Alexei Andreev- Autotech Ventures
Since the creation of the National Highway Traffic Safety Administration in 1970, much of the innovation we’ve seen in the automotive space has centered around the safety of vehicles. Seatbelts, anti-lock braking systems, crumple zones and other innovations have all increased the safety of vehicles and the survivability rates of accidents. Traffic-related fatalities have steadily declined — however according to the National Safety Council, vehicle deaths in 2018 were estimated at 40,000 for the third straight year. Could it be that we’ve reached peak safety when it comes to driving? Unless we as a society become more open to the idea of autonomous vehicles, this could very well be the case.
Whether or not autonomous vehicle technology will be capable of making roads safer is not the question. It most certainly will. The mainstream adoption of autonomous vehicles has little to do with technology, and much more to do with the way people perceive their own safety. Because safety, after all, is a relative measurement — the outcome of a social contract in which we decide what level of safety we’re comfortable with.
For example, airplanes are shockingly safe. The probability of dying in a commercial airplane accident is considerably smaller than the probability of falling at home and suffering a fatal injury and massively smaller than the probability of dying in a car accident on the way to the airport. Despite these probabilities, a fear of flying persists among many who would rather safely stay at home or travel by automobile.
Why? Because, safe as they may be, planes still fail. And when they do, it makes news and the public discourse is all about the technology failure associated with the accident. But technology will never be perfect, and so firmly planted on the ground, we simply feel safer, even as we’re flying down the highway in the middle of traffic. And yet, as recently as 2016, the NHTSA reported that 94-96% of traffic accidents are due to human error. We fall asleep, get distracted or agitated and get tunnel vision; as a result, we cause accidents. So isn’t it time to embrace an autonomous future?
Imagine AI enabled automobiles that would far exceed human performance as it pertains to safety, thanks to their ability to access and process data from multiple sensors to make decisions instantaneously.
Safety innovations of the past seem simple compared to what we’re working on now. Internet of Things (IoT) technologies will allow real-time diagnostics and prognostics on cars. Cloud computing and big data will enable vehicles to talk to each other. Augmented and virtual reality technologies stand to improve driver attention and data visualization while reducing cognitive overload behind the wheel. Which of these innovations will be the next seatbelt in terms of its impact is difficult to say. However, the cumulative impact of these new technologies will likely be just as significant as the mechanical innovations of the past.
It is no longer a question of if, but when. When will societies be willing to adopt and accept a new approach to safety? Even if AI-enabled cars increase safety by 10x, there will be the occasional failure. When technology replaces human error as the cause of traffic fatalities, will society be accepting of this? Who will be held liable in courts of law? We’ll have many questions to answer as a society, and I predict that autonomous adoption will be staggered. Some rational and data-driven societies will be more open to AI-driven vehicles if they’re statistically safer. Other societies, based on their historical and legal settings, may be unwilling, or unable, to change their perception. Time will tell.