91 percent increase in funding, according to ustwo.

But driverless cars raise so many questions; like how will I know it won’t crash? What will I do instead of driving? What about people marginalised by this technology?

ustwo design principal Tim Smith and interaction lead Harsha Vardhan tell me about their vision for driverless cars after nine months of extensive research with humans across all ages and sectors – from UX and AI researchers, professionals and experts in the autonomous vehicle trade, children and even London black cab drivers who fear for their jobs.

Tim, Harsha and the Auto team says autonomous technology is meant to help humans be more autonomous, not the vehicle, so their book argues more focus needs to be on serving people first and what increasing our mobility will look like.

Auto’s book of design principles, Humanising Autonomy: Where are we going? (you can download the full version online) explores our social needs, overcoming fear and anxiety over having no driver, designing a car that isn’t either intimidating or “too cute”, entertainment options within the car, and its accessibility for people often marginalised by technology advancements are all discussed.

Read more here: Humanising Autonomy: Where are we going?

Bearing all of these design principles in mind, ustwo has designed its own driverless car, and named it Roo. You can read the full description of how ustwo created Roo here. We feature Roo concepts throughout this feature. 

Humanising Autonomy

Harsha, Tim and the team carried out user testing in a number of vehicles including the Tesla with its Autopilot, both version one and two. 

"We used the Navya autonomous vehicle, running it’s trials within London’s Olympic park. In the course of our research we have also simulated rides in AVs to see how people might react to driverless transport – both within and outside the vehicle," says Harsha. 

Tim explains how the public responded in one user test. 

“We were driving people around, and the vehicle would come to a certain stop, say the taxi driver would swerve or stop if a dog was crossing the road. Immediately the driver relays feedback there, the passenger knows nothing is wrong, he just stopped in front of a dog, and everything’s okay,” says Tim, explaining the impact of removing a driver.

“However if you take the driver away, if the car was to screech to a halt it would take the same basic manoeuvre a human would, but it wouldn’t be swearing.

“Then people would wonder why it had stopped like that, and they'd start panicking and wondering if it’s the machine, if it’s a bug or an error, and then they would feel unsafe in the vehicle.”

Roo concept: If it’s a city car it won’t need much streamlining, so we can play with the shape a lot more, says ustwo

The driver isn’t there just to serve the purpose of brakes and acceleration, says Tim, but to be responsible for communication with the passenger. When designing autonomous vehicles, it’s important to build a form of humanised interaction into the machine so the passenger is aware of what’s going on with the vehicle, and this maintains trust.

“If trust isn’t established, then people just won’t adopt it,” he says.

When carrying out user research for the book, the Auto team found an overwhelming trend among almost all ages that the vehicle wasn’t trusted to transport people from A to B safely – they were certain it would crash.

Tim draws parallel to the early adoption of commercial aviation. The technology gained people’s trust once they had a safe and pleasant first flight. The same applies for autonomous cars. Any potential causes for anxiety ideally need to be mitigated in the first experience so people will use it again.

"I think the public will adopt to it as the most ‘boring’ way to travel safely. It will be less about safety concerns but more about the experience it offers – fast, efficient and not very unique," says Harsha.

Emotional Needs

There will always be resistance to driverless cars beyond the feelings of fear and anxiety. Apart from those who fall into the category of having a passion and joy for driving and its heritage, people also ride in taxis for the sole purpose of socialisation.

When we think about travel, it’s often about getting from A to B as fast as possible, but for some, there’s an emotional significance to the journey. ustwo spoke to one lady who suffers from isolation-based depression after recently becoming widowed.

“She uses a taxi to get somewhere she wouldn’t normally use a taxi to get to; for that human interaction. She gets a taxi just so she can have a meaningful conversation with another human being,” says Tim.

“In the book we draw on the study that suggests for human interaction is just as important as our need for water.”

The study is from a book cataloging studies by John Cacioppo, director of Chicago's Center for Cognitive and Social Neuroscience. He speaks about this in an interview with Fortune as well. 

Although many people fear autonomous vehicles are going to take over the roads and become the only mode of transport, Tim says the roads will always be shared by both manual drivers, autonomous vehicles and cyclists.

“I don’t think it will ever be 100 percent autonomous vehicles. Not because it won’t be possible, but because not everybody wants autonomous vehicle. It’ll be a choice,” he says.

Functional Needs

It may seem ridiculous, but Tim says it’s important for the exterior of an autonomous vehicle to be just right – not "so cute" that it gets bullied by other drivers, and not too imposing or scary for other drivers.

Google’s autonomous vehicle design (seen here) has been described as "so darn cute" from the exterior. It was designed not to scare pedestrians or increase fear around self-driving vehicles.

"I think that’s gone too far and fellow drivers are bullying it, because [people] know it’s autonomous, and they would never hurt a human but they know they can take advantage of that. They’ve got to challenge it and test it," says Tim.

"But you also can’t create a vehicle that other drivers fear, it needs to be respected so that if you are a passenger in an autonomous vehicle, and you just want to get to work on time, you won’t be either bullied or be too scary."

Aside from its emotional impact on pedestrians and passengers, Tim says autonomous cars could serve a number of different functional needs purely based on demand at that time – transforming from a commuter car, to taxi, to delivery car, to emergency vehicle all in one day.

ustwo's Roo AV design in taxi form

"It could be a commuter service ferrying six to eight people, but when commuting dies down, perhaps vehicles could shift into parcel delivery or mobile meeting rooms (no destination, just working space rolling around)," says Tim.

"Or if people can’t afford emergency vehicles, maybe they’re having a heart attack, if it speeds it could turn into ambulance. The outside of the vehicle could communicate its legal importance on the road and other autonomous vehicles allow it through the traffic automatically."

This chameleon-like concept demonstrates driverless cars wouldn’t just be a matter of being owned or a type of public transport.

"I think there’s more to it than that. Some people will own one, but it can be a lot of other things; royal mail vans, a pizza delivery van, a meeting room, emergency vehicle, or surveillance vehicle," says Tim.


There's an important piece of technology coming into fruition now, where the exterior of a car can act more as a screen than a static paint job.

There are a number of concepts which showcase this technology, says Harsha. From the "garish Toyota attempt", to something a little more subtler, and Audi's Swarm concept

"We can actively posit that this could be a future direction by manufacturers," he says. 

This same technology could also be used to change the nature of the car, making it less about personal ownership and more about efficiency for the transport network in the city.


Much focus can be laid on what passengers could do with their spare time once the responsibility of driving is taken away – will there be digital games on screens, like we see on planes? Or will people sleep and read a book like they do on trains?

But Tim says the experience of a driverless car isn’t the same as public transport, and the question shouldn’t be about what passengers can do, but more about what they need to do with the spare time.

"If you’re in the small environment of an autonomous vehicle, on quite chaotic roads with no driver, it’s not the same as a plane or train journey with huge amounts of cognitive offload."

He envisions commuter cars transporting people for journeys seven to 15 minutes long. The chance of sleeping or being productive in that time is very unlikely, says Tim.

The simple interior design of ustwo's Roo AV

"A screen inside a vehicle is probably not going to be used because everyone has a screen in their pocket. In New York City taxis, many of them have screens in the back but no one engages with them because they’re already on their own screen doing their own thing."

But Tim does see “amazing, nuanced entertainment options” within AV interiors as an opportunity for brands to differentiate from one another – something which the exterior of autonomous vehicles may not provide.

Harsha says it's easy to think about advances such as augmented reality (AR) being used for entertainment, but we can also think about its use as a means to prevent motion sickness within the vehicles, for example, augmenting the road and keeping people's eyes on it, so that they do not lose reference to a stationary plane, thus preventing sickness.

"It can also be quite educational - the means to see the city or the world with another lens for example, seeing London through the lens of an 18th century peasant."

I'm sure a lot of us can relate to artist Jon Burgerman's ideal AV interior


Whenever new technology is designed, accessibility has to be at the forefront. Adding additional technology for marginalised people (disabled, visually impaired, partially deaf) as an afterthought is no longer acceptable, says Tim.

The same entrance to an autonomous vehicle should accommodate both able bodied and disabled people, unlike London buses which can subtly discriminate disabled people by having a separate entrance for them in the middle of the bus that requires technology to lower it, as opposed to the front of the bus.

"It somewhat segregates disabled people by design," says Tim.

Another example is when visually impaired people feel for the wing mirrors on a car to position themselves alongside it. Even though autonomous cars may no longer need wing mirrors, ustwo included another design that stimulates the marker of a wing mirror in their own concept car.

Notice how none of these AV designs don't incorporate wing mirrors

"It will be a no-brainer for people with mobility issues, the blind or even children to use those vehicles on whim – an empowering agent in people’s lives," says Harsha.

"The barriers to adoption such as ’trust' can be slowly whittled down by people seeing the positive effects it brings to roads and quality of life, of course aided by good vehicle and service design. Even creating more jobs, especially in the service sector."

So, when will we see driverless cars?

Although many companies are already testing out autonomous cars, some estimate they won’t be on public roads until 2030.

"I personally think the adoption will also be culturally and geographically radically different - countries where autonomy can have huge impacts on safety might possibly see it last - like in India, where policy makers are against it’s implementation due to the loss in jobs and the technological complexity involved in seeing a safe vehicle on the roads given the chaos there," says Harsha.

"It harks back to the William Gibson quote about the future being here, but not uniformly distributed."

Note: We may earn a commission when you buy through links on our site, at no extra cost to you. This doesn't affect our editorial independence. Learn more.

Read Next...