Are we asking the right questions about self driving cars? How much should we be willing to change to speed the arrival of autonomous vehicles?
Creating autonomous vehicles (AV) is one of the most technically complex undertakings ever. It’s at least comparable to the space race or any other challenge humans have set themselves. But it faces even bigger obstacles than merely imbuing technology with the ability to make life or death decisions — it poses difficult questions for regulators, entrenched industries, vested interests and urban planners. And that’s before we turn to the question of public trust. In short, the world is not designed for AVs and we’re not ready for them.Yet surely we cannot allow the status quo. The daily carnage from human-driven cars should be motivation enough for radical change, and the promise of more affordable, more sustainable and more efficient mobility should be irresistible. But the irresistible object is encountering the immovable forces of inertia, infrastructure, tradition and intransigence.
If we could start over, deploying AVs would be easy. We could design and build streets easily and safely navigable for our robotic brethren. But we’ve had more than a century of people being used to driving and have created a world predicated on human-driven cars. Making infrastructure more robo-friendly for the less-adaptable-but-more-accurate and predictable AVs is a task that few seem to have an appetite for just yet. What if we improved road markings and layouts? What if we enabled traffic lights to communicate with cars, and all cars were technologically unable to break traffic laws?
Acceptability issues quickly arise when changes from normal, well established behaviours are proposed. People reach rapidly for talk of “rights” for things that are clearly merely norms. The greater good rarely prevails easily or quickly in the face of individuals’ attachment to what they’re used to. AV detractors often talk of their desire to maintain independence and control, citing a lack of trust in the technology, but offer little in the way of solutions to the prevailing toll of road death and injury. The AV industry craves popular and official acceptance and is often hesitant to require or even request unpopular changes for its benefit.
As I said in my book, Life As A Passenger, if autonomous technology had been available at the time the car was introduced, nobody would have thought human driving was a good idea, and the world would have evolved differently. Now though, we have massive industries that depend on human drivers. The flexible, common-sense nature of humans makes us capable of driving in situations and surroundings that can present an almost infinite array of variables. But we’re also prone to lapses of attention, poor judgement and selfish maneuvers. We’re rapidly approaching the point where self driving technology can cope with the vast majority of situations, but remains years away from achieving the flexibility to deal with all conditions. The quest for the hard-to-define “good enough” level of performance will delay the benefits of AVs for many years to come.
Waiting for the Robots to Come to Us
I want to pose the question about what we can do to enjoy some of the benefits of AVs without sacrificing safety or requiring overly onerous infrastructure or behavioural changes. This is the several trillion dollar question!
It’s interesting to consider Voyage, the self-driving startup that has established operations in The Villages, a 32 square mile retirement village in Florida — a closed environment that lacks many of the challenges of open roads. This provides an easier world in which to learn more about autonomous driving issues, but still provides enough inputs to advance the technology. Similarly, low speed shuttles (like those from Navya or EasyMile) operating in defined areas are a gentle introduction to the world of autonomous transport. It’ll also be instructive to see what can be learned from the Chinese experiment to dedicate several highway lanes to AVs.
We’re seeing massive efforts being put into building self driving technology that can coexist in the messy world of human driving. Rightly, much of the regulatory effort is focused on ensuring that AVs reach a very high safety threshold. But I wonder if we should make more effort, more changes to accommodate AVs into our world faster, rather than waiting for them to be perfect on our historically-anchored terms?
I wouldn’t like to think that things have to get worse before they get better. We shouldn’t accept that deploying autonomous cars will see a transition period where their coexistence with human drivers will increase collisions. Yet we can’t be naive enough to imagine that there will be a seamless, ubiquitous or fast departure from the dominant form of transport (and urban planning) of many decades.
So who is willing to move the goalposts? Will cities dare to inconvenience human drivers to make way for autonomous vehicles? Will regulators find a way to balance liability issues with ways to encourage development? Without relaxing the pressure on autonomous companies to reach the highest safety thresholds, is it time to have difficult conversations about ways we could make that outcome easier to achieve? Unless something changes, it’ll be a long slow road towards driverless cars and the attendant safety and time benefits.
David Kerrigan is the author of Life As A Passenger, which is available in English and Chinese. He presents at the Stanford Continuing Studies Self Driving Cars courses and writes extensively about the social impacts of emerging technologies at https://david-kerrigan.com