Google visits Detroit, JD Power says people want robocars, Thrun on Charlie Rose & More


A round-up of interesting robocar news this week:

More information has come from the Google team (to which I am a consultant) at the Society of Automotive Engineers conference in Detroit. In a speech there, covered in the Detroit Free Press and many others Anthony Levandowski outlined how Google has been talking to all significant car manufacturers about how they might work together to produce cars with Google's technology. Google is not looking to become a car manufacturer, but does want to see a real car on the roads -- and not next decade.

At the same time, talks with insurance companies about how to provide insurance for self-driving cars are also going on. Insurance companies pay the cost of all accidents, either directly through policies bought by the driver, or indirectly through insurance sold to manufacturers, and of course all these policies and cars are really paid for by car owner/drivers. As long as accidents are lowered, and the cost per accident remains the same, it's a win.

At the same time J.D. Power and Associates released a study on self driving car markets. This survey shows around a third of buyers would like to get self-driving functionality in their car, and about 20% would pay $3,000 for it. While advanced laser-based scanners cost much more than that today, I am confident that Moore's Law and higher volumes can bring things down to that price. These numbers are quite high for such a radical new technology. Such technologies normally only require a small volume of early adopters to get them going. The varoius basic autopilots announced by car manufacturers which require you to still keep your attention on the road will sell for well under $3,000.

Sebastian Thrun, leader of the Google X Lab, recently appeared on Charlie Rose where he spoke about the car, about Glass, and mostly about Udacity, his personal online education project. Sebastian also publicly posted that he took one of the Google self-driving Lexus cars up to lake Tahoe this weekend. I do think those long vacation home drives will be a big driver of people to pay serious money for a self-driving car. Saving time on the average 30 minute commute is one thing, but the 4 hour drive to Lake Tahoe is a real change, especially if you can use the time to interact with your family or get in serious reading or video watching. Of course, right now, Sebastian was keeping his eyes on the road in case he needed to intervene, since this is still a prototype.

Finally, NHTSA has released a report saying that robocars could eliminate up to 80% of crashes. While they won't get to that number right away, I think they can even do better in time. David Strickland, the head of NHTSA, has stated he has very high hopes for the technology, which is tremendous news, because it means that one my biggest fears in my early days of forecasting this technology -- too much government opposition -- seems less likely.

Some accidents are caused by mechanical failures (like tire blowouts or bad brakes) freak weather and other situations a self-driving car can't do much about. We may never get to zero. But this should still be the biggest lifesafer in the developed world until somebody cures some of the biggest diseases.


A small (perhaps nit-picking) point:

I agree with most of what you say, but at the moment, iatrogenic disease (and especially drug resistant hospital infections) which could be easily stopped with basic measures like universal hand sanitation in hospitals kill several times more people than car crashes. (We could also radically slow the progress of antibiotic resistance by not giving antibiotics to livestock on a routine basis).

This is not to say that I don't view the self driving car as a major advance, and indeed one I desperately want for both safety and personal comfort reasons. I am fully in favor of pushing the self driving car out as fast as feasible. However, using known, quite inexpensive techniques, we could save most of those people lost to medical accidents and hospital infections, and that would be a lot more lives saved.

I agree it is larger. Though it's not just one thing -- better handwashing would do quite a bit, but I'm not sure it would save the majority of those, but if you have numbers I am curious. I believe it's a wide variety of different sanitation problems, and also drug prescription and delivery errors and many other things. Note that 3 people close to me have died or lost limbs to such infections, so I do understand how serious it is.

One big difference is car accidents kill more young healthy people. While hospital infections kill allot sick people who are likely die of something else in the near future.

Could you elaborate on why lower accident rates are a win for insurance companies? They are only in business because of ill-fate. They make a margin on the volume of business they do. It is the risk of loss that compels people to pay premiums. If risk and loss go down, society is happier as a whole but insurance companies' business decreases.

Yes, it's not a win for certain elements, like auto body shops or emergency rooms. With insurance it's more complex. Claims will go down (which is good) but eventually premiums will go down. Though not in California, the largest market, where a silly constitutional amendment blocks giving a major discount for having a safer car. (The history of this is just insane.)

Of course insurance companies insure against more than car accidents, and even in cars, they make a fair bit selling comprehensive insurance and other types. But the insurance companies (and ER doctors) are not monsters, they don't want people to die just so they can have a business. They will migrate their skills into other markets they can go after. Markets change, this is the way of the modern world.

The insurance companies will make money because they wont drop insurance rate as much as the accident rate in auto cars drops.

Add new comment