Self Driving Cars: 5 Things We Need To Get Right
Autonomous vehicles (AVs) are all over the news these days, and soon on our streets. But there are five critical things we need to get right for them to succeed.
Yes we are close — but not there yet. Tesla has already AVs on highways and Google has self-driving cars that have crossed almost 2 million miles in controlled situations and conditions, mostly in Silicon Valley. But we are still very much developing systems that can transition among different environments (highway, city road, suburban road) seamlessly, can work as effectively in inclement weather like snow and rain, and can handle reasonably other drivers who are not obeying driving regulations. Looking at it another way, smart car still have the challenges of understanding the world around them (collecting data through sensors and processing them) and then deciding what to do. You will get very varied answers even among experts on how far we are, ranging typically from a couple years to a decade, but the consensus is that technology is actually the easier of our problems.
Should we be creating separate lanes just for self driving cars? Should they be co existing with manual cars on the same roads? Should we ban human drivers completely and just go completely with AVs? If a self-driving car gets into an accident who is to blame? There are an incredible number of questions that regulators need to address, and the pace of law-making is already falling behind the pace of technology evolution. The regulation will have tremendous impact on many other industries, including insurance, city infrastructure, zoning standards, ie various other parties will promote and resist AVs according to their interests. Right now places like Singapore and Milton Keynes in England have taken the forefront in terms of AV regulation and likely will set precedents for larger jurisdictions.
AVs are juicy targets for hacking and can obviously lead to tragedies. Car security will be far more challenging than planes — there are a lot more of them, making many more choices while moving, potentially communicating with many vehicles around them rather than a single tower. The good news is that we have built other industries, from cell phones to satellites, that we can learn from and there are smart people everywhere working on this next set of challenges. But we have barely scratched the surface and this is an area ripe for innovation and building large companies.
You can put technology and rules into a car but there is still the question on ethics. What should a car do if it has to decide between hitting a pedestrian or suddenly causing a pile up behind and potentially hurting more people? Or that braking the car suddenly may critically injure the passenger but save the biker just crossing in front? What does it mean to take an AV manufacturer to court? When human beings have to decide between two hard choices we tend to project and understand their dilemma, but we may not be ready to do that to non-human systems yet. One of the biggest selling points of self driving cars is that they will save lives overall since they will prevent many human accidents. But our expectations around machines are higher, just consider what happens when there is a first fatality involving a self-driving car.
Which brings to arguably the hardest and biggest challenge — public perception. We can get the technology as perfect as we can, enforce great regulation, imbue our AVs with fantastic security and even a sense of morality, but public perception can overrule all of these in how quickly AVs succeed. Adoption follows a U curve with early adopters, then a chasm of trust the innovation needs to cross before the mass market accepts it, and finally the late adopters. Laggards and counter interests will often criticize an invention, which can shape public perception very much against it. And when it comes to such a revolutionary, impactful technology like AVs, which affect the daily lives for billions around the world, the discourse and battle for adoption is going to be very public indeed. Companies, regulators, engineers, venture capitalists, and thought leaders better be prepared.
illustration credit: Avi Naim
illustration credit: Avi Naim