[ennui2]

Not quite so clear cut.

Maybe a 10 year old PC might be junk but anything from maybe 2009 onward can still be viable. A big part of overall computing speed is due to factors OTHER than the CPU, namely memory, GFX and storage. For instance, I have a little Gateway laptop from 2009 that originally had Windows 7 on it. It was supposed to only support 4GB but that was probably because 8GB sticks weren't around at the time. Turns out it does support 8GB and it has SATA so I maxed out the RAM and stuck an SSD in it with Windows 10. It's now snappier than it's ever been and it's got HDMI so it can drive 1080P.

My two main desktops are still running motherboards circa 2010, one with a dual-core and one with a quad-core i5. Both are running SSD, one with 8GB RAM and one with 16GB. The main one has a pretty contemporary graphics card and the other one is still good enough to run Overwatch. Both also were some of the first motherboards with USB 3.

You've gotta ask yourself what you need all that power to do. Unless you're editing 4K video or running a render-farm, any CPU/motherboard from around 2010 onward is "good enough".

So even if the CPU doublings are tapering off, the fact is it's already mostly overkill for what people need. That wasn't the case during the early days of the runup from the 8088 through the pentiums. During that period, each generation of PC unlocked new exciting applications that couldn't have been done before due to the lack of CPU power.

To get back to self-driving cars, these are all a part of IoT (internet of things). Ubiquitous internet started with the first iPhone and it "changed everything"(TM). As mobile bandwidth increased (now in 4G), a lot of computing power could be offloaded to servers. Things like maps were much more responsive. Realtime traffic data, etc...

Also, as in the post above, gfx cards are becoming more and more general purpose processors. They're really like highly parallel DSPs or RISCs rather than an x86 CPU, but for certain classes of problem they can whip x86.

The fact that a Tesla software update on legacy hardware will probably go a long way to preventing more autopilot fatalities shows that it's NOT simply a matter of faster and faster hardware. AI is mostly a software thing. You can take the same hardware and make it perform a lot better just by finding better ways to use it.

Now, given all the above, how intelligent do simple doomer dismissals from someone (who last programmed back when Vanilla Ice was a thing) saying that self-driving cars will create a child-ball-chasing armageddon sound?

[kublikhan]

Maybe a 10 year old PC might be junk but anything from maybe 2009 onward can still be viable. A big part of overall computing speed is due to factors OTHER than the CPU, namely memory, GFX and storage. For instance, I have a little Gateway laptop from 2009 that originally had Windows 7 on it. It was supposed to only support 4GB but that was probably because 8GB sticks weren't around at the time. Turns out it does support 8GB and it has SATA so I maxed out the RAM and stuck an SSD in it with Windows 10. It's now snappier than it's ever been and it's got HDMI so it can drive 1080P

I did say 10 year old computer, not 6 year old processor. If you go through the trouble to upgrade the ram, SSD, gfx, you have already upgraded most of the core parts of the computer. Took awhile to convince the wife how much better SSD is but now that I finally got her to upgrade she swears by them. And I have nothing against running old hardware. I was still using a 10 year old macbook with original parts until it died recently. It's just a matter of making sure the machine can handle the load you are throwing at it. BTW, did you do the upgrade yourself? I asked about upgrading the macbook but they said it was too old and they don't even have the parts for it anymore.

You've gotta ask yourself what you need all that power to do. Unless you're editing 4K video or running a render-farm, any CPU/motherboard from around 2010 onward is "good enough".

My Macbook upgrade was because the hardware died and they didn't make the parts for it anymore. I did not need a cpu power bump because I was mostly using it as a front end for my HTPC. Of course I couldn't do 4K with it, but since I don't have any 4K content it wasn't an issue. As for my desktop system, I had to upgrade it because my old system did not meet the minimum requirements for the software I was trying to install(PC games). Not running a rendering farm or doing 4k video editing, however my video encoding time did reduce considerably after the upgrade. For mobile devices, I usually end up upgrading those because the latest software does not run on the old OS I am running and the new OS will not install on my old device. Similar problem happened in the Apple lawsuit. Devices got upgraded to run a new OS that was designed to run on new hardware and it runs like crap.

Now, given all the above, how intelligent do simple doomer dismissals from someone (who last programmed back when Vanilla Ice was a thing) saying that self-driving cars will create a child-ball-chasing armageddon sound?

I agree with you and evilgenius on this point. I don't see computer power being the obstacle to self driving cars.

[evilgenius]

I think the only reason why we haven't already seen Moore's Law die is because of multi-core processors. Remember back when everything was single core and the clock speed of them was about as fast as it could go?

Moore's law was not about clock speeds. It was about transistor density. Since the time of single core processors and today, transistor density continued to increase apace with Moore's Law. And clock speed alone does not determine how fast your processor is. Lower clocked processors from today are faster than higher clocked processors from 10 years ago:
Why are newer generations of processors faster at the same clock speed?

Or check out real world results. A 9 year old Athlon X2 3.2 Ghz getting stomped by a modern Intel Core i7 At 3.1 Ghz:
iTunes
Adobe Premiere

Old single core laptops from that day with higher front side bus speeds and a fast processor are still nice things to have around.

Maybe if you are running 10 year old software. Running modern software on a 10 year old computer is not going to be pretty. Things got so ugly that Apple actually got hit with a Class Action Lawsuit because when their old devices got upgraded to the latest OS version they slowed to a crawl. And that's if you are even allowed to install on old hardware.

All true. I was trying to point out the time rather than get into all that. The time when multi-core came along was an important inflection point. Interestingly, an article I read back then, espousing the virtues of multi-core and how it would effect the future of computing predicted that Moore's Law would cease to rule by about 2019. Obviously, they couldn't know about everything that was to come. I think they thought we would never get below 20nm.

The single core thing is not for the likes of you. It makes more of a statement about how you don't really need the fastest computer to do things like post on this site, or open a word document. Even granny might not like the performance of a 386, but she might get along with something that ended that generation of single core thought. I guess it depends upon how much keeping up with the Jones's people do, and how much they are forced by Father Time to do as well.

[ennui2]

It's coming whether some people (who remain nameless) believe or like it or not.

https://www.wired.com/2016/09/self-driv ... ittsburgh/

[KaiserJeep]

People need to remember that the self-driving car does not have to be perfectly safe. All it has to do is be safer than a human driver. Which almost certainly, it already is.

[kublikhan]

Whatever. Speed is a function of density. And density is done. For those who are not AI fanboys, it sure is easy to find this

I already debunked that the last time it came up. That article was from 2014. Since then Intel has released 14nm. Which not only is an increase in density, but at a lower cost as well:

Intel 14nm continues to deliver lower cost per transistor. 14nm Intel delivers >2x improvement in performance per watt. Moore's Law continues!

14nm Process Technology: Opening New Horizons

Concerns over the immediate end of Moore’s Law remain overblown and sensationalistic.

Intel is also reporting that they have been able to maintain their desired pace at improving transistor switching speeds and reducing power leakage. Across the entire performance curve the 14nm process offers a continuum of better switching speeds and/or lower leakage compared to Intel’s 22nm process. Here we can see how the last several generations of Intel’s process nodes compare across mobile, laptop, and server performance profiles. All 3 profiles are seeing a roughly linear increase in performance and decrease in active power consumption, which indicates that Intel’s 14nm process is behaving as expected and is offering similar gains as past processes. In this case the 14nm process should deliver a roughly 1.6x increase in performance per watt, just as past processes have too.

Furthermore, these base benefits when coupled with Intel’s customized 14nm process for Core M (Broadwell-Y) and Broadwell’s power optimizations have allowed Intel to more than double their performance per watt as compared to Haswell-Y.

The end result is that while Intel’s cost per transistor is not decreasing as quickly as the area per transistor, the cost is still decreasing and significantly so. Even with the additional wafer costs of the 14nm process, on a cost per transistor basis the 14nm process is still slightly ahead of normal for Intel.

The fact that costs per transistor continue to come down at a steady rate may be par for the course, but that Intel has been able to even maintain par for the course is actually a very significant accomplishment. As the cost of wafers and fabbing have risen over the years there has been concern that transistor costs would plateau, which would lead to chip designers being able to increase their performance but only by increasing prices, as opposed to the past 40 years of cheaper transistors allowing prices to hold steady while performance has increased. So for Intel this is a major point of pride, especially in light of complaints from NVIDIA and others in recent years that their costs on new nodes aren’t scaling nearly as well as they would like.

Intel’s 14nm Technology in Detail

[vox_mundi]

Would you fly in a pilotless airliner?

Driverless trains and cars are already with us – but how soon before pilotless airliners? The biggest challenge may be getting passengers on board.

The technology already seems to be there – drones are not a particularly new invention. Armies use drones the size of aeroplanes in war zones, where they are controlled remotely or loiter flying a pattern. Even helicopters can be pilotless, like the K-MAX, which is as big as a standard helicopter, and delivers aid supplies with stunning precision to dangerous locations.

Aeroplane accidents are rare today, but when they happen, they are getting harder and harder to solve, says Tim Robinson, editor-in-chief of the Royal Aeronautical Society’s magazine Aerospace. That’s why investigations often focus on ‘human factors’, identifying psychological and physiological issues as a probable cause.

“So with pilots relying on autopilots for 95% of today's flights, the argument goes, why not make the final 5% – take-off and landing – automated?” says Robinson. “Computers fly ultra-precise, repeatable trajectories, do not fly drunk, do not get tired, do not get distracted and so the thinking goes could be safer than human pilots in the future.”

Computers also don’t suffer the kind of mental stress which leads to incidents such as co-pilot Andreas Lubitz, who in 2015 flew a Germanwings passenger jet into a mountain in the French Alps. Or Malaysia Airlines Flight 370, or EgyptAir Flight 990

Autonomous Vehicles to Include Walmart Self-Driving Shopping Carts?

Video - Self-driving shopping carts may eventually show up at a Walmart near you. The U.S. Patent and Trademark Office recently granted Walmart a patent for shopping facility assistance systems, devices, and methods, according to CNN Money.

The patent details how the facility assistance systems could be implemented. The shopping carts would work with detachable motors, which would be equipped with multiple sensors and video cameras, and be controlled by a facility circuit system.

If a customer wanted a cart, for example, they could summon one through an app or an in-store kiosk. The nearest motor would then find the closest available cart, dock with it, and drive it to the customer’s location. At that point, the motor would detach and drive itself to a holding area for the next request.

The system could manage the entire shopping cart inventory, moving them around as needed. It would replace the jobs of the cart retrievers who currently walk to shopping cart corrals and move long trains of empty carts back to store entrances.

Walmart is cutting 7,000 jobs due to automation

... Walmart decision may be a bit more alarming for those in the workforce. As the Wall Street Journal reports, the most concerning aspect of America’s largest private employer might be that the eliminated positions are largely in the accounting and invoicing sectors of the company. These jobs are typically held by some of the longest tenured employees, who also happen to take home higher hourly wages.

Now, those coveted positions are being automated. The Journal reports that beginning in 2017, much of this work will be addressed by “a central office or new money-counting “cash recycler” machines in stores.” Earlier this year, the company tested this change across some 500 locations. “We’ve seen many make smooth transitions during the pilot,” said Deisha Barnett, a Walmart spokeswoman.

[vox_mundi]

Michigan senate passes bills supporting autonomous vehicles

Michigan's legislature is taking the next step to ensure that the state remains at the forefront of driverless car research. Last week, the Michigan Senate approved four bills looking to push forward autonomous vehicle research in the state.

The package now moves from the Republican-controlled Senate, where it has been approved, to the House. The bill will allow for testing of the cars on 122 miles of roads throughout the state. If passed, the bill will also allow for the American Center for Mobility to redevelop the Willow Run Airport for more autonomous vehicle testing and research. ...

Google wants to amend Michigan autonomous vehicle bills

Google is contesting legislation the Michigan Senate passed last week to make the state a hub for autonomous car research.

The technology giant sees some of the language as excluding companies like Google that are newcomers to making cars.

In a letter to Rep. Bradford Jacobsen, R-Oxford, and members of the House Communications and Technology Committee, Google executive John Krafcik, head of Google's Self-Driving Car Project, took issue with two provisions and asked that they be amended.

The first defines a "motor vehicle manufacturer" as a company that has distributed motor vehicles before participating in research within Michigan.

"One interpretation of that definition would exclude companies, like Google, that manufactures autonomous vehicles but do not currently sell them," Krafcik wrote.

The second provision requires autonomous vehicles operating in Michigan to be "supplied or controlled by a motor vehicle manufacturer."

"That could be interpreted to exclude vehicles supplied by a vehicle manufacturer that another company, like Google, modifies with automated driving systems," Krafcik said. "We urge you to consider these small but crucial amendments to the bills to ensure that investment and deployment of autonomous vehicle technology is not inadvertently discouraged in the state."

Self-driving cars will soon be tested in the streets of Boston.

Boston’s streets will soon become a test track for self-driving cars, with officials announcing that they plan to launch a study of autonomous vehicles by the end of the year.

The city said Wednesday that it was chosen by the World Economic Forum for a program focused on the future of urban transportation. Details on how the yearlong initiative will work are slim, but city officials said a key part of the partnership will be working with the Switzerland-based organization to develop policy recommendations for the fast-emerging technology. The Boston Consulting Group will also participate.

Behind the wheel of Uber's new self-driving car, which hits the road today in Pittsburg

On Monday, I found myself behind the wheel of a Ford Fusion driving down a Pittsburgh street lined with gorgeous Romanesque brick factories. The sky was a cloudless blue, the Allegheny River sparkled seductively through the trees, and I was able to drink in all the splendor of the day without distraction because I wasn’t actually driving the car. This was one of Uber’s new self-driving cars, and I was behind the wheel, admiring the view. That is, until the Ford SUV in front of me stopped without warning.

Time slowed down. My brain issued the signal to brake, but before my foot could respond, the car braked on its own. It was abrupt but gentle, the kind of stop that would have caused my wife to raise a disapproving eyebrow, but only slightly. There wasn’t enough time to be amazed.

Autonomous Intersection Management: Traffic Control for the Future

Autonomous Intersection Management (AIM) is a new intersection control protocol that exploits autonomous vehicles' extraordinary capabilities of control, sensing, and communication to make traffic management at intersections much more efficient than traditional control mechanisms such as traffic signals and stop signs. This video illustrates the principle behind this new traffic control protocol and demonstrates its potential using Marvin, the autonomous vehicle developed at the University of Texas at Austin.

[vox_mundi]

GlobalFoundries to invest over $2B to make new 7nm computer chip and ‘post-silicon’ computer technology

GlobalFoundries says it will invest more than $2 billion to produce cutting-edge 7 nanometer computer chips at its upstate New York facility.

The company said Thursday that research and development now underway will lead to the semiconductor being made at its Fab 8 facility in the town of Malta, north of Albany.

The microchip industry is developing the 7 nanometer chips as it continues work on smaller, more powerful chips. The task has become more challenging because of physical and technological limits.

GlobalFoundries expects to be able to produce the chips for the market in 2018.

"It’s not clear, for example, if IBM’s 30nm pitch @ 7nm will be smaller than what Intel debuts at that node," writes Joel Hruska at ExtremeTech. "IBM is claiming that 7nm will deliver a 50% area scaling improvement over and above 10nm, and at least a 50% power/performance improvement 'for the next generation of systems that will power the Big Data, cloud and mobile era.'"

Intel's been pushing its 14nm process for a couple of years now as the Core M-based Broadwell family of microarchitectures. Your computer is more likely based on Intel's 22nm process (Ivy Bridge) or even its 32nm process (Sandy Bridge), which power the Core i3 and i5 processors.

In its fiscal 2Q15 earnings release, Intel divulged its plans about forthcoming nm products. The company plans to launch 10-nm chips by late 2017. After Skylake, Intel’s third 14-nm product, Kaby Lake, will hit the market. Kaby Lake is built on the same foundation of Skylake, “with key performance enhancements.” Intel’s 10-nm model, Cannon Lake, is expected to come during the second half of 2017.

Meanwhile, TSMC and Samsung are currently in a race to get 10nm chips to market, which will likely be won sometime in early 2016. Based on TSMC's just-unveiled 10nm validation chip, the process should enable clock rates about 20 percent faster than 14nm chips, with 40 percent less power consumption. Intel's own 10nm technology (dubbed Cannonlake) is likely delayed in favor of pushing additional 14nm products.
http://motherboard.vice.com/read/ibm-an ... moores-law

... Rather than pushing silicon indefinitely, IBM's second research project will look at other materials and techniques that might more easily take us to 7nm and beyond. Materials such as III-V semiconductors (notably gallium arsenide, GaAs) have around 10 times the electron mobility of silicon, allowing for smaller transistors with much higher performance and lower power consumption. Likewise, IBM is looking into graphene and carbon nanotubes, both of which have incredibly high electron mobility and can (theoretically) be fashioned into very small structures. IBM has already created a carbon nanotube transistor with a 10nm channel that showed no sign of performance degradation due to its diminutive size (CNTs are just single layers of carbon/graphene rolled up into a tube, and are thus very, very small).

But beyond better materials and the ever-shrinking transistor, IBM is also looking into different methods of computation entirely, such as neuromorphic computing (brain-like chips - TrueNorth already in production), quantum computing, and silicon photonics/optoelectronics (optical tech built into electronic chips). These methods won’t necessarily provide more gigahertz or consume less power, but they could offer wildly more capable computers that can process much more data than a conventional computer in the same amount of time. Finally, it’s important to note that IBM is already working on all of these technologies, and in many cases has been for years.
http://www.extremetech.com/extreme/1860 ... chnologies

... Instead of trying to shrink the features on a 2D surface, just build more 2D surfaces. In effect, it's a return to Moore's original idea that die size should have an impact on silicon scaling. It's just that in this case the die space comes from adding layers.

Some manufacturers have already started down this path. IBM's server operation sees 3D integration as vital to the development of its largest computers. Apple's iPhone managed to increase the density of the memories it used by having them stacked and then packaged together. Micron Technology and Samsung are going further by building 3D memory stacks that offer massive bandwidth by drilling holes for electrical connections through the stack. This reduces parasitic interference from the extra capacitance and inductance of long bond wires that are normally used to connect chips within a stack to the package.

A cube of chips could contain the entire brains and memory of a tablet and an array of other embedded systems devices.

TSMC will begin 10nm production this year, claims 5nm by 2020

According to TSMC’s recent conference call, it expects demand for 20nm to drop sharply as companies transition to 16nm and FinFETs. TSMC is still predicting it will ramp its 10nm production in 2017 — TechEye is reporting that the company has already taped out hardware on 10nm — with every intent of deploying at a high market share and holding that share throughout the period. The foundry’s goal is to begin 7nm volume production by 2018, with 5nm in production by 2020.

Decades ago, the process node (say, 500nm, or 0.5 micron) referred to the half-pitch and the printed gate length, as shown in the chart above. That began to change in the late 1990s — today, there is no single metric that defines a “node.” Instead, both the pure-play foundries and Intel use the term to mean “the next collection of technologies that collectively deliver a significant improvement in power consumption, die size, and frequency scaling.” The chart below shows the difference between 14/16nm between Intel, Samsung, and TSMC: ...

Of course, simply delivering a process with smaller feature sizes doesn’t mean that the process is intrinsically better — Samsung’s Apple A9 processors are smaller than TSMC’s but consume more power.

[evilgenius]

You know that intersections exist like the automated one in the gif above today, without self-driving cars. I once sat at a cafe in Paris and watched one for about an hour. The French had come up with the idea that taking away all of the controls would allow people to come up with their own answers. Some person had convinced them. it was about as scary as watching the gif, but nothing bad happened while I was there. People just have to go by what they know about how they should drive. If they are too aggressive, or too timid, it messes things up, but the flow soon returns to normal.

I contrast this with something I observe driving in the US, which is that when people see someone else getting away with a bad example no matter how good a driver they might normally be they tend to follow that bad example. I don't mean in the case of timidity, but excessive aggression. I don't know if this is related to how the alpha chimp is always the most aggressive one, bearing upon how much driving in the US is about a competitive attitude toward other drivers, or not. When I've compared my Paris example with what I've seen in the US I've had to come up with the conclusion that it wouldn't work here for the same reason that most computer ERP systems fail: people. I can't say that's entirely true because aggressive driving seems to wrankle me more than it does other people. I could be suffering from some form of confirmation bias. Suffice it to say that it is possible that the organic nature of driving is not the only component of the driving environment. There is probably a cultural one as well. It's this cultural element which will offer the stiffest resistance to the adoption of self-driving cars in some places, I think.

[vox_mundi]

Self-Driving Cars Are Learning To Drive Using Grand Theft Auto V

There are problems with the way humans drive that could be solved by self-driving cars, and even though Uber is looking to bring more autonomous vehicles to the road very soon, the technology hasn’t been perfected yet, and self-driving cars still have plenty to learn. Somebody’s got to educate them, and it looks like that somebody will be Grand Theft Auto V, the same game in which you can try to jump a car into a cargo plane.

Researchers at Intel Labs and Germany’s Darmstadt University are less interested in the carnage the game offers, however, and are focusing more on the gigantic and realistic 3D world filled with real-life objects like cars, people, bicycles, and other things you’d encounter on your daily commute (via Jalopnik). So with that in mind, it makes sense that the game is being used to train self-driving cars to recognize objects, which is a lot more practical and safe than these vehicles doing this learning out in the real world.

For example, below is a video of a car being driven in-game by artificial intelligence, which learned using what is called a neural network. As you can see, it stays in its lane and avoids other objects on the road, which seems like a good start to us:

This means that once self-driving cars are on the road, they shouldn’t be running down pedestrians on the sidewalk and jumping over buildings to avoid the police, so that’s good.

Forget GPS, Civil Maps gives self-driving cars street smarts

We tend to think of maps as two-dimensional renderings giving us a god's-eye view of roads and terrain features, but Civil Maps takes a different approach. Its maps are more akin to how we understand our location when viewing street-level landmarks.

With the Civil Maps model, a self-driving car compares what its sensors perceive of the outside world, including lane markings and signage, with an internal map, and correlates that reference with GPS coordinates. If the car's sensors pick up a sign or other object not represented in its onboard map, it uploads information about the object to Civil Maps' servers. As more cars detect the object, Civil Maps can dynamically update its map model and transmit it to other cars.

The secret sauce behind Civil Maps' technology is what Puttagunta describes as "converting machine vision data to map vectors." LIDAR sensors on a car, which use lasers to determine distances between objects, generate what's called a point cloud. The point cloud defines all the objects around the LIDAR sensor by their three-dimensional shape. In addition, different frequencies in the laser data define different colors, such as the letters on a sign.

Breaking down robotaxi economics

The vision of many of us for robocars is a world of less private car ownership and more use of robotaxis — on-demand ride service in a robocar. In this world, what does it cost to operate these cars? How much might competitive services charge for rides? How much money will they make? What factors, including price, will they compete on, and how will that alter the landscape?

Here are some basic models of cost. I compare a low-cost 1-2 person robotaxi, a higher-end 1-2 person robotaxi, a 4-person traditional sedan robotaxi and the costs of ownership for a private car, the Toyota Prius 2, as calculated by Edmunds. An important difference is that the taxis are forecast to drive 50,000 miles/year (as taxis do) and wear out fully in 5 years. The private car is forecast to drive 15,000 miles/year (higher than the average for new cars, which is 12,000) and still have many years and miles of life left in it. As such, the taxis fully depreciate in this 5-year timeline, the private car only partly.


(You can view the spreadsheet directly on Google docs and download it to your own tool to play around with the model.)
...

Ford’s self-driving car is … totally normal

When friends and family asked, “How was it?” and “What was it like?” I couldn’t quite answer. I didn’t know how to answer. It didn’t feel like much at all. A few words come to mind, one of them: boring.

But no, this is a self-driving car. It’s the future. The future isn’t boring, is it?

Maybe boring is the wrong word. Safe?

It felt safe, and ordinary, and that’s not how I expected self-driving cars to feel.

In order for thousands of pounds of programmed steel and glass to chauffeur millions of humans this way and that, they can’t thrill passengers. They must feel safe, and predictable, and thus: relatively unexciting.

This is Ford’s wheelhouse. Ford oozes normal. You may find hordes of GM faithfuls chanting “found on road dead” and “fix or repair daily,” but if this car company, a pillar of American industry and capitalism, was a flavor, it would be something plain that everyone is OK with — vanilla.

Uber drivers go to war against self-driving cars

While Uber Technologies Inc. sees the launch of its self-driving car program as a step forward in transportation, its drivers are not so sure.

Jim Conigliaro Jr., founder of the Independent Drivers Guild, which represents 35,000 Uber drivers in the New York City area, said he found the launch “concerning” particularly because Uber has been working with regulators to allow ride-hailing in cities on the premise that it brings jobs to the community.

“It feels weird because you know that at some point in the future you won’t have a job, you just don’t know when,”

While Campbell sees a silver lining of sorts with Uber’s planned mixture of self-driving cars and drivers, he said he is not convinced by the argument that the autonomous vehicles will create potential maintenance jobs for the displaced drivers.

“I don’t think many Uber drivers, as they are kicked out of driving for Uber, will want to go maintain those cars,” he said.

[vox_mundi]

Lyft’s president says ‘majority’ of rides will be in self-driving cars by 2021

Lyft President and co-founder John Zimmer released a 14-page document today in which he predicts that by 2021, "a majority" of rides on its network will be in autonomous vehicles. Also by 2025, Zimmer says personal car ownership in US cities will be a thing of the past.

The end of car ownership will change cities in "huge ways," he said, echoing those experts and academics that predict streets and parking lots will be transformed into housing and open spaces with the mass adoption of autonomous vehicles. It will also change the daily experience of riding in a car, he said.

Cars have really taken over our landscape ... Every hundred years or so, you get an opportunity for a redo

Zimmer posted his manifesto, titled "The Third Transportation Revolution: Lyft’s Vision for the Next Ten Years and Beyond," on Medium on Sunday.

Lyft will operate as a hybrid network, with drivers using autonomous vehicles, until the technology advances to the point where drivers, and even steering wheels and pedals, are unnecessary. At that point, Zimmer envisions a future where passengers can hail a self-driving Lyft by paying by-the-mile or by subscription model similar to Netflix and Spotify

No Parking Here

... A 2011 study at the University of California-Berkeley found that the United States has somewhere close to a billion parking spots. Since there are only 253 million passenger cars and light trucks in the country, that means we have roughly four times more parking spaces than vehicles. If you totaled up all the area devoted to parking, it'd be roughly 6,500 square miles, 50% bigger than Connecticut.

Social critics often complain that the interstate highway system deformed the United States by encouraging sprawl. But the metastasizing of parking has had equally profound effects. On an aesthetic level, it makes cities grimly ugly. Economically, it is expensive to build. A study by the Sightline Institute found that at least 15 percent of the price of rent in Seattle stemmed from developers' cost of building parking.

Those costs are passed on to tenants whether they own a car or not (on top of any per space fee the landlord charges)—padding rent by an average of $246 a month in Seattle and $225 nationwide.

And worst of all may be the emissions that parking causes. Studies have found that anywhere from about 30 to 60 percent of the cars you see driving around a downtown core are just circling, looking for an open space to claim. (An IBM survey found that worldwide, urban drivers spend an average of 20 minutes per trip looking for parking.) When Donald Shoup, an urban-planning professor at the University of California-Los Angeles, examined just one small business area near his university—Westwood Village—he found that "cruising" for parking, as he dubs it, burns 47,000 gallons of gas and generates 730 tons of carbon dioxide a year. What's more, all that asphalt traps heat and raises the temperature of cities during the summer. Environmentally, aesthetically, and economically, parking is a mess.

... One study suggests a single self-driving car could replace up to 12 regular vehicles. Indeed, many urbanists predict that fleets of robocars could become so reliable that many, many people would choose not to own automobiles, causing the amount of parking needed to drop through the floor.

"Parking has been this sacred cow that we couldn't touch—and now we can touch it," says Gabe Klein, who has headed the transportation departments in Chicago and Washington, DC. He sees enormous potential—all that paved-over space suddenly freed up for houses and schools, plazas and playgrounds, or just about anything. "All that parking could go away, and then what happens?" he asks. "You unlock a tremendous amount of value."

Self-Driving Cars Coming In Months, Not Years

[vox_mundi]

This MIT website makes you decide who a self-driving car should kill in an accident

http://moralmachine.mit.edu

[vox_mundi]

The Third Transportation Revolution: Lyft’s Vision for the Next Ten Years and Beyond

... To measure the health of our transportation system, let’s start by looking into how much money we spend on car ownership and how often we actually use our cars. It may shock you, but Americans spend more than $2 trillion every year on car ownership — more money than we spend on food. What’s even more staggering is that for all the money we spend on them, the 250 million cars in America are only occupied 4% of the time. That’s the equivalent of 240 million of the 250 million cars being parked at all times.

For the most part, your car isn’t actually a driving machine at all. It’s a parking machine.

America is running a failing transportation business.

We can’t be this inefficient anymore, because we’re about to hit an inflection point that will strain our cities’ resources like never before. The U.S. already has ten cities with more than a million people. And our urban population is growing fast. By 2050, almost 100 million more people will move to American cities.

We don’t have enough space, housing, or public transit to accommodate this population influx, especially while keeping cities livable and desirable places to be. And while fixing transportation won’t solve all these problems, it certainly doesn’t help to continue devoting so much of our space to unoccupied cars.

... A full shift to “Transportation as a Service” is finally possible, because for the first time in human history, we have the tools to create a perfectly efficient transportation network.

Ridesharing is just the first phase of the movement to end car ownership and reclaim our cities. As I mentioned before, the shift to autonomous cars will expand dramatically over the next ten years, transforming transportation into the ultimate subscription service. ...

... Lyft secured a $500 million investment from General Motors in January and is working with the automaker to integrate self-driving technologies into its ride-hailing service. Zimmer says those cars are now being tested on the roads in San Francisco and Phoenix, Arizona.

[Outcast_Searcher]

This MIT website makes you decide who a self-driving car should kill in an accident

http://moralmachine.mit.edu

This is a rather interesting site, IMO. It's not a long enough test to be definitive (as the site says), but the trends are still interesting if you do a few run-throughs.

I find my answers seem to differ from the general public's in that I tend to value more human lives highly, and I always value a human life more than dogs and cats. (Peoples' values are rather interesting).

The main problem I have with the site is they assume that whenever a vehicle hits a concrete barrier to avoid pedestrians, that all passengers die. Given modern car technology, this is totally silly.

One has to wonder if they could do a better job with this as to me, it creates congnitive dissonance, and I wonder if it impacts answers from people who don't read carefully enough and assume that surely (for example) 4 passengers in a car will generally fare better than 4 passengers run over by a car.

...

I wonder about crowd-sourcing the "solution" (values) used to decide who should die when a bad crash is unavoidable, vs. arbitrary legislation from our elected representatives, or arbitrary decisions from carmakers and insurance providers.

One would be about dollars and avoiding legal repercussions. One would tend to be more emotional, but be a "wisdom of crowds" decision, if controlled for internet shennanigans (i.e. cheating by voters).

[vox_mundi]

Three crashes so far with Google self-driving cars in Chandler

The self-driving Lexus sport-utility vehicles from Google are taking a beating in Chandler, with three crashes caused by humans last month, including what is thought to be the first drunk driver to hit one of the test vehicles....

In two of the accidents, Google's drivers were manually operating the vehicles, and in the third, the Google vehicle was rear-ended while operating in autonomous mode, according to the Mountain View, Calif.-based company owned by Alphabet Inc.

Google officials said they wanted to bring the program to a new environment to test the cars' responses to things like dust storms and golf-cart crossings in the metro Phoenix area, and with Chandler's technology industry anchored by Intel Corp., the city was a good fit.

But the vehicles are finding that careless drivers are often more dangerous than weather or the quirks of autonomous technology.

That appears to be the case in an Aug. 16 crash where a Google vehicle being driven manually on Ray Road near McKemy Avenue at 42 mph was rear-ended by a vehicle traveling 67 mph.

The 25-year-old driver of the second vehicle was arrested on suspicion of driving under the influence of alcohol, said Chandler police spokesman Sgt. Daniel Mejia. According to the police report, a passenger in the Google vehicle was treated for a concussion.

In another crash while in manual mode, Google said data from the vehicle — which is equipped with radar, cameras and a technology called LIDAR that measures distances with lasers — indicates the other driver was 100 feet from the stop line when the light turned red, but proceeded to run the light.

Google reports that its vehicle was traveling 2 mph at the time of the crash, and the other car was going 45 mph.

Lauren Barriere, a spokeswoman for the Google project, said the accident would have been avoided if the Google car had been in autonomous mode.

"One of the most useful features of our system is that we can go back and see what the car would have done in self-driving mode," she said. "In this case, our software would have predicted that the other vehicle would enter the intersection after the light turned red. As a result, our car would have yielded to the other vehicle due to its high speed until it crossed through the intersection."

On Aug. 22, another Google Lexus, operating in autonomous mode, was rear-ended on Desert Breeze Road in Chandler while it was stopped at the intersection with Ray Road. There were no injuries.

Two other rear-ending collisions involving Google vehicles took place in August in California. One Google car was hit while in manual mode and the other while in autonomous mode.

Self-Driving Car Guidelines Call for Information Sharing

The Obama administration is set to unveil a policy framework Tuesday to govern the development of self-driving cars, including new ways for automakers to share information on emerging technology with the government, according to people familiar with the announcement.

The U.S. Transportation Department’s guidance is expected to include recommendations for states to pass legislation on introducing self-driving cars safely on their highways, according to three people involved in their development who spoke on the condition they not be identified until the information is made public.

The guidelines are also expected to clarify how its current rules and regulations, which were formed in the 1960s, will be applied to emerging technology.

Earlier this year, the Transportation Department said it would allow automakers that can demonstrate they have developed a safe autonomous vehicle to apply for exemptions to certain rules. It marked a new approach to auto regulations designed to ensure the government doesn’t stand in the way of technological progress.

Mark Rosekind, NHTSA’s administrator, has said the self-driving car plan would be key to the agency’s attempts to focus on human error, which the agency estimates is a factor in 94 percent of fatal car crashes. Those crashes killed more than 35,000 people in the U.S. last year.

Here's all of the self-driving tech stuff, companies are currently using on their cars

MIT and DARPA Pack Lidar Sensor Onto Single Chip

At MIT’s Photonic Microsystems Group, the authors are developing a lidar-on-a-chip system that is smaller than a dime, has no moving parts, and could be mass-produced at a very low cost for use in self-driving cars, drones, and robots.

Uber plans to open office in Detroit

Ride-hailing company Uber Technologies Inc. will soon open an office in Detroit.

Uber's vice president of global vehicle programs, Sherif Marakby, announced the facility at an engineering conference Monday in suburban Detroit. The office will help Uber collaborate with automakers and suppliers based in the area.

Ties between Detroit and Silicon Valley are growing as research into self-driving cars accelerates. Google's self-driving car project recently opened an office in the Detroit area, while NVIDIA, which makes computer processors, has software engineers embedded at automakers.

Uber has said it doesn't intend to make cars itself, so it needs automotive allies. The company recently inked a deal with Volvo to develop self-driving cars.

Why most self-driving cars will be electric

... First are the regulatory reasons, namely gas mileage requirements. Then there are engineering reasons — electric vehicles are easier for computers to drive. And, of course, ride-hailing services will increasingly make up a higher percentage of daily miles driven, and it will be easier, cheaper and safer to recharge an unmanned car than to gas one up.

And then there are the engineering reasons.

"There are a lot fewer moving pieces in an electric vehicle. There are three main components — the battery, the inverter and the electric motor," said Levi Tillemann-Dick, managing partner at Valence Strategic in Washington, D.C., and author of "The Great Race: The Global Quest for the Car of the Future." "An internal combustion engine contains 2,000 tiny pieces that have to be kept lubricated and they break every once in a while."

Glen DeVos, Delphi Automotive vice president of engineering, said most hybrids and electric vehicles are configured for drive-by-wire, steering-by-wire, brake-by-wire, systems that structurally are compatible with automated driving. The by-wire technology replaces traditional mechanical control systems with electronic control systems. This flexibility expands the number of options for the vehicle's design. Eliminating mechanical linkages can reduce weight. ...

Oxbotica, Oxford University’s Self-Driving Car, Has a Brain

Oxford University’s Oxbotica is a self-driving car with a brain. The brain, called Selenium, processes data and learns from mistakes to improve itself. The creators of the two-seater hope to compete with giants in the market like Google.

“We love those guys [Google], because they made the market possible,” Oxbotica’s founder Paul Newman told London’s Financial Times. “They introduced this brilliantly into the public’s conception of what’s coming in the future.”

Oxbotica’s Selenium brain has applications beyond its self-driving automobile. It’s already used in warehouse forklifts, mining trucks, and on NASA’s Mars rover.

Self-Driving Car Employees Are Worth About $10m… Each

When Uber bought Otto for $700m last month, people viewed it mainly as a “talent acquisition,” meaning the Ube was interested in bringing on Otto’s 70 employees, not necessarily the core technology.

Same goes for GM’s $1B acquisition of Cruise in March. They wanted Cruise’s 40 self-driving car experts to come kick it in Detroit and do smart people stuff.

Which means…
The going rate for self-driving car talent these days is at least $10m per person, an astronomical figure that reflects an exponential rise in demand.

According to Sebastian Thrun, the mastermind behind Google’s self-driving car, “Companies are desperate for talent. But the skill set to build a self-driving car is a multidisciplinary skill set [and] that broad skill set is just not there.”

That’s because there’s only one machine learning program in the world and it’s at Carnegie Mellon. And by the way, that program is “scrambling to recover” after Uber poached away 40 of its researchers.

Fun fact: Carnegie Mellon is in Pittsburgh, PA which is also why Uber’s first self-driving tests with Volvo launched there. ...

and something Pete can relate to ...

Self-driving cars: 40 per cent Indians will be too scared to be productive, says study

According to the study nearly 50 per cent of Indians said there is likely to be an increase in the frequency and severity of motion sickness in a self-driving car.

[vox_mundi]

The Department of Transportation just issued a comprehensive policy on self-driving cars

The US Department of Transportation is attempting to get ahead of the curve. On Monday, it released a surprisingly far-reaching "Federal Automated Vehicles Policy."

DOT views AVs as a safety technology that could reduce some of the 38,000 traffic fatalities a year in the US, 95 percent of which are caused by human error. It also sees AVs as an accessibility technology that could provide personal transportation to whole populations (disabled, elderly, etc.) who have lacked it.

The DOT is not neutral toward AVs. It wants to get them on the road soon. That’s a big deal.

The policy package is composed of four sections:

- Vehicle performance guidance for manufacturers, developers, and other organizations outlining a 15 point “Safety Assessment” for the safe design, development, testing, and deployment of highly automated vehicles, including a request that automakers sign and submit this safety assessment to certify that their vehicles are ready for public roads.

- Model for state policy that presents a clear distinction between Federal and State responsibilities and recommends policy areas for states to consider with a goal of generating a consistent national framework for the testing and operation of automated vehicles while leaving room for learning and valuable state discretion. (... this basically means that when humans are driving, states are in charge, but when software is driving, the feds are in charge. Automation is part of vehicle safety, and that is NHTSA’s domain.)

- Current regulatory tools that National Highway Traffic Safety Administration (NHTSA) can use to aid the safe development of automated vehicles, such as interpreting current rules to allow for appropriate flexibility in design, providing limited exemptions to allow for testing of nontraditional vehicle designs, and ensuring that unsafe automated vehicles are removed from the road.

- New tools and authorities that NHTSA could consider seeking in the future to aid the safe and efficient deployment of new lifesaving technologies and ensure that technologies deployed on the road are safe. For example, NHTSA is seeking public feedback on whether to consider pre-approving novel automated vehicle technologies before they are allowed on public roads and whether to create a new Federal safety standard for revolutionary vehicle designs, such as ones that do not require a steering wheel or gas pedal.

Simultaneously with this policy, NHTSA is releasing a final enforcement guidance bulletin clarifying how its recall authority applies to automated vehicle technologies. In particular, it emphasizes that semi-automated driving systems – ones in which the human continues to monitor the driving environment and perform some of the driving task – that fail to adequately account for the possibility that a distracted or inattentive driver-occupant might fail to retake control of the vehicle in a safety-critical situation may be defined as an unreasonable risk to safety and subject to recall. This ensures that companies and Americans understand our expectations around safety. If NHTSA determines that a vehicle is unsafe, it will use its authority to remove it from the road

[Outcast_Searcher]

You know that intersections exist like the automated one in the gif above today, without self-driving cars. I once sat at a cafe in Paris and watched one for about an hour. The French had come up with the idea that taking away all of the controls would allow people to come up with their own answers. Some person had convinced them. it was about as scary as watching the gif, but nothing bad happened while I was there. People just have to go by what they know about how they should drive. If they are too aggressive, or too timid, it messes things up, but the flow soon returns to normal.

--- SNIP ---

Suffice it to say that it is possible that the organic nature of driving is not the only component of the driving environment. There is probably a cultural one as well. It's this cultural element which will offer the stiffest resistance to the adoption of self-driving cars in some places, I think.

Evil, I've seen this idea now multiple places. And to me whether by gif or cartoon graphic, given the timing involved, my cautious, common sense, pain sensing human brain is very uncomfortable with this aspect of the change. At least if ALL the traffic is automated, then short of Skynet stuff, no emotions come into play.

My issue is simply: what if something goes wrong? Say mechanical failure as simple as a blown tire or a wheel coming off the lugs, or a child runs into the street (I presume an animal just gets run over). I'd want to see a LOT of evidence that things could be safely managed/coordinated to avoid dangerous accidents. (At least in a normal situation with safe following distances, hard braking and perhaps a quick lane change (given the room between the cars) is feasible, especially at electronic communication/decision making speeds. In the mess above? Call it cultural bias, self preservation, common sense, or weenie cowardice -- I'd want to see some SERIOUS and FORMAL data that this is truly safe.

This is not to say that things better than ordinary X-way stop signs and traffic lights couldn't be used right away. For example, in my experience, once folks are used to them, traffic circles work wonderfully.

[ennui2]

I will be waiting for the "child chasing after ball" apocalypse with great interest.



Methinks the wait will be akin to waiting for Godot.