The New Autonomy

2018 was not a great year for the image of the self-driving car. Once heralded by futurists, manufacturers and the media as the most inevitable, invaluable and essential technology of the next few decades, enthusiasm seems to have stalled. Although technological development continues, it appears everyone is scaling back their expectations. In March, the unfortunate Elaine Herzberg became history’s first pedestrian victim of a driverless car, just over 121 years after Bridget Driscoll became the first pedestrian killed by a motor vehicle. Killed by an Uber-owned Volvo in Tempe, Arizona, Driscoll’s death caused headlines around the world. Up until that moment, the primary benefit of autonomy was seen to be safety, so the accident was a major setback, inviting a public relations backlash, a growth in general scepticism, an overall loss of confidence in the sector and the need of trying new transportation alternatives.

Slowly but surely, the leaders in the field have pivoted away from bold, sweeping statements about what their systems can do. Consumer bodies have also started to question the semantic disconnect between the systems currently available in modern cars and their actual abilities. It seems we wanted to believe a little too much. Is Tesla’s ‘autopilot’ remotely comparable to a true autonomous car? On the surface, at least, a Tesla’s ability to drive, steer and brake on its own appears like magic. 95% of the time, it just does fine. It’s the final 5% of unpredictability that causes problems. At last November’s D.Live technology conference in Laguna Beach, California, hosted by the Wall Street Journal, Waymo’s CEO John Krafcik said explicitly that “autonomy will always have constraints.” It seems there are constraints to autonomy that can’t simply be vaulted by AI. At least, not yet.

Many would say that doesn’t matter, because the algorithms that shape ride-sharing, and the labour practices and relentless demand that underpin it, have made the need for autonomy a moot point; Uber will still be in business regardless of who or what actually drives the cars. But there’s another, more important factor, to consider, and that is the shape of cities themselves. Algorithmic transportation relies on a status quo of car-accessible streets, resulting in increased, not decreased, local traffic.  Back in the 1980s, an American professor called Donald Appleyard discovered a correlation between traffic density and community cohesion. His book, Liveable Streets, published in 1982 by the University of California Press, showed that as more and more cars passed through a neighbourhood, the residents’ network of friends and acquaintances was eroded. Traffic trumped neighbourly connection.

Uber has three quarters of a million drivers in the US, all perpetually on the move, adding to traffic and congestion.

In London, the popularity of apps like Uber has seen the number of private hire drivers increase from 59,000 in 2009-10 to around 114,000 in 2017-18. In the early hours of the morning, the capital’s streets hum with Priuses awaiting their electronic summons or delivering passengers to far-flung suburbs. The benefits to its users are myriad, the downsides are shared by everyone else.

American cities were built for cars, and transportation alternatives are often scant.

In contrast, the grand boulevards of Paris or Barcelona were designed for horse-drawn traffic, first and foremost – cities with a medieval core were barely suited for more than a handcart. There are attempts to broach this gulf between transit on the mass, personal and micro level. Many global cities are now scattered with shared bikes or electric scooters as firms explore the role of ‘micromobility’ and give transportation alternatives to the cities. Elsewhere, transport remains a social device. Given that journeys are rarely 100% unique and many travel plans are shared by many, why not converge? In places without major transit infrastructure like light rail or even established bus services, ‘share taxis,’ typically small minibuses, are still the predominant form of mass transit. The tro tro of Ghana, the dalla dalla of Dar es Salaam in Tanzania, the Matatu of Kenya, as well as many other regional variations, all carry a significant percentage of local commuters, maximising the equation of precious road space and flexibility. In contrast, micromobility breaks the journey down, prioritizing convenience over image. Even this approach is not without its problems, as vandalism and cultural apathy ended bike share programmes in the UK, scooter share brands like Lime have struggled with technical and legal issues and China, the most enthusiast adopter of bike sharing, reportedly has mountains of discarded, unwanted bicycles.

There actually exists an alternative solution to modern urban travel. It might not be one that might not get you precisely door to door, but it is far more efficient in terms of the amount of space it takes up and people it can accommodate; it’s called a bus, and it’s worked effectively in cities for nearly 200 years. So why aren’t we investing more in smarter, cleaner buses as one of the best transportation alternative? Some places are (China, for example), but many other cities, especially in the US, are still in thrall to the idea of robotising their fleet of cars and taxis. The sad fact is that buses aren’t especially sexy; no venture capitalists are looking to sink hundreds of millions of dollars into an ancient technology. But when one considers that a city like Shenzhen in China has an all-electric fleet of 22,000 buses, you have to question to wisdom of constantly trying to reinvent the wheel. Existing technology, updated, paired with new approaches to frequency and smarter stopping and planning, can massive savings in both time and efficiency. In Autumn 2018, Barcelona finished an overhaul of its bus network, one that closed old lines, opened new ones and introduced a fleet of ‘Neighbourhood’ buses to join the suburban dots. Today, 95% of the city’s population has access to a better bus service. As a result, pedestrians are better served, neighbourhoods become more cohesive and reliance on cars – be they private, shared or ultimately even autonomous, is reduced.

Incremental change doesn’t have the kudos of technological revolution. But in order to break out of car-centric urban design, we don’t need smarter cars. Instead, we need to be cleverer consumers of transit, and that means making better use of the systems that have been shown to work for decades and try best transportation alternatives. A walkable city can become denser and more efficient, bringing benefits to all, and sharing low impact resources is the way forward.

Main Image: Illustration by Federico Babina