HomeDiagramsDatabaseMapsForumSkyscraper Posters
     

Go Back   SkyscraperPage Forum > Discussion Forums > Transportation

Reply

 
Thread Tools Display Modes
     
     
  #21  
Old Posted Jun 12, 2014, 4:51 AM
N830MH N830MH is offline
Registered User
 
Join Date: Jan 2011
Location: Phoenix, AZ
Posts: 1,847
What about Honda, Acura, Nissen, BMW and Luxus? Will they consider it? Will they have driverless?

Arizona, Georgia, North Carolina, South Carolina, Utah, Oregon, Florida, New Hampshire, Washington State, Ohio, Indiana, Tennessee, Rhode Island, Maine, Nebraska, Kansas, Missouri and New Mexico that they haven't signs the bills yet.
Reply With Quote
     
     
  #22  
Old Posted Jun 12, 2014, 7:33 AM
Hatman's Avatar
Hatman Hatman is offline
Registered User
 
Join Date: Aug 2013
Location: Salt Lake City, Utah
Posts: 973
Quote:
Originally Posted by N830MH View Post
What about Honda, Acura, Nissen, BMW and Luxus? Will they consider it? Will they have driverless?
Most of these companies are doing research and development on autonomous cars. Of these, Nissan is probably the most committed and farthest along:

http://www.torquenews.com/1080/nissa...h-isnt-problem

BMW, on the other hand, is doing... interesting things with their version of the technology:

BMW M235i prototype is world's first self-drifting car

Video Link


I cannot think of a single reason as to why this would be useful. Just typical BMW showing off, I guess. I love the worried look on the guy's face about 42 seconds in.
Reply With Quote
     
     
  #23  
Old Posted Jun 12, 2014, 8:03 AM
Hatman's Avatar
Hatman Hatman is offline
Registered User
 
Join Date: Aug 2013
Location: Salt Lake City, Utah
Posts: 973
I may be far too enthusiastic in pushing for autonomous cars, but every time I read something like this, I start to roll my eyes:

TAKING CONTROL OF AUTONOMOUS IMPLICATIONS
http://www.autonet.ca/en/2014/06/11/...utonomous-cars

Quote:
But what about the challenge of securing something that is more computer than car?

How will we protect the privacy of the vehicle's drive-route patterns and current whereabouts? Who will own that information - the owner or the automaker? If it's the latter, is it okay to sell that information?

For example, if there are two possible drive routes to the same destination, could a corporation pay the automaker to take the passenger past their place of business, in the hopes that they'll stop?

What about programming the car's software in the event of an accident? This programming is likened to a military targeting algorithm, which puts the self-driving car in a peculiar place, both legally and morally.

Protecting the car's software from attack will be imperative. Think that nightmare through: if the same software is used in all autonomous cars, and someone remotely accesses it, he or she now controls hundreds of fast-moving weapons in an urban environment.
But, thinking more slowly, I began to wonder how much of this is legitimately concerning.

1) Is route information really all that sensitive? I get that people don't want to feel like they have a tracker on their backs, but is having your route information logged in a computer somewhere (if this will even happen) really all that invasive?

2) Advertisers paying extra to have your car routed past their business. I cannot see this happening at all. My logic is this: How many billboards to you see along rail transit routes? Zero, because passengers in a train (or bus - or autonomous car) are not looking out the windows. They are looking at their phones or books or what-have-you's. Companies will realize very quickly that there's no sense in paying money to route someone past your business just so that you can ignore them - or worse, get annoyed at them for making your commute even longer. If any of this is even legal or possible at all.

Side note: Google was awarded a patent on a business model for allowing free transportation as part of advertising:


Source: http://www.theblaze.com/stories/2014...r-free-taxis/#
But this is a very different deal than paying to route someone's car. Especially if you own the car and it isn't a private taxi company.

3) The moral argument: If an autonomous car knows it is about to wreck with another car, and it knows it can crash in a way less damaging to itself but in a fatal way to the other car or it can equally distribute the risk between its own occupant and the occupant of the other car, what should it do?
Whatever the right answer should be, I have a feeling that legally the car will be required to protect its own occupants at all costs, no matter what will happen to the other car. This isn't really so different than what laws are in place right now - so why is this an issue for concern?

4) A hacker hijacking an autonomous car and using it as a weapon. I don't know a whole lot about hacking and hijacking, but I do know that autonomy implies an ability to make one's own decisions. If the car realizes it isn't acting independently, that someone is controlling it via the web, It should be able to detect this and shut down, right? Again, I'm not sure how easy or possible it is to hack into an autonomous (and presumably well-secured) system, but hacking, hijacking, and terrorism strike me as stuff out of TV dramas.

Just my thoughts.
Reply With Quote
     
     
  #24  
Old Posted Jun 13, 2014, 7:45 AM
Hatman's Avatar
Hatman Hatman is offline
Registered User
 
Join Date: Aug 2013
Location: Salt Lake City, Utah
Posts: 973
An article by Newsweek that I liked:
Google, You Can Drive My Car
http://www.newsweek.com/2014/06/20/g...ar-254475.html

Quote:
...
When a radical new technology arrives, at first we tend to think of it as a modification of an existing technology. Put a motor on a four-wheel chassis and all you’ve got is a carriage that doesn’t require a horse, right? Television seemed like radio with pictures. Cellphones seemed like telephones that could move around. Yet in each case, the new item opened up possibilities no one expected. Cars led to suburbs and shopping malls. Cellphones became pocket computers that are changing dating, banking, eating and just about everything else.

The first surge of autonomous vehicles probably won’t even carry humans. One of the most intense emerging battle zones in retail is same-day, nearly instant delivery—Wal-Mart and other brick-and-mortar retailers think they can fight Amazon by delivering orders from local stores in an hour or two. Amazon fired a shot back by saying it is working on delivery by drones that will land a package on your doorstep.

But the idea of drone delivery is wishful thinking, like a hangover-less bourbon. “The laws of physics still apply,” says Paul Saffo of Foresight at Discern Analytics. He doesn’t see how drones could ever carry enough packages to make the economics work, not to mention all the other attending problems. Who’s liable if the family dog attacks a drone, or when a sudden rain makes drones short out and drop pizzas on unsuspecting pedestrians?

What makes more sense for this upcoming battle? Autonomous vehicles built to drive up to your door with a package or food order and text you to come out and get it. To do that job, the vehicle doesn’t have to look like anything we’ve seen before. Throw out seats or headroom for a human. Make the things electric. Design something unique—maybe a cross between a U-Haul trailer and R2-D2. In a couple of decades, they will be whizzing all over city streets.
...
Reply With Quote
     
     
  #25  
Old Posted Jun 13, 2014, 7:49 AM
Hatman's Avatar
Hatman Hatman is offline
Registered User
 
Join Date: Aug 2013
Location: Salt Lake City, Utah
Posts: 973
Autonomy in the Military:
The Army's Drone Helicopters Will Airmail Driverless Cars to the Battlefield
http://motherboard.vice.com/read/the...he-battlefield

Quote:
Last month we learned that the military is developing a pilotless version of its iconic Black Hawk helicopter. Now we have some idea of how the new drone will be used.

For the next 18 months, a group of researchers will work to program the autonomous Black Hawk to deliver a driverless vehicle through the sky and lower it down on the ground where it will autonomously survey the land.

The car-toting robocopter will carry the all-terrain vehicle in a belly sling, scope out the best place to land, and then set the self-driving, sensing car free to search the area for dangerous contaminants.

"All of this will occur out of direct sight, without human intervention and without putting human lives at risk," explained a press release announcing the project, called Extended Operational Reach with Autonomous Air and Ground Vehicles.
...
Reply With Quote
     
     
  #26  
Old Posted Jun 13, 2014, 7:55 AM
Hatman's Avatar
Hatman Hatman is offline
Registered User
 
Join Date: Aug 2013
Location: Salt Lake City, Utah
Posts: 973
Lastly, this encouraging article from FORTUNE:
The end of driving (as we know it)
http://fortune.com/2014/06/12/end-of-driving/
Reply With Quote
     
     
  #27  
Old Posted Jun 17, 2014, 8:12 AM
Hatman's Avatar
Hatman Hatman is offline
Registered User
 
Join Date: Aug 2013
Location: Salt Lake City, Utah
Posts: 973
DUTCH UNVEIL FIVE-YEAR PLAN FOR SELF-DRIVING TRUCKS

http://www.autoworldnews.com/article...ing-trucks.htm

Quote:
Dutch officials have unveiled a plan to put self-driving trucks on the road from Rotterdam to other cities in the next few years, detailing steps that include computer simulations and truck tests on a closed track.

In a letter explaining the proposal, Infrastructure and Environment Minister Melanie Schultz van Haegen said the Netherlands is analyzing traffic laws to pave the way for testing autonomous cars on public roads, Reuters reported.

The infrastructure minister will submit a law by early next year to allow self-driving vehicles to be tested and plans to outline specific roads and conditions conducive to testing next year, according to the Associated Press.

The five-year plan was submitted to parliament by a collective of several industry and research groups, including Transport and Logistics Netherlands, DAF Trucks, Rotterdam Port and the Netherlands Organization for Applied Scientific Research (TNO).

"We want to do the first demonstrations in the beginning of next year and roll out the trial in a controlled environment as soon as possible," said Bastiaan Krosse, a spokesman for TNO, as quoted by Reuters.

While other European nations have launched similar projects, the Dutch proposal is unique since "no other project has a hard target of bringing this to market within five years, with the backing of the government," Krosse said.

To start, self-driving trucks would be tested through computer stimulations; real-life vehicles would later be put through their paces on a closed track prior to testing on public roads. The plan is for self-driving trucks to deliver goods from Rotterdam, Europe's largest port, within five years.

"There are countless benefits," to switching to autonomous trucks, said Marianne Wuite, spokeswoman for the Ministry of Infrastructure and Environment, as quoted by Reuters. "Self-driving cars need less space and therefore use asphalt more efficiently; they avert traffic jams and reduce accidents. They are also more environmentally friendly."
Reply With Quote
     
     
  #28  
Old Posted Jun 17, 2014, 8:20 AM
Hatman's Avatar
Hatman Hatman is offline
Registered User
 
Join Date: Aug 2013
Location: Salt Lake City, Utah
Posts: 973
How to get people to trust autonomous cars: Humanize them.

The Psychology Of Anthropomorphic Robots

http://www.fastcodesign.com/3031825/...morphic-robots

Quote:
...
Trusting the car to drive safely is among the biggest hurdles to the ascent of the autonomous car. It's an odd fear, considering how terrible human drivers are, but a natural one. I can attest first-hand that anxiety melts away when you actually ride in Google's self-driving car. But making consumers comfortable enough for a test drive (test ride?) will be a challenge for car makers preparing for that not-so-distant day when driverless cars hit the showroom.

[...]

The work of Waytz and colleagues suggests that simple anthropomorphic design elements might do the trick. In one recent study, the researchers recruited 100 test participants to operate a driving simulator through two courses. Some drove a normal manual simulator. Some operated a semi-autonomous simulator capable of controlling its own speed and its steering. Some operated a semi-autonomous car with a name (Iris), a gender (female), and a voice (pre-recorded human audio files).

Not only did test participants humanize Iris--they rated her as smarter and more capable of feeling, anticipating, and planning than the other simulators--they also trusted her more. In self-reports, participants operating Iris said they felt safer in the car and more willing to give up control, compared to those in the normal simulator. Their bodies confirmed the feeling: Heart-rate monitors displayed a lower change in arousal for Iris drivers, compared to both other simulator groups.
...
Reply With Quote
     
     
  #29  
Old Posted Jun 17, 2014, 2:02 PM
emathias's Avatar
emathias emathias is offline
Adoptive Chicagoan
 
Join Date: Sep 2007
Location: River North, Chicago, Illinois
Posts: 4,323
Quote:
Originally Posted by Hatman View Post
...
2) Advertisers paying extra to have your car routed past their business. I cannot see this happening at all. My logic is this: How many billboards to you see along rail transit routes? Zero, because passengers in a train (or bus - or autonomous car) are not looking out the windows.
...
While a lot of people are occupied with phones/books/whatever, a lot do look out the windows, and there are some billboards along rail routes in Chicago. The reasons there aren't more are probably more associated with these reasons:

1) Frequency of service - some commuter rail routes have pretty low frequency of service
2) Sightlines - even on high-frequency routes, it can be hard to locate a billboard in a place where it can be read from a train.
3) Use permits - local regulations about advertising sometimes prevent the construction of billboards in areas that would otherwise be effective.
Reply With Quote
     
     
  #30  
Old Posted Jun 18, 2014, 8:00 AM
Hatman's Avatar
Hatman Hatman is offline
Registered User
 
Join Date: Aug 2013
Location: Salt Lake City, Utah
Posts: 973
Quote:
Originally Posted by emathias View Post
2) Sightlines - even on high-frequency routes, it can be hard to locate a billboard in a place where it can be read from a train.
Okay, fair point. Passengers on rail transit look out the side of the train, not the front, so it's harder to make a billboard as visible. Fair point.
However, I still think billboards are going to slowly disappear as more and more people stop operating their cars and start watching TV on the go and the likes. And I still think the fear of corporations sponsoring routes that go past their businesses is absurd.
Reply With Quote
     
     
  #31  
Old Posted Jun 18, 2014, 8:11 AM
Hatman's Avatar
Hatman Hatman is offline
Registered User
 
Join Date: Aug 2013
Location: Salt Lake City, Utah
Posts: 973
While reading this article...

Driverless cars are only going to change just about everything

http://business.financialpost.com/20..._lsa=d366-a2f1

... I came across this video:

Video Link


If you can't see it, it basically notes the various infrastructure improvements that can be made possible with autonomous cars. These include:
1) Increased parking capacity, as parked cars can move out of each other's way in the garage in order to let other cars in or out.
2 Increased roadway capacity due to thinner (aka more) lane space. I'll continue to argue that the notion itself of lanes is too restricting, and that autonomous cars will fill the road space in the most efficient way possible, like a flex-road.
3) Increased grade-separation.
This is fascinating to me. Autonomous cars won't need overpasses/underpasses to be as wide as they currently are, nor do they need them to be so over-built. If you make a bridge for the weight of passenger cars, you can program all heavier vehicles never to go over it. Also, you can program all vehicles to go slower over the bridges, meaning they won't need to be as rigid. Lastly, because a computer is driving, sight-lines are no longer necessary, meaning that grades and crests can be much more pronounced, and that the lead-up to the bridge doesn't need to be as long.
All of this means that grade separation via bridges won't be nearly so expensive as it is now. Hence, I predict that with fully autonomous roadways, we will begin seeing much more grade separating at intersections and pedestrian crossings.
Reply With Quote
     
     
  #32  
Old Posted Jun 18, 2014, 4:40 PM
LouisVanDerWright LouisVanDerWright is offline
Registered User
 
Join Date: Jul 2012
Posts: 3,634
Quote:
Originally Posted by Eveningsong View Post
It would be nice if there are security personnel in driverless buses in 2075. Just so the riders could feel safe
Don't worry, by 2075 the autonomous vehicles will be equipped with tazers and sleeping gas to protect passengers and subdue unruly humans...

Quote:
Originally Posted by Hatman View Post
Finding a Better Word for 'Autonomous Car'

http://www.citylab.com/tech/2014/05/...us-car/371718/
That's a funny thought; the phrase "Autonomous Car" is the "Horseless Carriage" of our day...
Reply With Quote
     
     
  #33  
Old Posted Jun 18, 2014, 9:45 PM
ardecila's Avatar
ardecila ardecila is offline
atomic
 
Join Date: Jun 2006
Location: the city o'wind
Posts: 12,604
Quote:
Originally Posted by Hatman View Post
How many billboards to you see along rail transit routes? Zero, because passengers in a train (or bus - or autonomous car) are not looking out the windows.
There are plenty of these in Chicago along the elevated tracks. Not to mention all the ads inside the trains/buses, which can get pretty elaborate.
__________________
la forme d'une ville change plus vite, hélas! que le coeur d'un mortel...
Reply With Quote
     
     
  #34  
Old Posted Jun 19, 2014, 7:38 AM
Hatman's Avatar
Hatman Hatman is offline
Registered User
 
Join Date: Aug 2013
Location: Salt Lake City, Utah
Posts: 973
In the news today:

An article in favor of Autonomous Cars:
Driverless Cars Are Capable of Reducing Urban Traffic by 80 Percent
http://guardianlv.com/2014/06/driver...by-80-percent/

Quote:
Driverless cars are projected to be capable of reducing urban traffic by 80 percent. The fully automatic automotives could be the solution to the congestion problems in the United States, Asia, or any other place that experiences dense traffic on a daily basis.

Carlos Ratti, director of SENSEable City Lab run by MIT, said that the most exciting thing about these innovations is less cars on the road. He cites the utilization of smart technology in the vehicles’ software to make near perfect maneuvers and decisions. Ratti said that Asia is looking into the development of driverless cars, but the United States is much more involved in this development; with California recently issuing approval for tech companies to test and expand the technology with the intent of public use.

The combination of the vehicles’ abilities to communicate with one and another, and the increase of driverless taxis and buses will contribute to the lessening of traffic and accidents. The Singapore-MIT alliance for Research and Technology is currently testing a fleet of driverless cars created from golf carts in the National University of Singapore’s campus.

Nanyang Technological University is also experimenting with its own autonomous vehicles. The vehicles currently shuttle passengers between JTC Corporation’s CleanTech Park and NTU’s campus. Analysts believe as the technology becomes more apparent in the culture, cities will adapt quickly. The driverless cars being more efficient will mean people will reach places currently deemed too remote for businesses; subsequently, cities will expand to cover these areas.

Parking is a major contributor to congestion in metropolitan areas, driverless cars are capable of solving this problem and reducing urban traffic by 80 percent. A significant amount of traffic in cities is from drivers searching for an ideal parking spot (proximity to destination, payment, space). Google described that their driverless cars will be able to drop their passenger(s) off at their desired location, then go off on their own to locate a parking spot. The owner of the vehicle would then be able to call the car back to them via mobile app. This also enables the optimization of space utilization in urban environments.

The driverless cars also don’t employ an action called “rubbernecking.” Rubbernecking is when a driver of a vehicle slows down in order to observe an accident or situation on the road out of curiosity. Rubbernecking, according to a report from Brookings Institution, accounts for a quarter of traffic delays. Robotic cars are incapable of this.

It wouldn’t take a lot of driverless cars to be capable of more than just reducing urban traffic by upwards of 80 percent. According to Purdue University’s civil engineering professor Fred Mannering, if one in 10 U.S. vehicles were driverless, traffic fatalities and injuries, fuel consumption, and travel time would fall significantly. In total, saving $40 billion yearly. If half of the vehicles were driverless the savings would reach $200 billion.

Driverless cars have been tested with success by Google, MIT, and other technology companies. They estimate that in five years the vehicles will be fully incorporated into the public.

By Andres Loubriel
Also an article of fear:

Autonomous Vehicles Will Bring the Rise of 'Spam Cars'
http://motherboard.vice.com/read/aut...e-of-spam-cars

Quote:
he ultimate goal of self-driving cars is to remove the driver altogether, a move that'll free up the roads and create an entirely new aspect to the sharing economy. But the move towards automation could also inadvertently lead to the creation of a fleet of annoying, constantly-roving, empty vehicles. Spam cars, if you will.

It's an idea that was raised by Reddit user CombustibleCitrus on the SelfDrivingCars subreddit, and it's one that serious thinkers in the industry have already pondered:

"If driverless cars become common, I wonder if there's the possibility that some types of unmanned vehicles could become a nuisance. Will our neighborhoods be patrolled by driverless ice cream trucks, food trucks, vending machine trucks and Google Streetview cars all day long? What about driverless billboard trucks? What's to stop them from clogging up the highways during rush hour?," the Redditor wrote.

...
To this articles's credit, it goes on to show research from the same MIT team from the first article showing how these fears are almost entirely unfounded. The idea is that human drivers are already pretty cheap to hire - so autonomous cars - which are more expensive to buy but cheaper to operate - will hardly effect the economics of 'spam cars.'
Reply With Quote
     
     
  #35  
Old Posted Jun 19, 2014, 10:28 PM
Stryker Stryker is offline
BANNED
 
Join Date: Sep 2012
Posts: 3,559
I don't understand why this is so hard to understand, it'll be very much like your modern taxi. With one big hitch being a fraction of the cost, and far more predictable.

Once you think about it in those terms I believe its a lot more obvious.

The bigger question is how automated bus services might work. The biggest problem with bus services is the lack of comfort, inability for bus drivers to accounts for riders, and the wait times.

The singular taxi is good and all, but theres little desire for a car that can go 140 on the freeway, technically it's too expensive.

However high speed express bus services could be the norm.
Reply With Quote
     
     
  #36  
Old Posted Jun 20, 2014, 5:18 AM
Hatman's Avatar
Hatman Hatman is offline
Registered User
 
Join Date: Aug 2013
Location: Salt Lake City, Utah
Posts: 973
Quote:
Originally Posted by Stryker View Post
I don't understand why this is so hard to understand, it'll be very much like your modern taxi. With one big hitch being a fraction of the cost, and far more predictable.

Once you think about it in those terms I believe its a lot more obvious.
I think it is more complicated than that.

As a civil engineer, I enjoy thinking about the changes that autonomous cars can bring to our urban infrastructure. Infrastructure is simply a reflection of how people want to live - most people want to have a car and the personal mobility it brings, so our US infrastructure consists mostly of roads and parking lots. Autonomous cars sever the link between personal mobility and owning/operating a private vehicle. I think people will adopt autonomous technology pretty quickly, because who likes driving in trafic jams, or paying car insurance, or taking your car to a repair garage? Drivers are already so bored; they are already trying to text/eat/sleep/anything-except-focus-on-driving because the actual task of keeping your car between the painted lines is not what they like about driving. They like the freedom of mobility.

So once these cars are accepted enough that manually-driven cars are phased out, how does that change the infrastructure that was designed for human drivers? How do those changes affect the people living/working nearby? We've talked on this thread about densification of cities due to the in-filling of parking lots, and we have talked about what transit services may be replaced or enhanced by autonomous taxis (and autonomous buses too). But there are other things as well. Social changes, such as everyone, no matter how young, having access to the mobility previously reserved only for those old enough for drivers licences, or where illegal drug markets that used to be done in parked cars and parking lots will go (I'm trying to find the article that brought these points up without much success).

When the needs of the people changes, then the infrastructure changes too. How the people then react to those changes is the most interesting part.
Reply With Quote
     
     
  #37  
Old Posted Jun 20, 2014, 7:08 PM
eschaton eschaton is offline
Registered User
 
Join Date: Dec 2013
Posts: 1,578
Honestly, thinking about how cities will be retrofitted to deal with autonomous cars isn't too difficult, because it mostly means cities becoming similar to what they were like in 1900. Yes, there will be autonomous car traffic on the roads, but there will be little danger to pedestrians, and little reason for parking. Just start demolishing garages (both residential and commercial), narrowing roads, and filling in parking craters and you'll get most of the way there. Hell, once conversion towards automated systems is universal, you could begin seeing an argument towards eliminating highways, as AI should be able to drive on surface streets at near highway speeds (effective, not constant) without putting pedestrians in danger.

A deeper, and more difficult question, is what suburbs will start looking like, since they were built from the ground up with cars in mind, and there's going to be a lot less cars in the future.

Clearly garages (particularly modern 2-3 garage monstrosities) won't be needed for the most part. Some suburbanites might desire ownership of a car, and keep it in their garage, but others won't have any. Garages will have to be retrofitted to new uses - likely junk storage - and new construction in the suburbs will potentially have very different typologies than today.

Beyond this, I'm guessing "legacy" suburbs will remain largely the same as they are today. There's things you would likely not do in an autonomous car world, like 15-foot setbacks and cul-de-sacs, but we are stuck with them without heavy re-engineering. And I don't expect that self-driving cars will actually make people walk much more than today, so the trend towards walkable communities won't pick up dramatically. Cities may become places the car is (mostly) banished from, but suburbs won't change much besides home design.
Reply With Quote
     
     
  #38  
Old Posted Jun 20, 2014, 7:22 PM
Wizened Variations's Avatar
Wizened Variations Wizened Variations is offline
Registered User
 
Join Date: Apr 2008
Posts: 1,611
I wonder about the time when there are both autonomous and human controlled cars on the road.

Let's say an autonomously controlled car and a human driven car have a collision.


1) Is the rider of the autonomously driven car responsible for unforeseen bugs in the car algorithms?

Or in a differing vein:

2) What effects on human resumption of control would major electronic flux do to autonomously driven cars? What would the legal liabilities to, say Google, be? Would the definition of "Acts of God" have to expand?

There are others.
__________________
Good read on relationship between increasing number of freeway lanes and traffic

http://www.vtpi.org/gentraf.pdf
Reply With Quote
     
     
  #39  
Old Posted Jun 24, 2014, 7:09 AM
Hatman's Avatar
Hatman Hatman is offline
Registered User
 
Join Date: Aug 2013
Location: Salt Lake City, Utah
Posts: 973
^^^^ @ Eschaton

I think we are thinking along the same lines - except that I live in Salt Lake City, which is 95% suburbs. Suburbs, we can agree upon, stand to change the most by the rise of autonomy.
Reply With Quote
     
     
  #40  
Old Posted Jun 24, 2014, 7:21 AM
Hatman's Avatar
Hatman Hatman is offline
Registered User
 
Join Date: Aug 2013
Location: Salt Lake City, Utah
Posts: 973
A Little Startup Called Cruise From The Socialcam, Twitch Founders Is Tackling Self-Driving Cars Too
http://techcrunch.com/2014/06/23/cruise/

Quote:
Kyle Vogt, one of the founders behind Justin.tv, Socialcam and Twitch, is getting back to his undergraduate research roots in autonomous vehicles with a new self-driving car startup called Cruise. He’s recruited a team of engineers and roboticists from MIT.

The kicker here, though, is that Cruise isn’t making its own self-driving robotic cars in the way that Google has been approaching the space. Cruise has built a system that you can mount on an existing car and use to drive it down highways. It’s more than just standard cruise control, but it isn’t an entirely self-driving car, either.

[...]

ruise’s system has three components. There are sensor units that go on top of the car near the windshield and then there are actuators that control the steering and driving. There’s also a computer that goes in the trunk and takes up about one cubic foot of space.

When you drive onto a highway and merge into a lane, you’ll be able to hit a ‘Cruise’ button on your dashboard. The system will take control of the car’s steering, braking, and acceleration to keep you in your lane. But the system doesn’t take the place of a human driver, so you have to stay alert to make sure you’re following all traffic laws.



Video Link
Reply With Quote
     
     
This discussion thread continues

Use the page links to the lower-right to go to the next page for additional posts
 
 
Reply

Go Back   SkyscraperPage Forum > Discussion Forums > Transportation
Forum Jump


Thread Tools
Display Modes

Forum Jump


All times are GMT. The time now is 10:14 AM.

     

Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2017, vBulletin Solutions, Inc.