urbanSTL

From Bloomberg: Musk's Secret Plan to Curb City Traffic With Self-Driving ‘Bus’

Tesla founder Elon Musk:
“We have an idea for something which is not exactly a bus but would solve the density problem for inner city situations,” Musk said Thursday at a transport conference in Norway. “Autonomous vehicles are key,” he said of the project, declining to disclose more. “I don’t want to talk too much about it. I have to be careful what I say.”

“I very much agree with solving the high-density urban transport problem,” Musk said. “There’s a new type of car or vehicle that would be great for that and that’ll actually take people to their final destination and not just the bus stop.”

http://www.bloomberg.com/news/articles/ ... riving-bus

It's worth noting that the 1906 San Francisco piece was at least partially staged. If you watch closely you'll see the same cars driving through the shot several times. The Miles brothers, whose new SF studio made the fim, were trying to create a more vibrant and modern streetscape as a kind of early promotion; to help boost both their city and their new studio. There's a number articles on the thing floating around out there, several of which are in the footnotes of the Trip Down Market Street entry on Wikipedia. That said, it's still a darn fine demonstration of the way traffic fundamentally worked at the time. And still often does when it's a mix of pedestrian and slower wheeled transport. It's only when the relative speeds of the slowest pedestrians and fastest motorized vehicles become so dissimilar that more control becomes critical.

I'm personally a little skeptical that you'll ever be able to mix purely human pedestrian and bicycle with autonomous motorized transport quite that densely, as humans simply don't behave as predictably as all that. We tend to turn around instantly without thinking when we realize we forgot something at the table in the cafe and then, whoopsoedaisy, there goes your perfect robot record. (Unless you build in considerably more space.) Doesn't even really matter that it will be human error that caused it. It's always the motorist at fault, as the robot driver will have the absolute responsibility to not hit people, much the same as a human driver now. That said, that could be contemporary traffic in a lot of places simply sped up a little.

This video shows light early morning traffic at Ben Thanh Market in Ho Chi Minh. Things get considerably slower and heavier later in the day, but when traffic is moving it works like this. And the narrator does a nice job of showcasing interactions:


I have myself been both pedestrian and passenger in some of this very traffic. You can see the same circle from the scooter's eye level at about 3:50 into my own video. And we weave through traffic much like everyone you see there from start to finish. (Some heavy, some light, some early, some late.)


Anyway, the traffic isn't completely implausible, but I'd bet their footage is sped up just a little. Of course, if the pedestrians have very strict instructions to walk at a constant speed and direction, and the breaks are really quite good . . . still, I'd want more space than that. Even after having been through a fair patch of southern Vietnam many many times now.

This fall Audi is rolling out it's traffic signal countdown timers in 4 cities. Las Vegas, Seattle, Washington DC, and Portland. LA Times says LA will have to wait at least 2 years per a spokesman for Traffic Technology Services. “There are 114 agencies in the L.A. area and we have to work with each one,” he said. “Las Vegas has a single agency.” So how many agencies does our metro area have for traffic signal control? Is this another area where St. Louis will have to wait for life saving car technology because of our fractured government structure?

Here is the Audi video. Pretty cool stuff. Available now in some 4 cities.

While it is cool technology, Iit seems like an excuse to concentrate on the time down of the light and gun it when its green. Then they are no longer focusing on who is crossing the street

pat wrote:While it is cool technology, Iit seems like an excuse to concentrate on the time down of the light and gun it when its green. Then they are no longer focusing on who is crossing the street

I have witnessed this in Vietnam. There are countdown clocks at every traffic light. As soon as it get's close the engines start revving. (And someone often jumps the gun.) Pedestrians are pretty much on their own. (Even though there's a big billboard campaign trying to tell drivers to respect pedestrians.) I don't have a particularly good recording, as I was using a mediocre point and shoot camera, but you can see a little bit of it at 1:27 in a video I made of Saigon traffic:



You can're really hear the quality of hundreds of scooters and small motorbikes simultaneously revving their engines, but it's an impressive sound. A bit like living through a two stroke Sturgis Rally every day.

But with the stoplight countdown timer in the car, one would presumably slow down approaching a stoplight so it will be green when you get there, unless another car is in front of you. So you wouldn't tromps it off the line in either case.


Sent from my iPhone using Tapatalk

I do that anyway, honestly. No rush to get to a red light.

Denver planning for no parking needed in the future. http://www.denverpost.com/2016/10/15/de ... ving-cars/


Sent from my iPhone using Tapatalk

At this week’s U.S. Conference of Mayors, the assembled mayors from around the country unanimously adopted a resolution I authored to support the integration of AV technology within public transportation. Transit agencies please take note.

Among various applications within public transportation, autonomous vehicle technology could provide an ideal solution to the “first and last mile challenge,” increasing usage of new and pre-existing rail trunk lines and making public transportation a more effective and convenient option for many residents.

A municipal autonomous shuttle system would provide on-demand, point-to-point service for commuters throughout the city or any defined local area. It would take cars off the streets and in so doing would a) reduce traffic and b) save lives. But just as importantly it would increase mobility and serve the function which transit agencies, with their multi-billion dollar budgets, are actually supposed to fulfill.

A MASS system would also allow almost seamless connectivity to existing rail lines, while making public transportation a particularly attractive option for relatively short, local trips.
https://blog.connectedcarsworld.com/the ... portation/

How do AVs take cars off the streets?

Funny discussion about waymo self-driving cars in Phoenix. And what it could mean for bikes and pedestrians.

https://m.facebook.com/story.php?story_ ... 2472048401


Sent from my iPhone using Tapatalk

^Not too bad. And when traffic slows down, because people can safely use the streets, maybe more people will take the train.

Bloomberg - Even After $100 Billion, Self-Driving Cars Are Going Nowhere

https://www.bloomberg.com/news/features ... ng-nowhere

I am surprised Bloomberg never even mentioned vehicle-to-vehicle communication which could solve a lot these self-driving car problems.  Here is a new piece on the status of V2V in NY Times on October 5, 2022.  They're looking at 2025 at the earliest.  This may be behind the paywall.  

https://www.nytimes.com/2022/10/05/busi ... =share-url

or

https://www.nytimes.com/2022/10/05/busi ... ology.html

Context

https://slate.com/technology/2022/12/sa ... taxis.html

I think this shows why we need to go to the next step and add car to car communication, and car to infrastructure communication.  Everything communicates 2 ways except cars.  Elon, where are you?  If the construction crew (or the traffic light computer) had been temporality set to send messages to all nearby cars that the road was closed, they could easily map a route around it.  

And if you only obstruct the forward windshield/cameras, it stops for safety, as a human would.  (Or a human might jump out in a rage and shoot you for doing that.). I suppose it could try to use other cameras and back up to find another route, as a human might.  Never buy software older than version 2.1.   : )

Slate - The Safety Dance

Autonomous vehicle companies claim that “humans are terrible drivers” and their tech is needed to save lives. Don’t buy it.

https://slate.com/technology/2023/07/cr ... times.html

The goal should be crashless cars, not self driving cars. And cars need to communicate with each other and with traffic lights. Everything links to everything except cars. Why?


Sent from my iPhone using Tapatalk

Or we could stop chasing the same dragon of "safe roads are around the corner, just keep driving and building places around driving (and making us rich) for the time being," we've been after for a century.

I really don't think you're gonna get rid of cars (autonomous or not) because the one thing a car will give you that nothing else will is privacy(meaning the freedom to not be around other people).  I would really like to see a district do an autonomous only experiment: close the district to all cars except compact car size autonomus shuttles, small autonomous cargo vehicles, and buses (all under one system so that they share information) and take drivers out of the equation completely.  The way these autonomous companies are flexing safety these days is a fake-it-till-you-make-it pose, but it wont be for long.  I don't think we're far from robots being better drivers than humans--especially when we're meeting them more than half way. 

STLEnginerd wrote: ↑

Jul 18, 2023

I do get the point though.  But reducing car use might be a harder nut to crack than wide spread implementation of autonomous cars.

Also on the optimistic side autonomy might encourage more efficient ways to use cars collectively thus reducing cars overall.

Reducing car use has been done, while wide spread implementation of autonomous cars is still quite uncertain. Even if the tech were ready, then people would have to be convinced they want it, then it would take quite a while for the auto fleet to turn over. It's a similar time scale as the effective efforts to reduce driving took.

Using cars collectively I am not optimistic abut. How icky to get in a car that some stranger was just in alone?