BikePGH!

Is It O.K. to Kill Cyclists?

This topic contains 119 replies, has 25 voices, and was last updated by  gg 1 mo, 2 weeks.

Viewing 40 posts - 81 through 120 (of 120 total)
 
Author Posts
Author Posts

salty

Private Message

Nov 18 2013 at 6:40pm #

Stu – that was one of the suggestions made by Action United, and it would certainly help, although 15 seconds probably isn’t long enough for everyone to get across the intersection given how outrageously wide it is.

Whining about pedestrians jaywalking is about as productive as motorists whining about cyclists running red lights. Make people wait for a ridiculous amount of time to cross the gargantuan intersection, put them in direct conflict with cars so it’s unsafe even when the “walk” sign is on (which, BTW, is only on for the bare minimum amount of time – if it comes on at all, assuming you went to push some out-of-the-way button at the proper time), and then wonder why people cross illegally. Start by making it easier and safer for people to cross the street and there won’t be as much “jaywalking” either. Give people a big middle finger, and good for them for giving one right back.


byogman

Private Message

Nov 18 2013 at 9:30pm #

In my own example, I’m not entirely convinced the time of one is prioritized over the other. If I yield to people jaywalking, I’m likely to get stuck at the intersection, which would include the time when the ped would have a walk signal.

You can try and splice apart things by scenario to try and get to some sort of greatest good for the greatest number. More power to those with that attentiveness, energy and good will. It’s a rare trio.

Give people a big middle finger, and good for them for giving one right back.

This, however, requires only one attribute, ill will, and with cause, and is thus perhaps a more reasonable thing to expect.

But most interactions are emotionless, steered by scripts written by social norms. And there’s particular logic to trying to ingrain the most basic norm of all. Watch out for the vulnerable. That is, if our goal is safety.

There’s a nice bonus after safety that you see most clearly in pockets of density. The means of travel that leave you most unprotected not coincidentally also mean you require the least space. So by making that safer and more convenient, encouraging travel with less or no adornment, you maximize available space for all and make it easier to get from A to B generally.

These two paragraphs are #1 and #2 justifications for cities to invest in transportation cycling. The same logic also prioritizes unwheeled pedestrians over us. I’m ok with that, even if some do cross right in front of me while I have the green and kill my momentum.


salty

Private Message

Nov 21 2013 at 1:31am #

I’ve been reading Jeff Speck’s “Walkable City”, which I highly recommend – it’s excellent (thanks to the ELDI walkability workshops for the reference). I just got to a chapter (and 2 pages in particular) where he makes a lot of the points I was trying to make, a lot more convincingly than I can. Pages 185-186 in particular talk about push-to-walk buttons, shortened pedestrian walk cycles, lengthened vehicle signal cycles, right on red (and right/left on green) and how it adds up to an environment that’s hostile for pedestrians and encourages jaywalking. It’s as if he were describing Penn Ave exactly (and other factors like lane and roadway width are discussed elsewhere). So, why is it allowed to happen?

BTW, he also cautions against the “Barnes Dance” (dedicated all-way ped crossing, named after the person that popularized them), because it also makes pedestrians wait longer. His explanation there makes perfect sense too – if you are walking from point A to point B, and they’re not on the same road, you have to cross streets in two different directions. So, when you reach an intersection, you can choose which way to cross based on the signal and often you don’t have to stop walking at all. If you have the Barnes Dance you are much more likely to have to wait. In fact that happens to me all the time crossing 5th and Negley – I can usually cross both streets with minimal or no stopping. He says there it can still make sense where pedestrian density is very high (so the Target intersection might be a candidate, or maybe at certain times of day).

Anyways, my basic gripe is how often I’m stuck staring at a “don’t walk” sign for absolutely no reason other than whoever designed the intersection decided to prioritize (turning) vehicles over pedestrians.


edmonds59

Private Message

Nov 23 2013 at 8:08am #

This is heart rending but hopeful. Not cyclists specifically, but maybe this is the tipping point, our “Stop de Kindermoord”. Maybe the winds are changing: http://parenting.blogs.nytimes.com/2013/11/22/how-could-one-thoughtless-driver-take-this-child/?smid=tw-share&_r=0
and http://pix11.com/2013/11/21/family-of-3-year-old-run-over-by-suv-plea-for-change/#axzz2lNLNHsNh


buffalo buffalo

Private Message

Dec 11 2013 at 7:40pm #

BTW, if you too would like to read Walkable City, Pitt and CMU each have a copy, and the local public libraries have several; this link may be instructive: http://www.worldcat.org/oclc/778422668

See if you can pick up a copy of Vanderbilt’s Traffic while you’re at it.


jonawebb

Private Message

Aug 19 2014 at 3:49pm #

FYI, autonomous vehicles are being programmed to speed — because it’s “safer”: http://gizmodo.com/googles-autonomous-car-is-programmed-to-speed-because-i-1624025227?utm_campaign=socialflow_gizmodo_twitter&utm_source=gizmodo_twitter&utm_medium=socialflow


byogman

Private Message

Aug 19 2014 at 4:19pm #

OK, so the in order to have license to run these vehicles google needs to get the state/municipality to admit publicly that speed limits are a sham?

I mean, they are in practice, but it’s, um, interesting to be making something where safety is one of the bigger potential issues, both upside and downside, around that and what the implications would be if that were admitted.

I’d liked the happy face on the car better when it promised some traffic calming to come along with. Enough of these and they BECOME the prevailing speed. And if you’re free to do whatever instead of man the steering wheel, how important are those couple of minutes?

Finally begs the question if it’s not admitted, what if the self driving car is pulled over for speeding? Thought google had enough of a legal headache already with self driving cars!

I can understand go with the flow, I honestly drive that way, but for this case and generally as a model for the future, much more bad than good.


rgrasmus

Private Message

Aug 19 2014 at 4:38pm #

This survey (http://bikeleague.org/content/share-road-20) may be of use to advocate against this. I’m not sure where at in CMU this meeting is happening, but it seems very interesting.


AtLeastMyKidsLoveMe

Private Message

Aug 20 2014 at 9:11am #

PaDOT uses the 85% rule to establish speed limits, in disregard of safety concerns. So yes, the tail does wag the dog.


jonawebb

Private Message

Aug 20 2014 at 9:34am #

I don’t think they’re as concerned about safety as their marketing suggests. While you can make an argument that following the traffic is safer, try running that past a police officer who pulls you over for speeding. They’re speeding because they can get away with it, and because if they strictly followed speed limits it would make the cars slower and less appealing. They are not creating autonomous vehicles as a humanitarian gesture; they are doing it so they can sell them, which means they have to go as fast as other cars.


Mick

Private Message

Aug 20 2014 at 10:08am #

I guessing the first time Google get a multi-million dollar judgment against them, they might decide to obey the law.

That, and I imagine some clever little legal finesse could get them for conspiracy, which is felony (even if the crime being conspired to break is not).

Commit a felony where somone dies? In some states that is murder one.


StuInMcCandless

Private Message

Aug 20 2014 at 10:09am #

I really think it’s time we took PennDOT to task about the 85% rule. That’s the wall we end up bashing our heads up against for everything we try to accomplish. It’s been in place for so long, probably decades if not most of a century (1930s?), nobody remembers why it was implemented in the first place.

The thinking behind what led to the 85% rule needs rethought, and a different rule needs to come out of it.


Mikhail

Private Message

Aug 20 2014 at 10:35am #

What is PennDOT interpretation of 85% Rule? Is is the famous rule that 85% of project take 15% of resources and the rest 15% of the project take 85% of resources?


StuInMcCandless

Private Message

Aug 20 2014 at 10:52am #

Discussed here in depth, the idea is that the speed limit on a road should be determined by the speed of traffic that would pass on it, absent a speed limit sign. Take the 85%th percentile of those drivers’ speeds, and bingo, that’s your speed limit.


jonawebb

Private Message

Aug 20 2014 at 11:16am #

@Mick, your faith in the sensitivity of corporations to tort law is touching.


reddan

Private Message

Aug 20 2014 at 11:30am #

Hmmm…at the risk of asking a dumb question, do the autonomous vehicles that have been prototyped thus far have the ability to read speed limit signs? How are they expected to know how fast to go, other than based on surrounding traffic?

[I am not in any way justifying them NOT obeying speed limits...however, the question of how an autonomous vehicle is supposed to understand such things is troubling to me, given the wide variety of circumstances under which speed limits vary...temporary signs, construction flagmen, inability to see signs due to intervening traffic, etc.]


WillB

Private Message

Aug 20 2014 at 11:31am #

I continue to be surprised at the hostility of bike/ped advocates to autonomous cars. For all of the possible pitfalls (dealing with regulations to make sure they comply with speed limits, building in strong right of way protections, etc.), it would be hard for them to be worse than what we have now. Human drivers are wildly, catastrophically irresponsible and inattentive; why wouldn’t we work to get rid of them as fast as possible?


jonawebb

Private Message

Aug 20 2014 at 11:46am #

@reddan I remember research from a long time ago on reading street signs, so that has probably gotten better. But I think they also know the speed limit on any given street from the map and GPS.
@WillB All I’m pointing out is the difference between “can be” and “will be.” Sure, autonomous vehicles can be safer than human drivers, no question. But they will be safer only if we require them to be. The market won’t push things in that direction; safety has never been a primary issue driving car sales in this country (how many years did auto makers resist the introduction of air bags?) Autonomous cars will be safer only if we, as a society, decide safety should be a bigger concern than it is now.


Mick

Private Message

Aug 20 2014 at 12:01pm #

Safety hasn’t been an issue for car becasue teh problems were individuals. We would never, ever accept the abismal safety record that is standard for cars for airplanes or trains.

Some gets killed by a robot programmed to exceed the lawful speed limit and you can bet Edgar Snyder would be salivating on Google’s big pockets. (and good for him!)

Possibly even if the robot were not exceeding the speed limit at the time of the incident. “Evidence of willful disregard of the law is relevant to corporation’s claim of being a reponsible entity and so has relevance to the case.”

While corporations may not give a fig for the letter and the sprit of the law, they are very, VERY sensitive to it’s enforcement.


reddan

Private Message

Aug 20 2014 at 12:02pm #

@jonawebb: my question revolves more around transient changes to speed limits, like construction zones, parades/marches/bike races, and other such temporary events. I’m curious as to how such things are to be handled by autonomous systems? (Not concern trolling, just curious.)


Marko82

Private Message

Aug 20 2014 at 12:26pm #

@redan, it’s a speed limit! so you never go above that speed. I don’t see how that changes with construction zones, weather, etc. It is the maximum speed to be driven under ideal conditions. So a robot would still be allowed to notice that traffic is going slower than maximum and adjust speed downward accordingly. But the speed limit is, well, the limit.


jonawebb

Private Message

Aug 20 2014 at 12:36pm #

@reddan, a little Internet searching suggests figuring out how to deal with changing conditions is an active research problem.
@Mick, well, see http://www.brookings.edu/research/papers/2014/04/products-liability-driverless-cars-villasenor for example. Products liability law won’t work to restrain the industry, and for the same reason it won’t work to push it to make the cars as safe as we would like.


byogman

Private Message

Aug 20 2014 at 12:41pm #

I think these are fine examples of the sorts of things (well except for obscured sight lines, but if it catches one frame that should be enough) that just need to be handle-able by any software allowed on the road.

I’d hope google is there already than all these details needing to be manually mapped before they can be used and thus risk being out of date.

I’d also hope some of the more truly tricky stuff gets sorted out (someone waves you ahead, difference between pedestrian eyes up and pedestrian eyes in cellphone, cyclist approaches a pothole).

And I think it will ultimately, and pretty well. The concern I have revolves more around what’s going on legally, and philosophically.

It’s true you can’t build a system capable of interacting in the real world purely based on rules, you need a hierarchy of behaviors. But speeding by default doesn’t belong in there. And given that it is right now, it kinda makes me wonder what else is in there, by default.

Roll through stops? Ignore a pedestrian in the crosswalk because they’re not THAT far into the crosswalk? Go with the flow around a curve even though visibility and ability to react is greatly reduced? There are a lot of shades of grey prioritizing speed, some darker than others. In relatively unusual (there’s no window during the light cycle where there isn’t a pedestrian in the crosswalk) or emergency (about to be hit otherwise) situations you need to be able to relax some of these rules. But by default? Please no.

Self driving cars can, and probably will make us all safer. But how much safer? Enough safer that people driven cars can be curiosities with no place in the city and my kids (well, ok, probably long tech curve, grandkids) can ride down the road on their bicycles without any reason to fear? Because THAT is what we should be pushing for, not just a more attentive version of the status quo.


Ahlir

Private Message

Aug 20 2014 at 12:46pm #

A couple of observations

1. I understood that automated vehicles will first appear on limited access roads. like the turnpike, with the goal of increasing throughput and safety. I can go for a system that manages tailgating and makes driving less susceptible to (say) drowsiness. I don’t completely understand using such systems on legacy roads in crowded urban areas (DARPA challenges notwithstanding.)

2. Vehicle speeds in cities are out of control. Why we bother posting speed limits is beyond me. Either enforce the damn limits or reduce the limit to 20mph (like in some places) and “tolerate” moderate speeding.


WillB

Private Message

Aug 20 2014 at 2:02pm #

“2. Vehicle speeds in cities are out of control. Why we bother posting speed limits is beyond me. Either enforce the damn limits or reduce the limit to 20mph (like in some places) and “tolerate” moderate speeding.”

This is why think automated vehicles would be GREAT in cities, because it would be much easier to enforce speed limits without having to dedicated police resources to it, i.e. you just create a regulatory framework that the manufacturers adhere to.

As to the other questions about such vehicles in urban spaces, my impression is that the google car has already been doing this pretty effectively, and has not had any meaningful accidents. Though admittedly, I haven’t read up on the details lately.


WillB

Private Message

Aug 20 2014 at 2:04pm #

also to @jonawebb’s point, I very much agree about the importance of building pedestrian/bike safety priority into the regulations. I’m actually optimistic because I think that’s a lot easier than getting lots of individual drivers to behave themselves


jonawebb

Private Message

Aug 20 2014 at 2:12pm #

Speed limits in cities are a great example of the conflict between “can be” and “will be” in autonomous vehicle design. Don’t you think people who commute to work via the Glenwood bridge would be reluctant to buy an autonomous car if they knew it meant that they would cross the bridge at the posted speed limit (25 MPH) instead of going 50, 60, 70 MPH like everybody else?
BTW on the question of autonomous vehicles dealing with changing conditions, see ex. http://www.technologyreview.com/news/529466/urban-jungle-a-tough-challenge-for-googles-autonomous-cars/ for a recent review.


WillB

Private Message

Aug 20 2014 at 3:30pm #

well that certainly pours a bit of cold water on my hopes for getting human drivers off of city streets in the next 30 years.


Pierce

Private Message

Aug 20 2014 at 4:03pm #

I know nothing about autonomous cars, but wouldn’t it seem likely that they’ll have an autonomous mode and like an override mode? A la that movie with Will Smith, when they want to say, speed over the Glenwood Bridge, they’d just take control and do it, and then let the car take over when they’re bumper to bumper again

(I know the Google article says they’re thinking of having just AI driven cars, but me thinks the industry would choose to have both)

I think it’s more important to have less individual driving overall as opposed to getting everybody to switch over to auto-cars. If society prioritized other modes of transport, that too would have a positive effect and I think would work on slowing traffic down.


Ahlir

Private Message

Aug 20 2014 at 4:21pm #

– Speed can be controlled by governors, though these days we presumably want things to be “smarter”, somehow.

— In that Tech Review article: “research would now focus on making vehicles that are 100 percent autonomous—leaving no room for error”. This makes me want to cry; also, worry even more about the state of higher education in this country.


Steven

Private Message

Aug 20 2014 at 4:47pm #

Current speed limits are based on human reaction speeds. Automated cars will have different reaction speeds, maybe slower, maybe faster, but probably not the same. Shouldn’t limits for them be based on what’s safe for them, not what’s safe if a human had been driving?


Ahlir

Private Message

Aug 20 2014 at 5:03pm #

True, stopping distances may end up being different. All the same, as a biker I would prefer not being buzzed at 60mph, no matter how smart the car or how fast it can screech to a halt.

But now you have me thinking about all that research on empirical morality. You know, car is driving down the street. Suddenly a child runs out in front; there is no time to stop. What do you do? Run the kid over or swerve into the cyclist riding beside you? And what if you could access information about the individuals? Would you factor their expected social utility into your decision? What if the driver’s utility is even lower? Do you explode?


salty

Private Message

Aug 20 2014 at 11:14pm #

Right, there is a whole lot more to setting the proper limit than reaction time and stopping distance when there are people in the vicinity. I don’t think it is reasonable to set the limit higher than 25 (or better yet 20) on city streets, even if the cars are “perfect”. Of course, we already have higher limits than that, with drivers who drive a lot faster than the limit with impunity… and kill people with impunity.

Limited access highways, who knows – maybe some day they’ll be able to crank it up to 500mph or something.

I hope the “programmed in” speeding only applies to the latter. I exceed the posted limit on highways myself – I normally set the cruise control to 6-10 over. OTOH I try to strictly follow the speed limits in the city and I will be very disappointed if self-driving cars do not do the same.

I also think autonomous vehicles are our best hope at improved road safety – relying on human drivers sure as shit isn’t working. I also hope it one day spells an end to private car ownership, although that may be more of a pipe dream… But it would be a gigantic change if we didn’t have to waste so much space on car storage.

In case it is not obvious, I am not speaking for my employer at all.


jonawebb

Private Message

Aug 26 2014 at 11:29am #

Driverless cars in D.C. (from CMU!): http://www.washingtonpost.com/local/trafficandcommuting/driverless-vehicles-even-in-dc-streets-an-autonomous-car-takes-a-capitol-test-run/2014/08/25/6d26baa8-06a4-11e4-8a6a-19355c7e870a_story.html?hpid=z1
Nothing about them speeding, though. Stage 2.


Vannevar

Private Message

Nov 3 2014 at 9:19am #

A really interesting paper from the Rand Foundation on autonomous cars.
Abstract: http://www.rand.org/pubs/research_reports/RR443-1.html
Paper: http://www.rand.org/content/dam/rand/pubs/research_reports/RR400/RR443-1/RAND_RR443-1.pdf

I found the discussion of insurance and liability starting on Page145 quite interesting. Specifically: in an autonomous car world, does liability for damage due to collision shift from the User to the Manufacturer?

thanks to @TheIguana for the link.


WillB

Private Message

Nov 3 2014 at 9:47am #

“Nothing about them speeding”

It did say later on in the article that this car won’t break the speed limit.

The part of the piece that most interested me was the claim that a “highway pilot” function will be coming to production cars soon, which will basically allow drivers to completely check out during highway travel. The ability for the cars to navigate city streets is much farther off. The highway auto-pilot, though, is likely to save thousands and thousands of lives, so hopefully that comes sooner rather than later.


jonawebb

Private Message

Nov 3 2014 at 9:59am #

There was a CMU startup that commercialized a version of the highway auto-pilot some time ago. They detected drowsy or distracted driving and sounded an alarm. The system was an add-on for trucks. It looks like they were sold to another company. The lead researcher seems to be back at CMU.

http://investing.businessweek.com/research/stocks/private/snapshot.asp?privcapId=27328559

http://www.supplier.takata.com/astras/html/en/index.html

http://deanpomerleau.tripod.com/


Vannevar

Private Message

Nov 3 2014 at 1:24pm #

man I haven’t seen a Tripod URL since (seems like) 1963.


Vannevar

Private Message

Nov 3 2014 at 5:01pm #


gg

Private Message

Nov 4 2014 at 12:28am #

Van, that could be any of us. Sad stuff.

Viewing 40 posts - 81 through 120 (of 120 total)
 

You must be logged in to reply to this topic.


FBshare-1008x1008

Support BikePGH with a year end gift, and make Pittsburgh a better place to live, work, and play.

×