BIG DATA, AI & CONNECTIVITY TAKE THE WHEEL
According the US Census Bureau, each year Americans collectively spend an estimated 3.4 million years in their cars commuting, a mind-boggling tally that has increased 20% since 1980. That is a lot of time to be crawling in traffic — more than enough to inspire a small but growing cadre of engineers and entrepreneurs who see opportunity in gridlock.
Although a fully autonomous car is still on the horizon, human drivers have been ceding control, bit by literal bit, for well over a half century. Cruise control has been a standard feature for decades, while cars that can parallel park on their own or brake to avoid accidents have become increasingly popular. Apps such as Waze and Google Maps have taken over navigation duties. Meanwhile, in the six years since the first ridesharing app was launched (Sidecar), such services have become so popular that one L.A. developer has designed a parking garage that can easily be converted into a gym, mall or theater should the demand for parking spaces crash.
Safety (speaking of crashes) is another reason that self-driving cars have gained traction. Human drivers are a disaster. Last year in the US there were 40,000 traffic fatalities, more than two million injuries and auto repair costs spiraling into the hundreds of billions of dollars. Get humans out of the driver's seat and drunk or distracted drivers would no longer be a problem.
Replace them with a combination of connectivity, sensors, data analytics, robotics and artificial intelligence and humans can stay out of trouble, relax in seats designed to swivel for face-to-face conversation, stretch out for a nap, make phone calls or text to their heart's content.
Every part of an autonomous vehicle needs to be connected and in constant communication with every other part. Each vehicle, in turn, needs to be aware of all the others on the road as well as of the road infrastructure itself: streets, signs and traffic signals. If connectivity can be thought of as the central nervous system, then analytics and machine learning are the brains that transform data into action.
The auto business has traditionally been about steel rather than software: atoms rather than bits. It is no surprise, then, that it is tech companies with years of experience pushing the the boundaries of mass plug-and-play navigation and cross-device communication that are taking the lead developing smart cars able to function autonomously within a larger mobility network. Google, Apple, Facebook and Uber are working on self-driving cars, while Tesla, which could be described as a tech company that happens to make cars, leads the automaker pack.
Car manufacturers (also referred to as original equipment manufacturers or OEMs) have a competitive advantage in their knowledge and control of vehicle data. With such a broad, deep and detailed understanding of cars and drivers, such proprietary datasets could be leveraged to develop intelligent mobility platforms customized for individual consumer experience. It is a tall order technically that will likely require strategic partnerships with tech companies or, to follow BMW's lead, the development of tech-dedicated divisions, to accomplish.
Beyond technological challenges, psychological hurdles could also slow the adoption of self-driving cars. Humans may be spectacularly bad drivers, but they love driving and the feeling of control that comes with it.
According to a recent study by Deloitte, 75% of those surveyed thought self-driving cars were unsafe, despite their near perfect track record on road tests. Consumers have also become more leery about supplying companies with data, and the "data exhaust" from a connected car is substantial. Then there are cybersecurity worries.
Which leads to the biggest hurdles: regulation and liability. A connected vehicle is a hybrid: Data-driven technologies are regulated by the Federal Communications Commission. Cars and truck OEMs answer to the Department of Transportation. Sensors—the Internet of Things so critical to connectivity—fall under the umbrella of the Federal Trade Commission. Who is responsible for what when something goes wrong?
Panelist Chris Bonanti, a former Associate Administrator of the National Highway Traffic Safety Administration, recommends the development of performance-based standards to ensure minimum regulatory features are covered from a safety perspective. If autonomous driving delivers on its promise, it will help make the roads safer for everybody.
by Kara Voght, Medill School of Journalism
Director, Regulatory & Compliance
former Associate Administrator
Director, Product Management, BMW Technology Corporation
Head of Digital Ecosystem
New Mobility Consulting
New Mobility Consulting