Cars That Think iconCars That Think

An image of the far infrared camera called Viper that the startup Adasky has developed for self-driving cars.

Night Vision for Self-Driving Cars

Elon Musk famously thinks that cars can be made to drive themselves without relying on expensive laser-ranging lidars. But while Tesla is moving ahead with one fewer sensor than most self-driving car companies, a new startup wants them to add yet another—an infrared camera.

Adasky is developing a far infrared thermal camera called Viper that it says can expand the conditions that automated cars will be able to operate in, and improve safety.

“Today’s sensors are not good enough for fully self-driving cars and that’s where we come in,” says Dror Meiri, vice president of business development at Adasky. “We think infrared (IR) technology can bridge the gap from Level 3 all the way to Levels 4 and 5.”

Read More
An illustration shows two cartoon cars bumping up against each other.

Nvidia vs NXP—Whose Robocar Brain Will Win?

NXP and Nvidia have announced computing platforms for smart cars, and each company claims that its platform is the best by far.

Which company is right? Or is each right in its own way? The answer depends on how you assign a grade.

NXP today announced the S32x automotive processing platform, apparently an elaboration of the BlueBox system unveiled last year. We said then that BlueBox “knits together devices the company already sells,” and NXP appears to be doubling down on such interoperability.

“All the parts that come out will have the advantage of being compatible with this platform—this vision processor, this radar processor, this torque managment device,” said Matt Johnson, who’s in charge of NXP’s product lines, software, and automotive processors. 

Read More
An early lidar prototype from Strobe Inc., which resembles a silver metal box with a green lens, pictured with a Sharpie for scale.

GM Cruise Snaps Up Solid-State Lidar Pioneer Strobe Inc

General Motors announced on Monday that it had acquired Strobe Inc., a small California startup that has been developing a sub-$100 solid-state lidar for self-driving cars.

The terms of the deal were not announced but, given that Strobe was seeking only US $15 million in a funding round as recently as May, it is unlikely to be anywhere near the reported $680 million Uber paid for some of Waymo’s lidar team as part of its ill-fated Otto purchase.

Strobe looks to be a much better deal all around. The 12-person company is only three years old, and uses innovative frequency-modulated lidar technology first invented by its founder, Lute Maleki, at his previous firm called OEwaves.

Read More
photo of waymo vehicle in phoenix suburb Chandler

Report: Waymo to Announce True Robocar Service

Waymo is about to announce that its robocar ride-sharing program in Phoenix will dispense with its backup driver, according to a story in TheInformation.com. That would make it the first-ever commercial application of true self-driving technology.

The report cites two unnamed sources, apparently engineers unhappy with the fast pace that management has decreed for the rollout of the new service. The CEO of Waymo is John Krafcik, but the ultimate authorities are Sergey Brin and Larry Page, the heads of Alphabet, the holding company for Waymo and its big brother, Google.

Waymo declined to answer IEEE Spectrum’s request to confirm or deny the story. We’ll take that as a qualified “maybe.”

Five years ago, Brin famously promised to have a self-driving car on the road within five years. Maybe he thinks the time has come to put up or shut up. If the company isn’t ready to remove the backup driver this year, perhaps it would content itself with an announcement that it will do so next year.

Waymo is under pressure to show pre-eminence in the field that it pioneered, but so are the two dozen-odd other players. Among those rumored to be near some sort of demonstration of ready-for-the-road robodriving are NuTonomy, the ride-hailing service that debuted last year in Singapore; Uber; Tesla; and most of the big car companies. 

Audi took the lead three months ago with the unveiling of its Audi 8, which it says was the first car that has what it takes to reach Level 3 autonomy. That’s when the driver can read a newspaper but must always be prepared to take back the wheel within about 10 seconds if the car should have trouble.

What must an aspiring robocar company do to get ahead of the pack? I mean, beside administering sedatives to its lawyers and amphetamines to its publicists?

The first priority is to solve any outstanding tech problems. According to the account in TheInformation, Waymo’s cars still have trouble making left turns into traffic, particularly when there’s no left-turn arrow to serve as a guide. Another sticking point is navigating any area that Waymo’s superprecise maps haven’t sampled much, like shopping-center parking lots.

You must also choose your venue wisely, and this Waymo has done. The current ride-sharing problem is in the Chandler suburb of Phoenix, where the intersections are simple, the weather is balmy, and the legal environment is mild.  

Finally, you should listen to your own engineers. If they say you’re going too fast, slow down. Moving fast and breaking things isn’t good in the auto business. And if you don’t listen to your technical people, they may resort to leaking rumors to the press, as seems to have happened here. Worse, your best people may up and leave.

When a Tesla car equipped with Autopilot slammed into the side of truck, killing the driver, Tesla’s then-partner Mobileye threw a fit. The head of Mobileye accused Tesla of selling Autopilot as a self-driving technology when it was still just a driver-assistance feature. Tesla split with Mobileye, which Intel went on to acquire.

Court documents accidentally reveal costs for the company's self-driving car project between 2009 and 2015

Google Has Spent Over $1.1 Billion on Self-Driving Tech

Google has never publicly shared how much it spends on its self-driving cars. At first, Project Chauffeur was hidden away in Google’s ultra-secret X moonshot program. When that went public, its costs were bundled together in a vague “Other Bets” category that includes the company’s fiber Internet service, home automation, and life science spin-offs.

Now, a court filing in Waymo’s epic and ongoing lawsuit against Uber has accidentally revealed just how big a bet Google placed on autonomous vehicles. Between Project Chauffeur’s inception in 2009 and the end of 2015, Google spent $1.1 billion on developing its self-driving software and hardware, according to a recent deposition of Shawn Bananzadeh, a financial analyst at Waymo.

Bananzadeh was testifying as part of the lawsuit, in which Uber stands accused misappropriating trade secrets and violating patents from Waymo, Google’s self-driving-car offshoot. Because Waymo has yet to commercialize any of its technology in a meaningful way, the company thinks any damages in the case should be calculated on the basis of how much it spent building the technology in question.

When asked by an Uber lawyer how an estimate for developing one of the trade secrets, number 90, was arrived at, Bananzadeh replied: “My understanding is that it is a cost that captures the entire program spend from inception to the period of time where it stops.” He later clarified that meant from 2009, when Sebastian Thrun got the go-ahead for the project from Larry Page, to the end of 2015.

Throughout Bananzadeh’s deposition, every dollar amount was redacted to protect Waymo’s confidential commercial information. Every time, that is, except in the Uber lawyer’s very next question: “The calculation that was the basis of the $1.1 billion cost estimate for Trade Secret 90 is the same calculation that was done for Trade Secret 2 and Trade Secret 25?”

Waymo had apparently given an identical $1.1 billion cost estimate for each of the trade secrets being discussed. Bananzadeh was unable to provide a clear answer as to why that might be, except to say, “Insofar as it is part of the entirety of this self-driving system…. therefore, all of the costs of the program since inception… are what then informs that number.”

Waymo’s position seems to be that all of its trade secrets are inextricably linked to the whole self-driving car project, and any damages should reflect that fact.

In a filing, Otto Trucking called Waymo’s damages theory “entirely speculative” and “over the top,” and called on the court to forbid Waymo from offering any evidence or argument beyond the actual damages it has incurred.

Though $1.1 billion is unquestionably a massive figure, it actually seems quite reasonable compared to the recent over-heated market for self-driving car acquisitions. In March 2016, General Motors paid a billion for San Francisco–based Cruise Automation, a company that was a seller of after-market semi-autonomous vehicle kits. In February of this year, Ford invested the same amount in a joint venture with Argo AI, a two-month-old Pittsburgh start-up headed by a former Google self-driving car engineer. The largest self-driving acquisition to date, however, was Intel’s $15.3 billion purchase of Mobileye in March. The Israeli company had originally provided vision-based semi-autonomous technology for Tesla vehicles.

Uber shelled out a reported $680 million for self-driving truck maker Otto in August 2016, sight almost unseen. But it’s the circumstances surrounding the acquisition of Otto, and in particular its lidar technology, that are at the heart of Waymo’s case against Uber. Otto’s founder, Anthony Levandowski, allegedly had a draft contract for the purchase of the company before he even quit Google.

By spending its money earlier than others and mostly in-house, Google’s billion-dollar investment now looks relatively modest—almost a bargain. Waymo has, by far, the most sophisticated self-driving software. It has simulated over a billion miles of driving, and its cars have had the most self-driving experience on real streets (over 3 million miles in multiple cities).

The court case seems to suggest that Waymo has also built up an enviably solid platform of intellectual property. So, undesirable as this peek into its books might be for Google today, the company should pride itself on demonstrating that in-house R&D can still make a lot of financial sense.

A blue, medium-sized truck with the words "eCanter Powered by Positive Energy" on the side

Daimler Unveils First All-Electric Truck

Daimler unveiled the world’s first series production, all-electric truck at a gala launch in New York City yesterday.

The Fuso eCanter, a blocky, medium-duty truck made by Daimler’s Mitsubishi Fuso, was driven across town to emerge from behind a barrier amid a stream of bubbles and a crescendo of music. That made it impossible to take note of the truck’s most salient feature: its silence.

“How quiet is it?” I asked Marc Llistosella, the chief executive of Mitsubishi Fuso.

“Too quiet,” he responded. “We need a beep tone in early mornings and late evenings” to alert pedestrians to the truck’s presence.

Read More
A photo shows the black interior of a hot pink car, and a sign with the Lyft logo mounted to the dashboard.

Lyft and Drive.ai Gear Up to Test Autonomous Ride Sharing

Most autonomous car companies see ridesharing and taxi services as the clearest way for fully autonomous vehicles to become useful and cost effective. Unsurprisingly, ridesharing companies (whose largest expense is paying human drivers) are in enthusiastic agreement. Lyft in particular has formed partnerships with GM’s Cruise Automation, Waymo, and nuTonomy.

Last week, Lyft announced a new autonomous vehicle partnership, this time with Drive.ai, who plans to deploy their deep learning-based autonomous vehicles in and around San Francisco.

Read More
An illustration shows a woman with short hair wearing sunglasses and elaborate green earrings in front of a pink background.

Could a Pair of Earrings Hurt Waymo’s Lidar Trade Secrets Lawsuit?

A big part of Waymo’s trade secrets case against Uber and self-driving truck startup Otto Trucking is that the company takes protecting its technology very seriously.

“All networks hosting Waymo’s confidential and proprietary information [are] encrypted and [require] passwords and dual-authentication for access,” read its original complaint, in which it accused engineer Anthony Levandowski of downloading 14,000 of those secret files about lidar and other technologies. “Computers, tablets, and cell phones… are encrypted, password protected, and subject to other security measures. And Waymo secures its physical facilities by restricting access and then monitoring actual access with security cameras and guards.”

However, according to a recent court filing, Waymo also turns its secret lidar circuit boards into jewelry and gives them away to employees leaving to work for rivals.

Read More
Nissan Motor Co's revamped Leaf electric vehicle is displayed at the Makuhari Messe on September 6, 2017 in Chiba, Japan

Nissan’s Next-Gen Leaf Drives Farther, Thinks Deeper

Nissan was the first big car maker to offer an all-electric car and the first to give driverless technology its full commitment (as boldly asserted by then-CEO Carlos Ghosn). But this year Tesla and Chevrolet invaded the company’s niche by offering electric cars at comparable prices but with superior range and self-driving features.

The new Leaf answers all these challenges. Still, it can no longer claim to be the firstest with the mostest.

It was indeed first when it introduced the Leaf, back in 2010. True, Tesla had begun selling the all-electric Roadster two years earlier, but at US $100,000 that car was more in the way of a proof of principle. Nissan’s Leaf cost around US $35,000. And, with the continuing evolution of Nissan’s ProPilot driver-assistance package, the company has always been high up in the robocar peloton.

But Tesla’s new Model 3 costs no more yet can go some 350 kilometers (220 miles) on a charge. And the Chevrolet Bolt—also in the mid-$30k price range—goes a whopping 380 km. Both the Model 3 and the Bolt offer a suite of self-driving features, albeit some of the coolest ones come only as options.

The new Leaf has a battery pack rated at 40 kilowatt-hours, up from 30 kWh. That’ll take you a solid 240 kilometers (150 miles)—enough, Nissan says, to quell the range anxiety of customers in Asia and Europe. A performance version, due out in early 2018, will have 60 kWh and travel 480 km (300 miles), which ought to assuage the anxieties of American drivers.

The updated Leaf also offers a more ambitious edition of ProPilot, its answer to Tesla’s AutoPilot. The system can take over the entire job of parking and of driving in slow, bumper-to-bumper traffic and along single-lane roads. And in the Leaf all these features are standard.

And there’s one thing that no other company does: single-pedal driving. Press down and the car accelerates; lift your foot and it brakes. According to Automotive News, it’s an improved form of a similar system in the Nissan Note e-Power, a hybrid vehicle sold in Japan.

The winning driverless car

The Tech That Won the First Formula Student Driverless Race

Engineering student Manuel Dangel of Swiss Federal Institute of Technology (ETH) in Zurich and teammates were walking the racecourse at Formula Student Driverless in Hockenheimring, Germany, earlier this month when they realized that the computerized wheelbarrow they were using to map the course had gone haywire. [See "Students Race Driverless Cars in Germany in Formula Student Competition" 16 August 2017.]

As part of the track-drive event, one of several events that make up the entire competition, the rules permit teams half an hour to walk the racecourse and make measurements they might need to program their driverless cars. Because the track-drive event consists of ten solo laps on the same, unchanging course among traffic cones, “the basic strategy is to run within the map,” Dangel says. If you cannot make a map before the event, though, you have to switch to a more complex strategy.

Read More
Advertisement

Cars That Think

IEEE Spectrum’s blog about the sensors, software, and systems that are making cars smarter, more entertaining, and ultimately, autonomous.
Contact us:  p.ross@ieee.org

Editor
Philip E. Ross
New York City
Contributor
Willie D. Jones
New York City
 
Senior Writer
Evan Ackerman
Washington, D.C.
Contributor
Mark Harris
Seattle
 

Newsletter Sign Up

Sign up for the Cars That Think newsletter and get biweekly updates, all delivered directly to your inbox.

Load More