Robotic cars

Recent robotic car news

Published 5.12.2017
A brief look at some recent news in the robotic car market, and it's hard to mention robotic cars these days without talking about Waymo's lawsuit against Uber.

The judge in the Waymo-Uber suit continues to be impressed with the breadth of the evidence that Waymo has assembled. There is plenty of evidence that Anthony Levandowski stole the files ust before he quit Google. However, there is still no evidence that Uber has the files. The judge ordered Uber to expand its search, but given Uber’s history, does anyone reading this think Uber can be believed?

The judge also denied Uber’s request to have the case decided in arbitration, which means the case will be settled in public— although the two companies could still work out a settlement without an arbiter before the trial.

Not only will there be a trial, but Judge William Alsup also referred the case to the U.S. Attorney for investigation of possible theft of trade secrets, a move that suggests the Uber engineer at the center of the case could face criminal charges.

Levandowski has taken the 5th, and that was a factor in the judge’s decision to deny arbitration.

“Even though he is not a defendant here,” Alsup wrote in the court filings, “Levandowski’s assertion of his Fifth Amendment privilege has obstructed and continues to obstruct both discovery and defendants’ ability to construct a complete narrative as to the fate of Waymo’s purloined files. As a practical matter, it is hard to imagine how consolidating proceedings as to Levandowski and defendants, whether here or in arbitration, could alleviate these difficulties.”

Here's a different take: Robotic cars will kill Uber? TTJ has been saying for a while that Uber’s future depends on robots. the company desperately wants to rid itself of its expensive drivers, and can not be late to the party. Ford has clearly indicated that it is going to challenge Uber as a ride provider.

Hence the need to steal the technology needed from Waymo. Waymo seems to have found a carmaker in Fiat willing to cede some control of the car experience. Fiat has not robotic car effort, and is not looking to steal the technology but buy it. Waymo has indicated that it does not want to make cars.

Uber’s philosophy, both internally and in its pitch to consumers, is that it’s a hassle to own a car. The irony is, for the pay-by-the-ride future of transportation to be realized, someone has to own a lot of cars. Chances are, it won’t be Uber.

Rather than product providers, carmakers are to become service providers, at least in this thesis. I am not convinced.

All of this is part of what auto makers and ride-hailing companies anticipate will be a larger transition to “transportation-as-a-service”—potentially the end of widespread vehicle ownership in developed countries. Subscribing to such a service for all of a person’s transportation needs within a typical American city could cost anywhere from 10% to 25% what an average consumer now spends on owning, maintaining and insuring a vehicle, says Tony Seba, co-founder of technology think tank RethinkX. Cost savings on that order could lead to rapid adoption akin to the touchscreen smartphone revolution, he argues.

The point in this article is that currently, Uber offloads the cost of the vehicle and maintenance onto drivers. No drivers, then Uber has to swallow those costs. If Uber isn’t going to own the vehicles, then who is? Switching between ride services is easier than switching between cell phones. Thus the author concludes that Uber is doomed. Even if it gets its robotic cars up and running.
Are humans the hold up for robotic cars? No, not really. The thesis is that robotic cars will be programmed to drive perfectly, and that is a poor mix for driving with humans. However, Google is already aware of this problem and is training its cars to drive more like humans.

Programmers are optimistic that someday the cars will be able to handle even Beijing's traffic. But the cost could be high, and it might be a decade or more before Chinese regulators deem self-driving cars reliable enough for widespread public use, said John Zeng of LMC Automotive Consulting.Intel's Winter expects fully autonomous cars to collect, process and analyze four terabytes of data in 1 ½ hours of driving, which is the average amount a person spends in a car each day. That's equal to storing over 1.2 million photos or 2,000 hours of movies. Such computing power now costs over $100,000 per vehicle, Zeng said. But that cost could fall as more cars are built.

Developers working with artificial intelligences (AI) are having the systems learn to drive by mimicking humans. Among the significant sticking points with AI is that engineers can't explain how the systems make the decisions they do.
This is not surprising. Amazon is looking into robotic cars to deliver its packages quickly and cheaply. No way does Jeff Bezos want to pay drivers. For the time being, the company is not looking to build its own robots, but rather looking to see how it could leverage robotic technology, most of which does not exist.

The initiative, still in its early phases, could help the Seattle-based company overcome one of its biggest logistical complications and costs: delivering packages quickly. Amazon could use autonomous vehicles including trucks, forklifts and drones to move goods. In addition, driverless cars could play a broader role in the future of last-mile delivery, enabling easier package drop-offs, experts say.

No one actually knows what Amazon is up to, however. A robotic truck wouldn’t be held to the same rules as a human driver, who can only drive 10 hours. Four days to cross the US becomes 1.5. Amazon is also interested in drones, and that is where (for now) their biggest interest lies.
Apple is starting to put robot cars on the road in California, with former NASA roboticists behind the wheel. Apple has some catching up to do.

Apple Inc. ’s AAPL +1.05% plan for autonomous vehicles calls for putting more-senior engineers in all of its cars than some of its rivals are using for road tests, a move that suggests the company is still in the early phases of testing its technology, analysts say.

In a permit issued April 14 by the state of California, obtained Friday through a public-records request, Apple identifies six employees, including roboticists who worked at the National Aeronautics and Space Administration, who will be in the front seat of three Lexus sport-utility vehicles outfitted with technology to make them autonomous.

Apple is using Bosch veterans as well, and Bosch has become a big supplier.