It happened. A self-driving car has killed a pedestrian. Skeptics are calling for development to end.
According to reporting from ABC News, Elaine Herzberg was struck by an Uber autonomous SUV as she attempted to cross Mill Avenue on Sunday evening in Tempe, Ariz., just outside of Phoenix. She later died from her injuries.
The terrible news feeds into public fears. But it will not stop the development of self-driving cars.
Most people now carry computers in their pockets that are far more powerful than the one that put a man on the moon. It’s easy to take them for granted, since they are so ubiquitous.
On the other hand, self-driving cars have emotive value. They are the computer-driven future we were promised in “The Jetsons” way back in the 1960s. They represent humans accomplishing something that should not be possible.
That value shows no signs of corroding. That’s because the future, in this respect, has not arrived yet. And in all honesty, they are not coming to the masses anytime soon.
Related story: Porsche’s Electrifying New ‘Mission’
Initially, only very expensive cars will get fully self-driving packages. Most cars will get a watered-down version. Cars will self-steer and brake to avoid accidents. They will have better cruise control. They will take over when we are distracted. It will make them much safer, cheaper to insure and more pleasurable to drive.
Development of fully autonomous technology is about the future of mobility. Uber, Alphabet (GOOGL), Baidu (BIDU) and even Apple (AAPL) want a piece of the operating system of mechanized movement. It is foundational, and important. It is also likely to be a business far larger than simply selling cars.
In The Rise of Mobility as a Service, Deloitte, the global consulting firm, predicts this trend will change the way people move around. It will create a valuable new ecosystem — one that consists of car and bicycle sharing, taxi service, mass transit, digital payment systems and smart cities.
It is a multitrillion-dollar opportunity already set in motion. It begins with self-driving cars.
The Tempe incident is a setback. According to witnesses, Uber’s Volvo SUV didn’t even slow down as it hit Herzberg even though a human engineer was in the driver’s seat, and the vehicle is equipped with state-of-the-art LiDAR, an advanced radar system that uses light detection and ranging to detect objects.
On Monday, Tempe police chief Sylvia Moir told the San Francisco Chronicle that an early probe of the accident showed no fault by Uber. Moir explained that Herzberg emerged abruptly from shadows behind a dimly lit center median. She then pushed a bicycle laden with plastic bags into oncoming traffic.
“It’s very clear it would have been difficult to avoid this collision in any kind of mode,” Moir said.
However, the sole reason Uber’s tricked-out Volvos have LiDAR is to see, and make sense of obstacles that are poorly lit. Avoiding the accident might have been impossible. But the vehicle should have automatically slowed down upon detection.
The promise of self-driving cars is not that they can perform as well as humans. To gain public confidence, they must be much safer than a human driver.
In 2015, Elon Musk, the co-founder of Tesla (TSLA), suggested safe, fully autonomous cars would eventually lead the National Highway Traffic Safety Administration to ban human drivers.
Forget that this is a strange comment coming from a chief executive of a car company. Musk’s point was that autonomous cars will become so safe that riders would fear human drivers.
Months later, Google said its program had only 13 minor fender benders despite logging 1.8 million miles of autonomous driving since 2009. During that time, the autonomous vehicle was not deemed at fault for a single incident.
In contrast, the NHTSA Motor Vehicle Crash Overview reported 6.3 million traffic accidents in 2015, an increase of 3.8% over the previous year. The number of fatalities was 35,092, a 7.2% increase, and the fastest rise since 1965.
The data does not lie. Humans are terrible drivers. And they are getting worse.
In 2017, the NHTSA began advocating for greater automation, and the removal of “unnecessary” barriers to self-driving vehicles. The experts argue more technology will save lives.
And that means any furor over the Tempe incident will likely be very short-lived. Government agencies and big businesses are lobbying for more automation, faster.
The best way to play this trend is still Nvidia (NVDA). The semiconductor company bet big on deep learning in 2010. Its form of embedded artificial intelligence is cornering the market for the first batch of self-driving solutions, due to reach high-end vehicles next year.
The Santa Clara company reports it has working relationships with 320 automakers, tier-1 suppliers, developers and researchers.
In 2017, Nvidia partnered with Bosch, the world’s largest auto parts company. The goal is to eventually bring the same cutting-edge technology to the larger market.
Pullbacks in Nvidia due to news events like the Tempe crash are a buying opportunity. My subscribers are sitting on a nice 277% gain in this name, and there’s plenty more upside to come. Not just for this company, but for others that become a major part of this valuable ecosystem. To get opportunities like these delivered to your inbox each week, click here.
Jon D. Markman