Not far from where I live, there's a traffic light that regularly backs up, especially the left turn lane. The navigation apps started suggesting a shortcut around that light that involved a left turn cutting through a neighborhood on a gravel street and arriving at very busy street with poor sight lines and just a stop sign. Eventually the city put up plastic bollards to block the left turn.
UPS figured out decades ago that left turns in urban settings are usually suboptimal. Why don't nav applications allow drivers to avoid left turns except at intersections with left arrows?
It’s apparently used to apply an additional routing penalty for “difficult intersections”, assigned by map editors, which provides a robust way to avoid these situations, including difficult left turns: https://www.reddit.com/r/waze/comments/ffypgb/avoid_difficul...
A parking lot of a busy supermarket trivially handles the foot traffic, easily funneling all of it through a handful of choke points without any (pedestrian) traffic jams. No, the only time you see a traffic jam in a parking lot is when the cars are moving, because the density is so much worse.
> You do realize that the traffic will be moving at the speed of walking then, or ?
walkable paths do not usually overlap with cars/bicycle lane/bus lanes.
but they are very effective to reduce the use of vehicles for routine activities (grocery, schools, catching public transport, meeting friends, going out for a drink, etc.)
Also traffic speed is actually not that relevant, in Rome, where I live, most of the time cars move at an average speed lower than 10 km/h, the real issue is avoiding congestions, not making cars move faster.
15-20 km/h is more than enough to do what you have to do in a reasonable time.
where I live in Rome to reach the station I can either take the subway (5 stops, 11 minutes), the bus (~20-25 minutes), a cab (~15-20 minutes) or my car ~30-35 minutes and forget about parking it.
fast tracks for cars can be moved outside of the city, like the great ring around Rome.
You do realize that the traffic will be moving at the speed of walking then, or ?
You don't understand. They see this as a feature.
They will not be happy until everyone (except them) is living in Kowloon and eating bugs. Individual mobility and autonomy is a problem to be solved, not a solution to be sought.
How exactly is individual mobility and autonomy enhanced on a weekday at 5pm on the 405 stuck in traffic for an hour or more? And then you have to park at either end of that commute.
You can get from The Battery to Inwood Hill Park (opposite ends of Manhattan) in an hour or less without a car. After you exit the terminal, no extra parking step needed. No car insurance. No gas. Just individuals mobilizing to their destinations with greater autonomy.
A car definitely takes longer and leaves you fixated on driving rather than reading a book, browsing the news, or streaming a TV show, especially at 5pm.
I use public transport almost exclusively but I can see why it wouldn't work for everyone. Juggling two kids and bags of shopping while hopping on and off packed busses/trains sounds like a nightmare. Not to mention delayed services. Hours of my week are wasted and London's is considered one of the best public transport systems in the world...
Car centric design is why you have to juggle the kids in the first place. I went to school on foot since my first day in school, later by bicycle — and this was (and still is) normal where I grew up.
Most kids in european cities go to school themselves from a certain age onwards, in my case that was after kindergarden, elswhere it might be a few years later, depending on the regional culture.
Current public transport is like transport belts in Factorio. Fixed routes.
This works ok in a simple old-fashioned economy (a few large employers, large hubs where everyone goes to the same place).
What we need is a more flexible, on-demand model of public transport. Something like logistic bots in Factorio. Lots of small people-movers which go direct from point to point, on demand. Buses and trains were already obsoleted by cars and can only survive by targeted anti-car policy.
Some kind of taxi (which would need to be automated to be at all economical for regular use) or electric scooter hire is the next stage of evolution.
Logistic bots do require a lot of energy, but that's the nature of progress.
> Buses and trains were already obsoleted by cars and can only survive by targeted anti-car policy
Where? In France most people would take the train to travel a long distance (2h+ drive). I can be in Paris in 2 hours at 320km/h using the high-speed train whereas it would take me 6 hours at 130km/h using a car. Not counting the nightmare of going through the Parisian périphérique.
To be completely fair, Paris is an exception. France is an unbelievably centralised country, and everything goes to Paris. Bordeaux is similarly well served, but past that... Good luck in the southeast and its broken 80km/h lines, good luck in the southwest and its complete lack of lines... Hell, even Lyon takes a very long time to get to if you're not coming from Paris.
Closer to 4h30 by car, although the train does drop you off directly in Part-Dieu. Still, this doesn't take into account the need to go to the train station.
>Nantes: 4h50 by train, 6h45 by car
From personal experience, I can promise you that 6h45 is horribly optimistic.
But really, that only furthers the point: Lyon is the second biggest city in France, and it's more or less equivalent to take the car to get there, with farther places slowly winning out on time. That's _dreadful_.
Berlin, and Germany in general seems to be a better role model then. Germany is the opposite of France, very decentralised with multiple population/economical centers dotting the country: Hamburg, Berlin, Frankfurt, München, Stuttgart, Köln/Dortmund/Düsseldorf/Essen corridor.
> Current public transport is like transport belts in Factorio. Fixed routes. This works ok in a simple old-fashioned economy (a few large employers, large hubs where everyone goes to the same place).
It works just fine in new-fashioned economies as well because good public transit is made of networks and criss cross.
As an example, even in the middle of the night, the majority of Toronto's population has public transit available just a short walk away:
> Then in 2005, several routes were added so that east–west service would run on every grid street instead of alternate ones in much of the city. This has brought the Blue Night Network up to a total of 24 routes, serving 97% of the city's population within a 15-minute walk.
Stumble out of a bar at 3 AM? You can take the bus/streetcar home.
During the day the network is much more extensive, with more routes operating, so most people probably have to walk <10 minutes to get to a route (some routes though having more frequent service than others).
Transfers add latency and uncertainty. It's just about tolerable on a metro system with trains every 5 minutes, but that's super expensive and only the largest, well-run cities have it.
The largest cities have the greatest need. Small towns that aren't high population centers aren't so strangled by traffic hopefully. And also hopefully are made to be more walkable and not spread out so unnecessarily.
Rural areas are rural. One car or truck every mile or so isn't the problem we're talking about here. It's the hundred cars per mile chokepoints that are the issue.
> What we need is a more flexible, on-demand model of public transport. Something like logistic bots in Factorio. Lots of small people-movers which go direct from point to point, on demand. Buses and trains were already obsoleted by cars and can only survive by targeted anti-car policy.
Why?? Cities like Berlin work perfectly fine with a massive public transport network, you get to almost anywhere in 30-45 minutes (except for the most far away places).
I don't need point-to-point on demand, I just need to be able to get into public transport in 5-10 minutes and get through the network to destination B in 30-45 minutes. That moves a lot of people efficiently, there's no need for this hyper-individualisation of transportation...
When it comes to public transit, "flexible" is usually a negative attribute. If you rely on public transit, you want reasonable guarantees that the service will be there in the future. If it's too flexible, someone will eventually ruin the service for you by changing it too much.
Public transit is also called mass transit, which gives a better idea of what the service is about. Traffic is a necessary evil. It's not valuable in itself but only as a means to an end. Mass transit works best in urban areas, where it maximizes throughput while minimizing the space wasted for traffic. The resulting urban topology is different from that arising from individual modes of transport. Instead of a grid of streets, you get a network of walkable neighborhoods. And much less traffic per person.
Not a viable solution in America. Americans are allergic to public transportation and density. Even in places where public transportation is relatively good by American standards it's seen as a money pit.
IIRC very few public transport systems make a profit on running costs - never mind capital costs. The rest operate as money pits.
Useful systems have huge positive externalities that can't be captured by charging passengers at a rate that reflects those benefits (I'm not sure if this is due to a free-loader effect, irrationality on the passengers part or something else).
I think the only systems that have profited on scale commensurate with their benefits are those that were a real estate play, eg.: the Metropolitan Line[1] in London, Los Angeles trams, Hong Kong subways. I might be wrong about these examples, either due to misremembering or to falling to a just-so story.
Edit 2: The point, I think, is that if you rely on direct profits for the transport system you give up on huge benefits and accept huge costs across society.
> Now it’s packed with cars that use it as a shortcut from congested Mission Street to heavily traveled Market Street. Residents must struggle to get to their homes, and accidents are a daily occurrence.
Of course people who used to have preferential access to a convenient shortcut would prefer that things remain that way. But why should they be the only one that get to use that road? Why are the needs of other people who want to get from A to B any less important. Odd view this is.
Because they live here, and because particle pollution and sound pollution actively kills people. Additionally, these roads were built assuming low amounts of traffic.
To continue on with your logic: why are pedestrians complaining that I'm driving on their sidewalk? What are their needs to have a safe place to walk at more important than my needs to get to my office really fast?
You can hardly call it a "shortcut" when it is the only way to reach your destination.
All roads are designed for a specific purpose, and roads designed for destination traffic simply aren't capable of handling large volumes of through traffic. Other people are supposed to go the long way because that's literally what the roads are designed for.
Tech being forced to interact with government to generally improve society seems to be such a foreign concept with the tech world that now I'm not sure how the anti-gov propaganda can be unwound.
My current burb has been fighting "cut through" traffic since before phone navigation was a thing. They do it with "traffic calming" (filling streets with obstacles to through-traffic) and designs that give people no alternative to getting on the main stroad. The result has been sheer hell. The frustration level is so high that whenever the main stroad is not gridlocked it is like Fury Road: harrowing for drivers and 100% lethal to peds and cyclists. But I guess that one guy on the side street who was shaking his fist at people speeding past his house is happy now.
I don't begrudge people who want their residential neighborhood streets to be residential. The issue is usually the development choices. The main cut through like you're describing near me recently had large bumps added to prevent speeding. No problem in my book except they should probably remove the traffic lights at either end of that street now. The lights either encouraged it to be a cut through or were added because it was, but in this case, there are not enough residents on that strip versus others to justify one, especially now.
> But I guess that one guy on the side street who was shaking his fist at people speeding past his house is happy now.
That’s me. My UK cul-de-sac street is a 20mph limit. It’s a thin street with cars parked on both sides with just enough room to drive down the middle. There are kids and pets.
People fly down the street at terrifying speeds, putting on seatbelts, fiddling with their phones and all sorts between two lines of parked cars.
My neighbour on one side uses my dipped kerb to mount the pavement at 15mph and drive along the pavement to park outside his own house. My other neighbour has a hedge, such that my son could be running up the drive on to the pavement, where we have taught him it is safe, and my other neighbour could be driving along the pavement and hit him without ever seeing him.
The street I live on just did this. I live on a residential street one block off of a main road, they added "Road Closed to Thru Traffic" construction barriers to both entrances of my street. They've been there for 2 weeks with no sign of construction.
* they can only load-balance in their own datasets, which means that users of different apps can over-compete for the same small amount of road capacity
* load-balancing is often implemented in a naïve way that fails to account for how particular roads aren't meant to take increases in traffic
This reads like whinery by someone who doesn’t know how navigation apps work. All the big apps have a way to designate streets and places as something they should never route through, either because it would make the app route people over parking lots, or because it is truly a disturbance to the locals.
In the case this writer is complaining about it’s probably a one-off event so this wasn’t set up. Big whoop.
The sinalling and designation must adapt to technological realities.
Streets not meant to be used as transit roads should be formally designated as local traffic only. Idealy this should be the default, prohibiting transit passage everywhere except on specially indicated roads.
This article is not well sourced and is inconsistent with lived experience.
Ten years ago navigation apps would almost always recommend freeways even when traffic was stalled and there was a faster route. Now I regularly get guided to exits to take routes through residential neighborhoods during rush hour.
Er, that's actually what the article says. Ten years ago routing was done with fixed databases, whereby a "freeway" (motorway) was always deemed faster. Now it's done with live data, so if the freeway is recorded as slow by cars currently there, apps send you through residential routes. This creates problems for the residential areas, effectively scaling up "rat racing" problems to unsustainable levels.
The author studies this problem for a living, so I'd be wary of calling them up on receipts.
Definitely agree that to whatever degree (if at all) vendors are required to work with city planners and transportation networks, it’s not enough. Transportation and logistics are too complex and delicate for the disruption approach of the tech sector and it can severely degrade or even endanger people’s lives.
I live on a small side street of an exit just before a horribly designed junction of two highways. A couple years after we moved in, we noticed that suddenly rush hour traffic was overflowing through the exit. Our street isn’t even remotely designed for that kind of traffic. Now we had an endless train of honking, angry drivers that effectively blocked us in our drive way. If someone on our street has an emergency, they’re fucked.
For whatever reason, it’s gotten better the past couple years and only pops up when there’s an accident on the interstate. Hoping and praying is a poor mitigation strategy, though.
I personally get around most of the time without GPS. I'll pull up a map and figure out directions, mapquest style. I think it keeps your brain plastic to navigate the world around you from memory, and I thing using a GPS is unhealthy for all but long distance trips on unfamiliar roads.
For those scenarios, I use OSM. I have been a contributor, I love it, it is more accurate than google and I couldn't give a damn about traffic information. I'm fine with the dynamic nature of finding out what traffic is like when I'm on the road, life is boring when experiences are entirely predictable.
"But now online navigation apps are in charge, and they’re causing more problems than they solve."
We shouldn't lose perspective. That statement is not even remotely true. Compared to the time and the gas savings, the problems described in the article are relatively minor.
53 comments
[ 3.0 ms ] story [ 105 ms ] threadIt’s apparently used to apply an additional routing penalty for “difficult intersections”, assigned by map editors, which provides a robust way to avoid these situations, including difficult left turns: https://www.reddit.com/r/waze/comments/ffypgb/avoid_difficul...
Cars will always be a perpetually morphing problem whose repetitively common theme is "too few people moved per unit area".
De-prioritize cars. Re-prioritize walkable, bikable, and train-accessible cities.
You do realize that the traffic will be moving at the speed of walking then, or ?
And mixing people with bikes and cars is a bad idea.
walkable paths do not usually overlap with cars/bicycle lane/bus lanes.
but they are very effective to reduce the use of vehicles for routine activities (grocery, schools, catching public transport, meeting friends, going out for a drink, etc.)
Also traffic speed is actually not that relevant, in Rome, where I live, most of the time cars move at an average speed lower than 10 km/h, the real issue is avoiding congestions, not making cars move faster.
15-20 km/h is more than enough to do what you have to do in a reasonable time.
where I live in Rome to reach the station I can either take the subway (5 stops, 11 minutes), the bus (~20-25 minutes), a cab (~15-20 minutes) or my car ~30-35 minutes and forget about parking it.
fast tracks for cars can be moved outside of the city, like the great ring around Rome.
https://en.wikipedia.org/wiki/Grande_Raccordo_Anulare
You don't understand. They see this as a feature.
They will not be happy until everyone (except them) is living in Kowloon and eating bugs. Individual mobility and autonomy is a problem to be solved, not a solution to be sought.
You can get from The Battery to Inwood Hill Park (opposite ends of Manhattan) in an hour or less without a car. After you exit the terminal, no extra parking step needed. No car insurance. No gas. Just individuals mobilizing to their destinations with greater autonomy.
A car definitely takes longer and leaves you fixated on driving rather than reading a book, browsing the news, or streaming a TV show, especially at 5pm.
Most kids in european cities go to school themselves from a certain age onwards, in my case that was after kindergarden, elswhere it might be a few years later, depending on the regional culture.
Current public transport is like transport belts in Factorio. Fixed routes. This works ok in a simple old-fashioned economy (a few large employers, large hubs where everyone goes to the same place).
What we need is a more flexible, on-demand model of public transport. Something like logistic bots in Factorio. Lots of small people-movers which go direct from point to point, on demand. Buses and trains were already obsoleted by cars and can only survive by targeted anti-car policy.
Some kind of taxi (which would need to be automated to be at all economical for regular use) or electric scooter hire is the next stage of evolution.
Logistic bots do require a lot of energy, but that's the nature of progress.
Where? In France most people would take the train to travel a long distance (2h+ drive). I can be in Paris in 2 hours at 320km/h using the high-speed train whereas it would take me 6 hours at 130km/h using a car. Not counting the nightmare of going through the Parisian périphérique.
- Toulouse: 4h by train, 5h30 by car
- Nice: 5h30 by train, 5h by car
- Strasbourg: 4h15 by train, 5h15 by car
- Le Havre: 5h05 by train, 7h by car
- Nantes: 4h50 by train, 6h45 by car
Closer to 4h30 by car, although the train does drop you off directly in Part-Dieu. Still, this doesn't take into account the need to go to the train station.
>Nantes: 4h50 by train, 6h45 by car
From personal experience, I can promise you that 6h45 is horribly optimistic.
But really, that only furthers the point: Lyon is the second biggest city in France, and it's more or less equivalent to take the car to get there, with farther places slowly winning out on time. That's _dreadful_.
Planes are best for longer distances, and some sort of personal transportation is best for short distances.
It works just fine in new-fashioned economies as well because good public transit is made of networks and criss cross.
As an example, even in the middle of the night, the majority of Toronto's population has public transit available just a short walk away:
> Then in 2005, several routes were added so that east–west service would run on every grid street instead of alternate ones in much of the city. This has brought the Blue Night Network up to a total of 24 routes, serving 97% of the city's population within a 15-minute walk.
* https://en.wikipedia.org/wiki/Blue_Night_Network
Stumble out of a bar at 3 AM? You can take the bus/streetcar home.
During the day the network is much more extensive, with more routes operating, so most people probably have to walk <10 minutes to get to a route (some routes though having more frequent service than others).
Rural areas are rural. One car or truck every mile or so isn't the problem we're talking about here. It's the hundred cars per mile chokepoints that are the issue.
Why?? Cities like Berlin work perfectly fine with a massive public transport network, you get to almost anywhere in 30-45 minutes (except for the most far away places).
I don't need point-to-point on demand, I just need to be able to get into public transport in 5-10 minutes and get through the network to destination B in 30-45 minutes. That moves a lot of people efficiently, there's no need for this hyper-individualisation of transportation...
Public transit is also called mass transit, which gives a better idea of what the service is about. Traffic is a necessary evil. It's not valuable in itself but only as a means to an end. Mass transit works best in urban areas, where it maximizes throughput while minimizing the space wasted for traffic. The resulting urban topology is different from that arising from individual modes of transport. Instead of a grid of streets, you get a network of walkable neighborhoods. And much less traffic per person.
Non-car centric development.
The Oh the Urbanity channel has a video on the (misguided) idea that "urban living" = Manhattan / Hong Kong apartment blocks:
* https://www.youtube.com/watch?v=BCmz-fgp24E
Want a front yard, back yard, and garage (attached to a laneway)? Plenty of that was built pre-WW2:
* https://en.wikipedia.org/wiki/Streetcar_suburb
* https://www.youtube.com/watch?v=MWsGBRdK2N0&t=1m8s
Examples (Streetview):
* https://www.google.com/maps/place/125+Hampton+Ave,+Toronto,+...
* https://www.google.com/maps/place/50+Geoffrey+St,+Toronto,+O...
* https://www.google.com/maps/place/70+Jackman+Ave+Toronto,+ON
See also:
* https://www.youtube.com/watch?v=cO6txCZpbsQ&t=9m28s
Fifteen minutes pedalling in one direction is downtown, fifteen minutes in the other is farm land.
Useful systems have huge positive externalities that can't be captured by charging passengers at a rate that reflects those benefits (I'm not sure if this is due to a free-loader effect, irrationality on the passengers part or something else).
I think the only systems that have profited on scale commensurate with their benefits are those that were a real estate play, eg.: the Metropolitan Line[1] in London, Los Angeles trams, Hong Kong subways. I might be wrong about these examples, either due to misremembering or to falling to a just-so story.
[1] https://en.wikipedia.org/wiki/Metro-land
Edit: formatting.
Edit 2: The point, I think, is that if you rely on direct profits for the transport system you give up on huge benefits and accept huge costs across society.
Of course people who used to have preferential access to a convenient shortcut would prefer that things remain that way. But why should they be the only one that get to use that road? Why are the needs of other people who want to get from A to B any less important. Odd view this is.
To continue on with your logic: why are pedestrians complaining that I'm driving on their sidewalk? What are their needs to have a safe place to walk at more important than my needs to get to my office really fast?
All roads are designed for a specific purpose, and roads designed for destination traffic simply aren't capable of handling large volumes of through traffic. Other people are supposed to go the long way because that's literally what the roads are designed for.
https://www.wired.com/story/99-phones-fake-google-maps-traff...
Open up an API for governments to drop in temporary blockers in the map (with appropriate lead-up times too).
Competitors could come to some agreement on using each other's apis for emergency or traffic jam alerts.
That’s me. My UK cul-de-sac street is a 20mph limit. It’s a thin street with cars parked on both sides with just enough room to drive down the middle. There are kids and pets.
People fly down the street at terrifying speeds, putting on seatbelts, fiddling with their phones and all sorts between two lines of parked cars.
My neighbour on one side uses my dipped kerb to mount the pavement at 15mph and drive along the pavement to park outside his own house. My other neighbour has a hedge, such that my son could be running up the drive on to the pavement, where we have taught him it is safe, and my other neighbour could be driving along the pavement and hit him without ever seeing him.
People just don’t think.
Apps do load-balance, but the main issues are:
In the case this writer is complaining about it’s probably a one-off event so this wasn’t set up. Big whoop.
Streets not meant to be used as transit roads should be formally designated as local traffic only. Idealy this should be the default, prohibiting transit passage everywhere except on specially indicated roads.
Ten years ago navigation apps would almost always recommend freeways even when traffic was stalled and there was a faster route. Now I regularly get guided to exits to take routes through residential neighborhoods during rush hour.
The author studies this problem for a living, so I'd be wary of calling them up on receipts.
I live on a small side street of an exit just before a horribly designed junction of two highways. A couple years after we moved in, we noticed that suddenly rush hour traffic was overflowing through the exit. Our street isn’t even remotely designed for that kind of traffic. Now we had an endless train of honking, angry drivers that effectively blocked us in our drive way. If someone on our street has an emergency, they’re fucked.
For whatever reason, it’s gotten better the past couple years and only pops up when there’s an accident on the interstate. Hoping and praying is a poor mitigation strategy, though.
For those scenarios, I use OSM. I have been a contributor, I love it, it is more accurate than google and I couldn't give a damn about traffic information. I'm fine with the dynamic nature of finding out what traffic is like when I'm on the road, life is boring when experiences are entirely predictable.
We shouldn't lose perspective. That statement is not even remotely true. Compared to the time and the gas savings, the problems described in the article are relatively minor.