We continue with some old business, in “Tesla’s Self-Driving System Cleared in Deadly Crash” (Neal E. Boudette, The New York Times, January 19). The federal government concluded that that company’s “autopilot” feature, despite being overnamed and being engaged in a fatal May collision, did not need to be recalled. It has been modified to shut off once the driver ignores three warnings to put their hands on the steering wheel, and two competitive systems, from General Motors and Audi, will monitor driver’s eyes (!) for the same purpose.
“Are General Motors and Lyft About to Crush All Self-Driving Rivals?” John Rosevear posted this question on February 19th in The Motley Fool. GM is planning on making “thousands” of driverless Bolts next year for Lyft’s ridesharing, which the latter company apparently and rather reasonably considers critical not only to their prosperity but to their survival. In an area this fast-moving they can’t be sure, so the short answer to Rosevear’s query is “no.” Meanwhile, per Yahoo Tech three days later, competitor Uber is now testing its self-driving cars in Arizona, where the governor took a ride in one and praised it.
In the February 23rd New York Times, John R. Quain raised the issue of product standardization between driverless vehicle manufacturers. True, we will need that, but it’s too early to try to establish such conventions. We don’t know which self-driving technologies will win out on merit and in the market, which are two different things. It will be later in the process, maybe two years from now, before official standards are agreed upon and put into place.
The consortia and partnerships are well under way, and so are the legal cases. Per The New York Times (“A Lawsuit Against Uber Highlights the Rush to Conquer Driverless Cars,” February 24), Google’s Waymo driverless vehicle organization has brought an action against Uber for allegedly stealing trade secrets. With top technicians getting high seven-figure bonuses to change companies, and nondisclosure agreements insufficient to clearly define the thin and permeable line between personal knowledge and proprietary information, this was neither the first nor the last. The same goes for such ploys as, attributed to Uber executive Anthony Levandowski, trying to pay off “distracted” employee’s girlfriends to break up with them, and, as did Uber’s Otto division, filming videos of legally unapproved self-driving trucks. It’s an exciting high-stakes game out there – don’t be surprised to see many more of these sorts of things.
On March 8, we learned from Fox Business that “Uber Self-Driving Cars Are Coming Back to California Roads.” They were barred from them in December, but now have a permit to test there again, if for only two units, to become the 26th company allowed. That state has also proposed that cars without steering wheels be testable there (USA Today, March 10th).
My latest “why didn’t I think of that” moment came from John R. Quain’s March 9 New York Times “Cars Will Talk to One Another. Exactly How is Less Certain.” Since driverless vehicles will be receiving a lot of electronic information, it only makes sense for them to broadcast it to others nearby. Per Quain, they can relay news of “disabled cars and vehicles that are braking hard ahead, as well as slippery road conditions.” And more. They may not be able to stop stupid human drivers from trying to beat trains to intersections, but can at least tell them they are there. Two great advantages this communication will have – and I am almost certain it will be built into all American cars within ten years – is that it works even in areas without cellular service, and does not require infrastructure improvements. Perhaps it could also issue warnings about people cruising, in the left lane, at two-thirds of the speed limit.
This month’s most massive transaction was, as reported in The Wall Street Journal on March 13th (“Intel Joins Silicon Valley’s Race to Make Best ‘Server on Wheels,’” Ted Greenwald), was Intel buying Israeli car-camera company Mobileye for $15.3 billion. The two together will form only part of a driverless-vehicle consortium, but, with Intel’s chips, it’s a powerful combination. Intel’s estimate of “autonomous-driving systems, services and data” reaching $70 billion by 2030 strikes me as conservative, and, just maybe, if enough others share Intel CEO Brian Krzanich’s view that “what’s under the hood” will soon refer to computing features, millennials may find more interest in driving.
We finish with “The 5 Biggest Challenges to the Driverless Car Revolution,” or at least how Justin Loiseau saw them in his March 11th Motley Fool piece: pricing, consumer understanding, “safety/security issues,” regulation, and technology. It’s way too early to worry about the first. The second is also premature, as we don’t know exact capabilities yet. The third is more a matter of perception, as driverless vehicles should start cutting highway deaths as soon as next year. The outlooks for the last two, though, are clearly positive. The bad news for investors, to which The Motley Fool caters, is that we can’t yet tell the Fords and Chevrolets from the Stutz’s and Hupmobiles. But driverless cars are the McCoy, the real deal, the genuine article; if you have an appetite for risk, you should place some bets somewhere.