Self-Driving Vehicles
11-09-17, 09:46 PM
#466
Here's an interesting article:
http://www.thetruthaboutcars.com/201...aced-mobility/
Carlos Ghosn has effectively relegated the old, fun Nissan of the 80's and 90's to history and replaced most of their vehicles with CVT crossovers (excepting the Z and GTR) but this is refreshing coming from a well known CEO such as him. Note that he's not saying Nissan isn't getting into "mobility" themselves because they most certainly are. He sees both traditional ownership and mobility coexisting.
We'll see if he turns out to be right.
http://www.thetruthaboutcars.com/201...aced-mobility/
Carlos Ghosn has effectively relegated the old, fun Nissan of the 80's and 90's to history and replaced most of their vehicles with CVT crossovers (excepting the Z and GTR) but this is refreshing coming from a well known CEO such as him. Note that he's not saying Nissan isn't getting into "mobility" themselves because they most certainly are. He sees both traditional ownership and mobility coexisting.
We'll see if he turns out to be right.
11-09-17, 10:39 PM
#467
it will start in cities. We can see how even today citiy governments hate cars - they have more and more sections that are pedestrian only, where parking availability sucks, and in cities like london they have fees for the privilege of bringing your car in. So as self driving uber cars emerge it will be pointless, stupid, and problematic to 'drive' there and then those cars will be banned. But it will be a long time until farmer joe doesn't drive his old f150 somewhere 
although even there, joe may have more knowledge than we realize because self driving tractors are already out there. https://www.digitaltrends.com/cool-t...ving-tractors/
although even there, joe may have more knowledge than we realize because self driving tractors are already out there. https://www.digitaltrends.com/cool-t...ving-tractors/So what are people doing some of the time? Using ride shares. I've used them myself but really only for situations in which I know I probably don't want to park my car on a dark side street in a sketchy part of town. Depends on how much those fares are, too. They aren't always cheap and thus don't make sense most of the time. But I'm sure once Uber and Lyft no longer have to actually hire drivers using their own cars and underpay them then the cost per fare will go down (a bit of sarcasm there).
The reality that it isn't all that great for rideshare drivers aside, this is a trend now in cities. You don't always use these services but sometimes you do when you are going to an event somewhere where you know you'll have a nightmare finding parking. Or maybe because you'd rather just leave your car home in the garage that night rather than find the one poorly lit side street to park it on with a meter. Many, many times though, that is still the preferable option to a rideshare.
You're not wrong about some people in local government who seem to hate cars and would prefer there be less of them or to disallow them. Folks like those usually make themselves sound totally inflexible on the subject and it's hard to imagine they can accept that some people actually like to drive vehicles even in a big city.
I can actually see some reason why this might happen in an extremely dense part of a huge city. But for that to work the public transit would have to be as good as it is in Manhattan. And if a ridiculously vast no-vehicle area is drawn up it will force some people to move further away from where they are and consequently it will make properties for sale and rent closer to the exclusion line exponentially more expensive-- because outside of that line you'd be able to keep a car to actually *leave* the city or drive anywhere outside the exclusion area. That would make some people's commutes a lot more interesting and troublesome.
Not to say it doesn't work in some areas. Japan's major cities are laid out like this and everyone uses the train there just as in Manhattan. But there are also still a huge number of cars. Also, many cars outside of the cities because you generally need one or some sort of service to get you around where everything is more spread out in your city/town.
In areas outside of big cities-- yeah, not realistic. Autonomous farming equipment... yep, that's already happening. It's also far easier to code and build CPUs for those processes than it is to program Level 5 self driving cars and trucks. The more present disconcerting issue with farming equipment which could have been applied to anyone trying to self diagnose, repair or modify their own vehicles is John Deere claiming that their farmers are not allowed to do those same things to the farming equipment they legally own and paid for. Congress shot that down thankfully... for a few years until it comes up for review.
Last edited by KahnBB6; 11-09-17 at 10:45 PM.
11-10-17, 09:24 AM
#468
Lexus Fanatic
iTrader: (20)
I can actually see some reason why this might happen in an extremely dense part of a huge city. But for that to work the public transit would have to be as good as it is in Manhattan. And if a ridiculously vast no-vehicle area is drawn up it will force some people to move further away from where they are and consequently it will make properties for sale and rent closer to the exclusion line exponentially more expensive-- because outside of that line you'd be able to keep a car to actually *leave* the city or drive anywhere outside the exclusion area. That would make some people's commutes a lot more interesting and troublesome.
The more present disconcerting issue with farming equipment which could have been applied to anyone trying to self diagnose, repair or modify their own vehicles is John Deere claiming that their farmers are not allowed to do those same things to the farming equipment they legally own and paid for. Congress shot that down thankfully... for a few years until it comes up for review.
11-10-17, 05:15 PM
#469
on a related note, a lot of my family lives about 30 miles outside london england. they're all close to train stations with fast service into london. their house prices are OFF THE CHARTS. train fees aren't cheap either + you have the 'tube' (subway) on top. my brother-in-law commutes in every day, has done for years. the upside is he can get stuff done on the train.
Meanwhile, watch what will actually happen for values of the properties in areas where cars can drive and where parking spots or garages as part of them do exist. This speaker knows these realities very well but that isn't was being sold in the talk.
I wouldn't be so sure. That could have happened but it was shot down with a specific DMCA exemption. Now there is much greater and widespread awareness of the issue and the precedent it could set.
11-10-17, 07:44 PM
#470
LMFAO
https://techcrunch.com/2017/11/08/dr...ithin-an-hour/
A driverless shuttle set free in downtown Las Vegas was involved in a minor accident less than an hour after it hit the streets, reported the local NBC affiliate KSNV. Not really the kind of publicity you want, or that self-driving cars need.
The shuttle, an egglike 8-seater Navya, is operated by the AAA and Keolis. It was a test deployment along half a mile of the Fremont East “Innovation District,” so this thing wasn’t cruising the strip. Probably a good thing.
Now, it must be said that technically the robo-car was not at fault. It was struck by a semi that was backing up, and really just grazed — none of the passengers was hurt.
Like any functioning autonomous vehicle, the shuttle can avoid obstacles and stop in a hurry if needed. What it apparently can’t do is move a couple feet out of the way when it looks like a 20-ton truck is going to back into it.
A passenger interviewed by KSNV shared her frustration:
Surely this situation is not so rare that the shuttle’s designers did not allow for it? Moving the car out of the way of an oncoming vehicle seems like a pretty elementary safety measure.
A City of Las Vegas representative issued a statement that the shuttle “did what it was supposed to do, in that its sensors registered the truck and the shuttle stopped to avoid the accident.” It also claims, lamely, that “Had the truck had the same sensing equipment that the shuttle has the accident would have been avoided.”
Not if it failed to react properly, as arguably was the case with the shuttle. Testing will continue, but I have to say I wouldn’t get on this thing until they demonstrate that it can do more than just stop.
https://techcrunch.com/2017/11/08/dr...ithin-an-hour/
A driverless shuttle set free in downtown Las Vegas was involved in a minor accident less than an hour after it hit the streets, reported the local NBC affiliate KSNV. Not really the kind of publicity you want, or that self-driving cars need.
The shuttle, an egglike 8-seater Navya, is operated by the AAA and Keolis. It was a test deployment along half a mile of the Fremont East “Innovation District,” so this thing wasn’t cruising the strip. Probably a good thing.
Now, it must be said that technically the robo-car was not at fault. It was struck by a semi that was backing up, and really just grazed — none of the passengers was hurt.
Like any functioning autonomous vehicle, the shuttle can avoid obstacles and stop in a hurry if needed. What it apparently can’t do is move a couple feet out of the way when it looks like a 20-ton truck is going to back into it.
A passenger interviewed by KSNV shared her frustration:
The shuttle just stayed still and we were like, ‘oh my gosh, it’s gonna hit us, it’s gonna hit us!’ and then.. it hit us! And the shuttle didn’t have the ability to move back, either. Like, the shuttle just stayed still.
A City of Las Vegas representative issued a statement that the shuttle “did what it was supposed to do, in that its sensors registered the truck and the shuttle stopped to avoid the accident.” It also claims, lamely, that “Had the truck had the same sensing equipment that the shuttle has the accident would have been avoided.”
Not if it failed to react properly, as arguably was the case with the shuttle. Testing will continue, but I have to say I wouldn’t get on this thing until they demonstrate that it can do more than just stop.
11-10-17, 07:50 PM
#471
Lexus Fanatic
11-10-17, 07:52 PM
#472
There is a funny article by one of NYT's lunatics. I bolded out the most ridiculous part - the guy is clearly out of touch with reality.
https://www.nytimes.com/interactive/...-policing.html
STREET SIGNAGE IS the iconography of the automobile age. It’s like highly functional pop art: silhouettes of schoolchildren, white arrows, rectangular cries of WRONG WAY and, most central of all, the ubiquitous stoplight. The traffic light might be the first part of that iconographic world to be transformed, or vanish altogether, once we are fully in the age of autonomous cars. Robots, after all, won’t need signs to optimize the way they move through urban landscapes.
Urban-transportation experts have been busily creating computer simulations to show how this might work. In one model, each crossroads would have an “intersection manager,” a computer that senses the approaching traffic and uses wireless communication to talk to the oncoming cars. When each self-driving car is perhaps 300 yards away, it sends a request to the intersection manager — to turn right, say, or to move on through. The intersection manager then does an on-the-fly calculation to route that vehicle most efficiently, like an omnipotent and tireless traffic cop.
The result? A ballet of cars whizzing and weaving past one another in the intersection. Some slow down as they approach; others pass straight through. But crucially, compared with today’s intersections, many fewer cars come to a complete halt.
This could significantly speed up traffic throughout an entire city. Peter Stone, a computer scientist at the University of Texas at Austin who works with one model, has found that the “delay” time at intersections shrinks remarkably. “Right now, it takes me an average of 20 minutes to get to work, but with autonomous-car intersections, it might be half that time,” he says. Safety would be enhanced, too: Forty-three percent of car crashes in the United States occur at intersections, and Stone predicts that robot vehicles would crash only if there was a mechanical error. Better yet, autonomous intersections could produce an estimated 20 to 50 percent less carbon dioxide, because there would be fewer idling cars and jack-rabbit starts. “That’s the most expensive and most polluting part of driving,” says Remi Tachet des Combes, a mathematician who created robot-intersection models while at the Massachusetts Institute of Technology.
For the human passenger, though, a robotized intersection could be mildly terrifying — like flying through a crowded asteroid belt, trusting the A.I. to find the right path. “At first I think it will be freaky,” Stone admits. “Some people will need the window darkened so they don’t freak out.” But in the long run, we’ll probably shrug, get used to it and barely look up from our games of Candy Crush as we zip through. And pedestrians? They would probably push a button at the intersection to request their turn — or even use a smartphone app.
More subtly unsettling, however, might be the spectacle of a city devoid of stoplights. Indeed, devoid of all major street signs: no huge billboards across highways naming the exits, no complex merge instructions. Those signs are expensive to build and maintain. They’re designed for humans, and GPS-brained robots don’t need them to know where they’re going. Certainly, human pedestrians and cyclists will still need guideposts, but as Stone suspects, far fewer, and smaller, ones.
A world with almost no street signs would feel strange. It could make a city less cluttered and more attractive. But it might also leave us feeling unmoored. Social critics worry that GPS has already eroded our knowledge of the city; some studies have found that the more we rely on devices, the less we deeply intuit where we are and how to navigate on our own. “We become more helpless,” as Greg Milner, the author of “Pinpoint,” a history of GPS, told me. If robots rule the roads, we might get where we’re going a lot more quickly — but end up not knowing precisely where we are.
Clive Thompson is the author of “Smarter Than You Think: How Technology Is Changing Our Minds for the Better.”
https://www.nytimes.com/interactive/...-policing.html
STREET SIGNAGE IS the iconography of the automobile age. It’s like highly functional pop art: silhouettes of schoolchildren, white arrows, rectangular cries of WRONG WAY and, most central of all, the ubiquitous stoplight. The traffic light might be the first part of that iconographic world to be transformed, or vanish altogether, once we are fully in the age of autonomous cars. Robots, after all, won’t need signs to optimize the way they move through urban landscapes.
Urban-transportation experts have been busily creating computer simulations to show how this might work. In one model, each crossroads would have an “intersection manager,” a computer that senses the approaching traffic and uses wireless communication to talk to the oncoming cars. When each self-driving car is perhaps 300 yards away, it sends a request to the intersection manager — to turn right, say, or to move on through. The intersection manager then does an on-the-fly calculation to route that vehicle most efficiently, like an omnipotent and tireless traffic cop.
The result? A ballet of cars whizzing and weaving past one another in the intersection. Some slow down as they approach; others pass straight through. But crucially, compared with today’s intersections, many fewer cars come to a complete halt.
This could significantly speed up traffic throughout an entire city. Peter Stone, a computer scientist at the University of Texas at Austin who works with one model, has found that the “delay” time at intersections shrinks remarkably. “Right now, it takes me an average of 20 minutes to get to work, but with autonomous-car intersections, it might be half that time,” he says. Safety would be enhanced, too: Forty-three percent of car crashes in the United States occur at intersections, and Stone predicts that robot vehicles would crash only if there was a mechanical error. Better yet, autonomous intersections could produce an estimated 20 to 50 percent less carbon dioxide, because there would be fewer idling cars and jack-rabbit starts. “That’s the most expensive and most polluting part of driving,” says Remi Tachet des Combes, a mathematician who created robot-intersection models while at the Massachusetts Institute of Technology.
For the human passenger, though, a robotized intersection could be mildly terrifying — like flying through a crowded asteroid belt, trusting the A.I. to find the right path. “At first I think it will be freaky,” Stone admits. “Some people will need the window darkened so they don’t freak out.” But in the long run, we’ll probably shrug, get used to it and barely look up from our games of Candy Crush as we zip through. And pedestrians? They would probably push a button at the intersection to request their turn — or even use a smartphone app.
More subtly unsettling, however, might be the spectacle of a city devoid of stoplights. Indeed, devoid of all major street signs: no huge billboards across highways naming the exits, no complex merge instructions. Those signs are expensive to build and maintain. They’re designed for humans, and GPS-brained robots don’t need them to know where they’re going. Certainly, human pedestrians and cyclists will still need guideposts, but as Stone suspects, far fewer, and smaller, ones.
A world with almost no street signs would feel strange. It could make a city less cluttered and more attractive. But it might also leave us feeling unmoored. Social critics worry that GPS has already eroded our knowledge of the city; some studies have found that the more we rely on devices, the less we deeply intuit where we are and how to navigate on our own. “We become more helpless,” as Greg Milner, the author of “Pinpoint,” a history of GPS, told me. If robots rule the roads, we might get where we’re going a lot more quickly — but end up not knowing precisely where we are.
Clive Thompson is the author of “Smarter Than You Think: How Technology Is Changing Our Minds for the Better.”
11-10-17, 08:01 PM
#474
Lexus Fanatic
Yes, I know, but you are still one of the few people to see the truth about the self-drivers....that they may (?) eventually show promise years from now, but are still a long way off from that. They are certainly not perfected to the point that I'd want to depend on them. Indeed, at the time, I even had some qualms about reviewing an Infiniti Q50 Adaptive-Drive option, with the electronic steering. I chose a conventional-steer version instead, without that option....though, today, I'd probably be a little less-apprehensive about it, as that system has been on the road now, several years, without (apparantly) an major accidents attributed to it.
11-10-17, 08:05 PM
#475
Yes, I know, but you are still one of the few people to see the truth about the self-drivers....that they may (?) eventually show promise years from now, but are still a long way off from that. They are certainly not perfected to the point that I'd want to depend on them. Indeed, at the time, I even had some qualms about reviewing an Infiniti Q50 Adaptive-Drive option, with the electronic steering. I chose a conventional-steer version instead, without that option....though, today, I'd probably be a little less-apprehensive about it, as that system has been on the road now, several years, without (apparantly) an major accidents attributed to it.
Regarding self driving vehicles, the only way they can be truly viable is when they are controlled by computers with level of AI that fully matches human intelligence, and on top of it has sensors that are reliable in all weather and road conditions.
Last edited by Och; 11-10-17 at 08:08 PM.
11-10-17, 08:17 PM
#476
Lexus Fanatic
I'm not familiar how the adaptive drive option in the Q50 works, does it completely eliminate physical linkage between steering wheel and steering rack? In that case, it better be engineered to outlast the rest of the car, otherwise as these cars get older we'll start seeing accidents caused by failure of this system.
https://www.reuters.com/article/us-n...-idUSKCN0Z221O
https://jalopnik.com/the-infiniti-q5...get-1484200188
11-10-17, 11:41 PM
#477
Seems like the engineers of self driving cars actually have more common sense than people that believe in autonomous car lunacy.
https://www.wsj.com/articles/teslas-...ers-1503593742
PALO ALTO, Calif.— Tesla Inc. Chief Executive Elon Musk jolted the automotive world last year when he announced the company’s new vehicles would come with a hardware upgrade that would eventually allow them to drive themselves.
He also jolted his own engineering ranks.
Members of the company’s Autopilot team hadn’t yet designed a product they believed would safely and reliably control a car without human intervention, according to people familiar with the matter.
In a meeting after the October announcement, someone asked Autopilot director Sterling Anderson how Tesla could brand the product “Full Self-Driving,” several employees recall. “This was Elon’s decision,” they said he responded. Two months later, Mr. Anderson resigned.
In the race to develop autonomous vehicles, few companies have moved faster than Tesla, an electric-car pioneer that this year surpassed General Motors Co. as the nation’s most-valuable auto maker.
Behind the scenes, the Autopilot team has clashed over deadlines and design and marketing decisions, according to more than a dozen people who worked on the project and documents reviewed by The Wall Street Journal. In recent months, the team has lost at least 10 engineers and four top managers—including Mr. Anderson’s successor, who lasted less than six months before leaving in June.
Tesla said the vehicle hardware unveiled in October will enable “full self-driving in almost all circumstances, at what we believe will be a probability of safety at least twice as good as the average human driver.” The self-driving feature is subject to software development and regulatory approval, and “it is not possible to know exactly when each element of the functionality described” will be available, Tesla noted.
A Tesla spokeswoman attributed the turnover in the Autopilot group—which has more than 100 people developing hardware and software—to fierce competition for talent at large technology companies, conventional auto makers and startups. Tesla has hired more than 35 people to join its Autopilot group so far this year, and brought on more than 50 in 2016.
In an email, Mr. Musk said he was unhappy with previous Journal articles on the company. “While it is possible that this article could be an exception, that is extremely unlikely, which is why I declined to comment,” he wrote.
The hurdles to putting self-driving cars on the road on a mass scale are huge, but so are the potential rewards. Advocates say autonomous cars will help minimize congestion and pollution—and would likely make automobile travel much safer. Human error causes 94% of crashes, according to government statistics.
Companies working on self-driving technology—ranging from Ford Motor Co. and GM to Alphabet Inc.’s Waymo LLC and ride-hailing firms such as Uber Technologies Inc.—are dealing with a tricky balance. If they make their technology too proficient, drivers may be tempted to stop paying attention or take their hands off the wheel. Automobiles that drive themselves under all circumstances face more testing and government review. Most companies are initially releasing either semi-automated vehicles that require engaged drivers or autonomous cars that operate under restricted conditions.
Tesla decided to introduce semi-autonomous technology rather than wait because “when used correctly, it is already significantly safer than a person driving by themselves and it would therefore be morally reprehensible to delay release simply for fear of bad press or some mercantile calculation of legal liability,” Mr. Musk said last year.
Mr. Musk has positioned himself as a disruptive force across industries. The South African-born engineer, who made more than $100 million on an early investment in PayPal Inc., is chief executive of Space Exploration Technologies Corp., or SpaceX, which aims to colonize Mars. He said he is building a “Hyperloop” that will whisk commuters from New York to Washington in 29 minutes. In July, Mr. Musk announced Tesla is building the world’s largest lithium-ion battery storage project in Australia.
“Elon’s world is a tough world, and that’s fine with me,” said Ross Gerber, chief executive at Gerber Kawasaki Wealth & Investment Management in Santa Monica, Calif. “They’re hard-charging, trying to change the world. That’s why I invest in them.”
Tesla’s board includes James Murdoch, the chief executive of 21st Century Fox, which shares common ownership with Wall Street Journal parent News Corp .
Mr. Musk, who is 46 years old, became chief executive of Tesla in 2008, the year the startup produced its first electric sports car, the Roadster. He gained a reputation for working around the clock, sometimes spending the night on the factory floor.
In October 2014, Tesla introduced hardware and later added software updates that enabled capabilities such as automatic emergency braking and a collision-warning system—features some rival car makers had also begun incorporating.
Tesla engineers were already hard at work on a breakthrough intended to propel the company closer to the industrywide Holy Grail: putting a driverless car on the road.
Tesla dubbed its project “Autopilot,” after the technology that aids airplane pilots. In 2015, Tesla hired Robert Rose, a former software engineer at SpaceX, to run the Autopilot group.
The pressure was high to get testing done and develop the product. In a 2015 email, Mr. Rose urged engineers to get started on validating the technology, or to proving it works in tests, “RTFN”—right the f— now.
Some engineers and suppliers pushed back. Issues they debated included the amount of time a driver would be given to retake the wheel if a car’s autonomous driving features stopped working; mechanisms to keep drivers engaged; and whether the technology should be allowed on all roads.
Weeks before the October 2015 release of Autopilot, an engineer who had worked on safety features warned Tesla that the product wasn’t ready, according to a resignation letter circulated to other employees and reviewed by the Journal.
Autopilot’s development was based on “reckless decision making that has potentially put customer lives at risk,” the engineer, Evan Nakano, wrote.
Tesla declined to comment specifically on Mr. Nakano. “We actively encourage development teams and suppliers to highlight concerns and issues so that they can be comprehensively addressed during development,” a spokeswoman said.
Tesla said it based its design on millions of miles driven by employees and other early testers, followed by performance validation over millions of additional miles. It said Autopilot has been tested over more than 1 billion miles.
At least one early test drive was harrowing. In May 2015, Eric Meadows, then a Tesla engineer, engaged Autopilot on a drive in a Model S from San Francisco to Los Angeles. Cruising along Highway 1, the car jerked left toward oncoming traffic. He yelped and steered back on course, according to his account and a video of the incident.
On the same trip, he said police pulled him over for suspected drunk driving. He said he was sober and shot an email warning colleagues, “Do not use Autopilot this weekend.”
Mr. Meadows said he was later dismissed for what he was told were “performance issues.” Tesla declined to comment on Mr. Meadows but noted that the incident happened months before the release of the technology, giving the company plenty of time to work out problems that had been discovered during test drives.
As the team ironed out the technology, some enlisted the help of suppliers to settle disagreements. One engineer contacted MobileyeNV, the Israel-based company that made Autopilot’s cameras, and expressed fear that the equipment could be unsafe if used by drivers who weren’t fully engaged, according to people familiar with the matter. Mobileye Chairman Amnon Shashua contacted Tesla in May 2015 and was reassured that the technology would be deployed safely, according to a Mobileye securities filing.
As the release approached in October 2015, however, a Tesla engineer reported to Mobileye that the product was being released in a way that would allow the car to drive itself without hands on the wheel. Mr. Shashua flew to California and suggested precautions, a person familiar with the matter said.
Mr. Musk said “activation of Autopilot would be ‘hands-on,’ ” Mobileye said in the securities filing. “Despite this confirmation, Autopilot was rolled out in late 2015 with a hands-free activation mode.”
Tesla said drivers were “responsible for, and ultimately in control of, the car,” when it rolled out Autopilot.
Tesla owners’ manuals describe Autopilot as a collection of “driver assistance features” and note that motorists are responsible for staying alert, maintaining control of the vehicle and driving safely. The initial version of Autopilot warned drivers to retake the wheel if their hands weren’t detected.
Debates raged industrywide, as car makers and tech companies balanced technological advances that boost safety with the potential for dangerous misuse by drivers.
“This is the worst subject in the world to be adventurous with,” said Scott Keogh, Audi AG’s top U.S. executive, in an interview. The luxury auto maker plans to roll out fully self-driving cars in 2020.
Alphabet’s Waymo decided its autonomous system should be free from human interaction partly after its own employees using automated-driving technology became overconfident, engaging in dangerous behaviors such as taking their eyes off the road and reaching for briefcases.
“They were just human,” said John Krafcik, Waymo’s chief executive, speaking at a January automotive conference in Detroit. “They began to trust the technology.”
Shortly before the release of Autopilot in October 2015, Mr. Rose left the company. Tesla said Mr. Rose’s departure wasn’t due to any disagreement involving Autopilot but declined to elaborate further.
Mr. Anderson stepped into the job. A Ph.D. from Massachusetts Institute of Technology with a string of patents and papers on autonomous technology, he had helped launch the Model X.
Under Mr. Anderson’s leadership, engineers continued working on Autopilot improvements and other self-driving technology advances. They also had to deal with dark periods.
In May 2016, Joshua Brown, a former Navy Seal, activated the Autopilot system in his Tesla Model S while driving on a Florida highway. Autopilot didn’t see an 18-wheel truck crossing the road against a brightly lit sky, Tesla said. The vehicles collided and Mr. Brown was killed.
The following month, U.S. National Highway Traffic Safety Administration officials alerted Tesla that they were about to publicly disclose they were investigating Autopilot. During a conference call, Mr. Musk complained they were unfairly singling out Tesla for one incident when traffic fatalities claimed tens of thousands of lives annually in other companies’ vehicles, said people familiar with the exchange.
In September 2016, Tesla upgraded the system. It disabled automatic steering, preventing reactivation until the vehicle is parked, if a driver ignores repeated warnings to keep hands on the wheel. Mr. Musk said the update’s enhanced radar likely would have prevented the crash.
Mobileye and Tesla parted ways and Mobileye made public its concerns. Tesla said Mobileye had promoted the technology until Tesla began building competing equipment in-house, an account Mobileye disputes. Intel Corp. reached a $15.3 billion deal to buy Mobileye earlier this year, joining the driverless car race.
In January, U.S. traffic safety regulators closed their investigation into the May 2016 fatal crash, noting: “A safety-related defect trend has not been identified at this time and further examination of this issue does not appear to be warranted.” The probe found the truck should have been visible to Mr. Brown for at least seven seconds before the collision, and that Tesla made an effort during the design process to prevent drivers from misusing Autopilot. The investigation also found the rate of Tesla vehicles crashing had fallen nearly 40% since the company installed its automatic steering feature.
Still, regulators expressed worries about how automated-driving technologies are being marketed.
“We are concerned about drivers operating these vehicles having a good understanding of the capabilities and limitations of the systems,” then-NHTSA spokesman Bryan Thomas said at the time. “It’s not enough to put it in an owners’ manual and hope that drivers will read that and follow it.”
In October 2016, Tesla announced an upgrade of Autopilot. All new vehicles were being built with eight cameras that provide 360-degree visibility at up to 820 feet of distance. Tesla vehicles previously featured just one forward-looking camera, as well as a forward radar and 12 long-range ultrasonic sensors positioned to see 16 feet around the car in every direction.
The new cars still come with just one radar sensor, but it has enhanced processing to provide additional data about the driver’s surroundings. The radar can see through heavy rain, fog, dust, and even a car in front of it, according to Tesla.
Tesla also said it updated the cars’ 12 ultrasonic sensors to improve the distance at which they can detect hard and soft objects.
For $5,000, Tesla customers can buy an option called “Enhanced Autopilot.” That, Tesla said, would give them access to four of the car’s eight cameras as well as the radar, 12 sensors, and the new onboard computing system.
For another $3,000, drivers get the right to activate the rest of the cameras when Tesla enables a full self-driving system. Customers could wait to buy the option, but it would be more expensive.
The announcement shook up some engineers, because they believed that the product that was released wasn’t designed to be self-driving, according to several people interviewed.
The marketing was a factor in the decision by Mr. Anderson and at least two other engineers to leave the company, according to people familiar with the matter.
Mr. Anderson in December launched his own company, Aurora Innovation, with Chris Urmson, the former head of Google’s autonomous driving team. In a blog post, Mr. Anderson said the new company is developing self-driving technology “the right way.”
Buoyed in part by the promise of autonomous-technology enhancements, Tesla delivered more than 47,000 vehicles in the first half of this year.
In August, Tesla confirmed a new upgrade to its latest Autopilot hardware.
By the end of this year, Mr. Musk plans to demonstrate a car driving itself from Los Angeles to New York.
https://www.wsj.com/articles/teslas-...ers-1503593742
PALO ALTO, Calif.— Tesla Inc. Chief Executive Elon Musk jolted the automotive world last year when he announced the company’s new vehicles would come with a hardware upgrade that would eventually allow them to drive themselves.
He also jolted his own engineering ranks.
Members of the company’s Autopilot team hadn’t yet designed a product they believed would safely and reliably control a car without human intervention, according to people familiar with the matter.
In a meeting after the October announcement, someone asked Autopilot director Sterling Anderson how Tesla could brand the product “Full Self-Driving,” several employees recall. “This was Elon’s decision,” they said he responded. Two months later, Mr. Anderson resigned.
In the race to develop autonomous vehicles, few companies have moved faster than Tesla, an electric-car pioneer that this year surpassed General Motors Co. as the nation’s most-valuable auto maker.
Behind the scenes, the Autopilot team has clashed over deadlines and design and marketing decisions, according to more than a dozen people who worked on the project and documents reviewed by The Wall Street Journal. In recent months, the team has lost at least 10 engineers and four top managers—including Mr. Anderson’s successor, who lasted less than six months before leaving in June.
Tesla said the vehicle hardware unveiled in October will enable “full self-driving in almost all circumstances, at what we believe will be a probability of safety at least twice as good as the average human driver.” The self-driving feature is subject to software development and regulatory approval, and “it is not possible to know exactly when each element of the functionality described” will be available, Tesla noted.
A Tesla spokeswoman attributed the turnover in the Autopilot group—which has more than 100 people developing hardware and software—to fierce competition for talent at large technology companies, conventional auto makers and startups. Tesla has hired more than 35 people to join its Autopilot group so far this year, and brought on more than 50 in 2016.
In an email, Mr. Musk said he was unhappy with previous Journal articles on the company. “While it is possible that this article could be an exception, that is extremely unlikely, which is why I declined to comment,” he wrote.
The hurdles to putting self-driving cars on the road on a mass scale are huge, but so are the potential rewards. Advocates say autonomous cars will help minimize congestion and pollution—and would likely make automobile travel much safer. Human error causes 94% of crashes, according to government statistics.
Companies working on self-driving technology—ranging from Ford Motor Co. and GM to Alphabet Inc.’s Waymo LLC and ride-hailing firms such as Uber Technologies Inc.—are dealing with a tricky balance. If they make their technology too proficient, drivers may be tempted to stop paying attention or take their hands off the wheel. Automobiles that drive themselves under all circumstances face more testing and government review. Most companies are initially releasing either semi-automated vehicles that require engaged drivers or autonomous cars that operate under restricted conditions.
Tesla decided to introduce semi-autonomous technology rather than wait because “when used correctly, it is already significantly safer than a person driving by themselves and it would therefore be morally reprehensible to delay release simply for fear of bad press or some mercantile calculation of legal liability,” Mr. Musk said last year.
Mr. Musk has positioned himself as a disruptive force across industries. The South African-born engineer, who made more than $100 million on an early investment in PayPal Inc., is chief executive of Space Exploration Technologies Corp., or SpaceX, which aims to colonize Mars. He said he is building a “Hyperloop” that will whisk commuters from New York to Washington in 29 minutes. In July, Mr. Musk announced Tesla is building the world’s largest lithium-ion battery storage project in Australia.
“Elon’s world is a tough world, and that’s fine with me,” said Ross Gerber, chief executive at Gerber Kawasaki Wealth & Investment Management in Santa Monica, Calif. “They’re hard-charging, trying to change the world. That’s why I invest in them.”
Tesla’s board includes James Murdoch, the chief executive of 21st Century Fox, which shares common ownership with Wall Street Journal parent News Corp .
Mr. Musk, who is 46 years old, became chief executive of Tesla in 2008, the year the startup produced its first electric sports car, the Roadster. He gained a reputation for working around the clock, sometimes spending the night on the factory floor.
In October 2014, Tesla introduced hardware and later added software updates that enabled capabilities such as automatic emergency braking and a collision-warning system—features some rival car makers had also begun incorporating.
Tesla engineers were already hard at work on a breakthrough intended to propel the company closer to the industrywide Holy Grail: putting a driverless car on the road.
Tesla dubbed its project “Autopilot,” after the technology that aids airplane pilots. In 2015, Tesla hired Robert Rose, a former software engineer at SpaceX, to run the Autopilot group.
The pressure was high to get testing done and develop the product. In a 2015 email, Mr. Rose urged engineers to get started on validating the technology, or to proving it works in tests, “RTFN”—right the f— now.
Some engineers and suppliers pushed back. Issues they debated included the amount of time a driver would be given to retake the wheel if a car’s autonomous driving features stopped working; mechanisms to keep drivers engaged; and whether the technology should be allowed on all roads.
Weeks before the October 2015 release of Autopilot, an engineer who had worked on safety features warned Tesla that the product wasn’t ready, according to a resignation letter circulated to other employees and reviewed by the Journal.
Autopilot’s development was based on “reckless decision making that has potentially put customer lives at risk,” the engineer, Evan Nakano, wrote.
Tesla declined to comment specifically on Mr. Nakano. “We actively encourage development teams and suppliers to highlight concerns and issues so that they can be comprehensively addressed during development,” a spokeswoman said.
Tesla said it based its design on millions of miles driven by employees and other early testers, followed by performance validation over millions of additional miles. It said Autopilot has been tested over more than 1 billion miles.
At least one early test drive was harrowing. In May 2015, Eric Meadows, then a Tesla engineer, engaged Autopilot on a drive in a Model S from San Francisco to Los Angeles. Cruising along Highway 1, the car jerked left toward oncoming traffic. He yelped and steered back on course, according to his account and a video of the incident.
On the same trip, he said police pulled him over for suspected drunk driving. He said he was sober and shot an email warning colleagues, “Do not use Autopilot this weekend.”
Mr. Meadows said he was later dismissed for what he was told were “performance issues.” Tesla declined to comment on Mr. Meadows but noted that the incident happened months before the release of the technology, giving the company plenty of time to work out problems that had been discovered during test drives.
As the team ironed out the technology, some enlisted the help of suppliers to settle disagreements. One engineer contacted MobileyeNV, the Israel-based company that made Autopilot’s cameras, and expressed fear that the equipment could be unsafe if used by drivers who weren’t fully engaged, according to people familiar with the matter. Mobileye Chairman Amnon Shashua contacted Tesla in May 2015 and was reassured that the technology would be deployed safely, according to a Mobileye securities filing.
As the release approached in October 2015, however, a Tesla engineer reported to Mobileye that the product was being released in a way that would allow the car to drive itself without hands on the wheel. Mr. Shashua flew to California and suggested precautions, a person familiar with the matter said.
Mr. Musk said “activation of Autopilot would be ‘hands-on,’ ” Mobileye said in the securities filing. “Despite this confirmation, Autopilot was rolled out in late 2015 with a hands-free activation mode.”
Tesla said drivers were “responsible for, and ultimately in control of, the car,” when it rolled out Autopilot.
Tesla owners’ manuals describe Autopilot as a collection of “driver assistance features” and note that motorists are responsible for staying alert, maintaining control of the vehicle and driving safely. The initial version of Autopilot warned drivers to retake the wheel if their hands weren’t detected.
Debates raged industrywide, as car makers and tech companies balanced technological advances that boost safety with the potential for dangerous misuse by drivers.
“This is the worst subject in the world to be adventurous with,” said Scott Keogh, Audi AG’s top U.S. executive, in an interview. The luxury auto maker plans to roll out fully self-driving cars in 2020.
Alphabet’s Waymo decided its autonomous system should be free from human interaction partly after its own employees using automated-driving technology became overconfident, engaging in dangerous behaviors such as taking their eyes off the road and reaching for briefcases.
“They were just human,” said John Krafcik, Waymo’s chief executive, speaking at a January automotive conference in Detroit. “They began to trust the technology.”
Shortly before the release of Autopilot in October 2015, Mr. Rose left the company. Tesla said Mr. Rose’s departure wasn’t due to any disagreement involving Autopilot but declined to elaborate further.
Mr. Anderson stepped into the job. A Ph.D. from Massachusetts Institute of Technology with a string of patents and papers on autonomous technology, he had helped launch the Model X.
Under Mr. Anderson’s leadership, engineers continued working on Autopilot improvements and other self-driving technology advances. They also had to deal with dark periods.
In May 2016, Joshua Brown, a former Navy Seal, activated the Autopilot system in his Tesla Model S while driving on a Florida highway. Autopilot didn’t see an 18-wheel truck crossing the road against a brightly lit sky, Tesla said. The vehicles collided and Mr. Brown was killed.
The following month, U.S. National Highway Traffic Safety Administration officials alerted Tesla that they were about to publicly disclose they were investigating Autopilot. During a conference call, Mr. Musk complained they were unfairly singling out Tesla for one incident when traffic fatalities claimed tens of thousands of lives annually in other companies’ vehicles, said people familiar with the exchange.
In September 2016, Tesla upgraded the system. It disabled automatic steering, preventing reactivation until the vehicle is parked, if a driver ignores repeated warnings to keep hands on the wheel. Mr. Musk said the update’s enhanced radar likely would have prevented the crash.
Mobileye and Tesla parted ways and Mobileye made public its concerns. Tesla said Mobileye had promoted the technology until Tesla began building competing equipment in-house, an account Mobileye disputes. Intel Corp. reached a $15.3 billion deal to buy Mobileye earlier this year, joining the driverless car race.
In January, U.S. traffic safety regulators closed their investigation into the May 2016 fatal crash, noting: “A safety-related defect trend has not been identified at this time and further examination of this issue does not appear to be warranted.” The probe found the truck should have been visible to Mr. Brown for at least seven seconds before the collision, and that Tesla made an effort during the design process to prevent drivers from misusing Autopilot. The investigation also found the rate of Tesla vehicles crashing had fallen nearly 40% since the company installed its automatic steering feature.
Still, regulators expressed worries about how automated-driving technologies are being marketed.
“We are concerned about drivers operating these vehicles having a good understanding of the capabilities and limitations of the systems,” then-NHTSA spokesman Bryan Thomas said at the time. “It’s not enough to put it in an owners’ manual and hope that drivers will read that and follow it.”
In October 2016, Tesla announced an upgrade of Autopilot. All new vehicles were being built with eight cameras that provide 360-degree visibility at up to 820 feet of distance. Tesla vehicles previously featured just one forward-looking camera, as well as a forward radar and 12 long-range ultrasonic sensors positioned to see 16 feet around the car in every direction.
The new cars still come with just one radar sensor, but it has enhanced processing to provide additional data about the driver’s surroundings. The radar can see through heavy rain, fog, dust, and even a car in front of it, according to Tesla.
Tesla also said it updated the cars’ 12 ultrasonic sensors to improve the distance at which they can detect hard and soft objects.
For $5,000, Tesla customers can buy an option called “Enhanced Autopilot.” That, Tesla said, would give them access to four of the car’s eight cameras as well as the radar, 12 sensors, and the new onboard computing system.
For another $3,000, drivers get the right to activate the rest of the cameras when Tesla enables a full self-driving system. Customers could wait to buy the option, but it would be more expensive.
The announcement shook up some engineers, because they believed that the product that was released wasn’t designed to be self-driving, according to several people interviewed.
The marketing was a factor in the decision by Mr. Anderson and at least two other engineers to leave the company, according to people familiar with the matter.
Mr. Anderson in December launched his own company, Aurora Innovation, with Chris Urmson, the former head of Google’s autonomous driving team. In a blog post, Mr. Anderson said the new company is developing self-driving technology “the right way.”
Buoyed in part by the promise of autonomous-technology enhancements, Tesla delivered more than 47,000 vehicles in the first half of this year.
In August, Tesla confirmed a new upgrade to its latest Autopilot hardware.
By the end of this year, Mr. Musk plans to demonstrate a car driving itself from Los Angeles to New York.
11-11-17, 07:08 AM
#478
Lexus Fanatic
Have to agree. Musk may have bitten off more than he can chew. What's probably more important than self-driving cars, right now, for Musk, is simply finding a way to pay his present bills. It's going to look pretty silly if he has to follow GM (once) and Chrysler (several times) in going to Uncle Sam for a loan. One thing that will help, of course, is the rate-reduction for corporations that may be coming up in the tax-bill....dropping it to 15 or 20%, from the present 35%, will make a significant difference.
11-11-17, 11:28 AM
#479
Lexus Fanatic
I don't think self driving cars are going to be as successful as people think. Most people don't want to sit and let someone else do their travelling. People who buy cars enjoy driving otherwise they would take the transit. Also, from what I have seen over my life, very few people are able to drive the speed limit and not rush and take short cuts here and there while they drive. Most people will not tolerate self driving, speed limit and never taking the short cuts on their commute. There is something very special about driving that self driving can't replicate.
I do however think self driving will be a somewhat useful supplement to driving ever day.
I do however think self driving will be a somewhat useful supplement to driving ever day.
11-11-17, 01:26 PM
#480
Lexus Fanatic
I'm not completely opposed to self-drivers, assuming the technology to make them safe is successful. I can foresee at least some handy uses for them, particularly for those persons who are too old, too feeble, too forgetful, or too uncoordinated to drive themselves any longer, or don't want to trust themselves to risky cab or Uber drivers. Still, even with a self-driver, they have to be competent enough to be able to at least select the destination and program the car's GPS computer for the trip.