Could self-driving buses bring vehicle autonomy home?
spectrum.ieee.org
spectrum.ieee.org
"Bus driver" is not exactly a high-paying job, yet one that is important for social cohesion: passenger safety and informal surveillance of roads and neighborhoods sure are crucial!
When it comes to anything 'self-driving', the main goal should be to reduce road traffic and neighborhood pollution-by-parking-spaces. This does nothing for either purpose.
That's all without value?
Do you think 'surveillance' is always bad? Like: nobody should mention the tree blocking the road, because, well... the implications?
No. I would love for there to be many times more traffic cameras and license plate readers. Video-footage from self-driving vehicles could go a long way to getting dangerous drivers off of the road.
Actual case study, maybe slightly summarized an exaggerated:
-Prosecution: here is a video of Defendant repeatedly running over several pedestrians
-Defense: it was late at night, visibility was really low, and, besides, my client was very drunk. And they really, really need their car and license for their job! They would be ruined otherwise!
-Judge: well, good points, are we missing something here?
-Prosecution: They. Ran. Over. Several. Pedestrians. Forwards, then backwards!
-Judge: yeah, OK, sure, they should cover the immediate medical costs for those unfortunate bystanders, but other than that I order a safe-driving course, 12 months of alcohol interlock and 36 months of suspended license revocation that nobody will care about, ever. Next case!
[1] https://www.dnainfo.com/chicago/20170126/old-town/ryne-san-h...
[2] https://www.nbcnewyork.com/news/local/crime-and-courts/nyc-c...
In my city, 70% of the operating budget is labor.
> passenger safety
An excellent way to ensure passenger safety is not to share an enclosed space with someone dangerous.
> When it comes to anything 'self-driving', the main goals should be to reduce road traffic and neighborhood pollution-by-parking-spaces.
When it comes to transportation systems, the goals should be: safety, utility, reliability, convenience, speed, flexibility, efficiency, environmental impact (pollution, noise, etc.).
Link? Because, well, your city is certainly special! Unless, like, you group things like 'pension liabilities', and 'insurance involving personnel' under 'labor', which is... creative but not accurate.
https://www.civicfed.org/civic-federation/blog/cta%E2%80%99s...
Paying the bus drivers does not account for 70% of the budget, I assure you.
I've dealt with self-reporting systems in this area, and I do not have the words to tell you how much less I trust self-reporting than an 8 hour test-drive by someone that knows the bus.
This skips over all the local police's missing person reports that show up for every driver that can help. Completely unrelated to the job of driving - but when it's my kid that doesn't come home, arguably more important to me.
Drivers are sticking around, and I think a lot of people are confused about how much they do to keep things running
US citizen? Can I ask, why is it that US citizens are so worried about dangerous people in public transport? Is it a rational concern or is it more of a -no offecne meant- preconceived idea? Is it something that's evenly geographically distributed or something you find more in certain places than others?
I'm asking honestly and curiously and hoping more to clear up my own preconceptions than to enforce them. I do find it strange; because I use public transport basically all the time (I don't own a car). I live in the UK and some years I commuted daily by train to work and back. Otherwise I take the bus when I have far to go. I have consistently traveled between the UK and Greece by train and ferry boat for some 15 years now. I've never felt that I was in any kind of danger, although to be fair some times the London trains can be a bit of a pain in the ass on a Saturday night (too many drunken brits, but they're mostly harmless, and really just rowdy).
I've been on some pretty wild lines too. For a few years me and a friend would take the night train from Paris to Venice, and back on the return. That's a sleeper train with six-person cabins. We rode it on through the years of increasing immigration waves from Africa through Italy to the North of Europe, so I've traveled with people with ... varied life histories for sure. Still, all this time, me and my friend's most memorable experience remains the time a Chinese gentleman was asked to open his luggage by border guards in Switzerland, at which point everyone in the compartment (and possibly the ones next door) was very clearly able to smell the large fresh fish he was carrying in his bag. They took him off the train because he didn't have papers. For himself, not the fish. I hope he reached his destination. I bet everyone would have been eager to fast-track him along.
I mean, it's not the boring safety of sitting in your car but it's not dangerous. Only interesting. And the night train, when it careens through the Alps you're like a baby in the cradle, swaying gently and lulled by the sound of the rails. I miss that train. I think the company went bankrupt during the pandemic and they don't run anymore. That has complicated me and my friend's travel quite a bit in the last few years.
So I have good experiences of public transit. What's the problem with US public transport? Why are people so worried about it?
Yes. I understand that there is some enforcement on trains in Europe that you be a paying passenger. In metro trains in the US this is less common or non-existent. Because of the US inability to deal with drug, mental-health, and homelessness issues, public transit often ends up being shelter for a lot of people who have nothing to lose. The risk of being murdered is low, but the risk of an unnecessarily unpleasant experience is high: begging, harassment, groping, smoke-filled train cars, food scraps, having someone spit in your face, urine, vomit, and feces are stories every commuter has. Inter-city trains are generally fine.
There is enforcement in Europe, more or less depending on where you are. In the UK it's not unknown or even not uncommon for homeless folks to board busses or trains and maybe taking a nap (or a good night's sleep) instead of sitting outside and freezing. In the UK and in Greece beggars are also a common occurrence.
>> The risk of being murdered is low, but the risk of an unnecessarily unpleasant experience is high: begging, harassment, groping, smoke-filled train cars, food scraps, having someone spit in your face, urine, vomit, and feces are stories every commuter has. Inter-city trains are generally fine.
That certainly sounds extremely unpleasant. What I'm trying to understand is how high is the risk, really, and whether it is higher in some cities than others. I mean because I still can't imagine public transport where people will spit in your face in the bus or in the train. Although I've seen people getting sick (and then jumping off at the next stop, bastards) on London busses. In London, being puked on of a night out is a real risk but then it's nothing to do with public transport.
Anyway, how high is the risk of such unpleasant experiences in the U? Why is it that way?
Urbanists: we should make mass transit more affordable and convenient, and people will use it more.
Industry: Ok we have tech that can reduce the cost of operating buses by 50% and let them operate round the clock.
Urbanists: No no not like that. I didn't hear the words "trains" or "light rail" anywhere.
I personally think we should celebrate any improvements in the cost and availability of public transit. And that includes buses.
"Train brain"
The future could be or a mass genocide to lower quickly earth population or an "urbanized" world of single family homes and sheds. That's why all who knows push "urban air mobility" which can't be in dense areas and can't be "collective" as well, because it's a way to reduce the need of infrastructure again and again.
We need a cluster world, not a mainframe one.
Putting the soap box aside for just a moment however, I could see this being feasible for static bus routes, assuming your city maintains the roads and signals along that route so the ML algorithms piloting the multi-ton death machine on open roads isn’t distracted by a new pot hole or missing road stripe. Since no city in America actually does this though, these things seem about as reliable as the “self-driving” Teslas under Vegas. Then you throw in the need to retrain the algorithm every time the route changes, and the millions in expenses that’ll cost versus the few thousand it takes to retrain a human driver on a new route…
Yeah, nevermind, breaking out the soap box again. Stop trying to short-change infrastructure investments and just invest in the infrastructure already. It won’t make some SV VC fund richer, sure, but it’ll at least go towards something useful and necessary for the city’s occupants.
So light rail or trolleys is the same but worse and more expensive (plus they also still have drivers).
For this to get used, it needs to:
- not be beholden to traffic or congestion. Like a subway or the L? Or dedicated lanes for busses could work and may be drastically cheaper
- needs to be able to service lots of destinations and changing needs. Subways do this by building new lines and stations (again, $$$)
-needs to be relatively clean and safe. This is an actual concern people have that has to be addressed.
This is discussed a bit in urban planning circles, one of the reasons trams or light rail is preferred is political, it's much harder for some austerity government to come into power in a city and remove tramlines than remove bus and bike lanes.
The brutal core of this argument remains those who demand we take urban planning and development seriously to benefit the masses, versus those who seek fortune and riches bootlicking those who would benefit from chaos and disorganization. I hate to be that blunt about it, but I think the HN crowd can swallow that bitter medicine better than most.
If you’re all-in on a car-focused city, then you’re going to be disappointed. If you have gripes with your current mass transit (like Boston’s infamous Green Line), then it’s better to address those gripes directly than throw the baby out with the bathwater in favor of something newer, shinier, and worse.
This bus-hate is irrational and self-defeating. A successful mass transit system has multiple modes of transport. Making one of them cheaper than it is now can only be a good thing.
Just as driverless trains are a thing, it seems obvious that buses should also mature to be driverless?
It’s the same blinders I see time and time again: you cannot solve an inherently political problem through technology alone.
One of the innovations of modern self driving systems is not requiring any accomodation from the infrastructure. It's the same idea as "smart endpoints and dumb pipes".
You mean roads? Yeah I guess if a city isn't up to maintaining its roads, it's pretty f*cked anyway.
> you cannot solve an inherently political problem through technology alone
Often technology moves a problem away from the realm of politics. Politics is a game of allocating limited resources. If technology brings abundance, the question of allocation is way less important.
Plus, even once you've designed a system that can do all that safely, you'll still need staff on board if you want to collect fare revenue.
Fare evasion is an interesting point. Of course you can't completely avoid it, as it happens on rail too. But you could install a turnstile on the bus.
Google has spent less so far on building fully self-driving cars than the state of California has spent on HSR and they have only built 22 miles out of 171, never mind that the selected route is clearly the outcome of some stupid politicking.
Light rail also kinda sucks if I am being honest, at least in every city I have lived in.
Jane Jacobs argued that trains are suitable for well established routes of travel. First the bus, then the train.
Totally, and this is true on a microscale too. 1,000s of train commuters are regularly delayed by a "sick passenger" on a rail line. With buses, following busses can go around any stopped bus.
The maneuvers I see them do regularly are unlikely to be done by any current self-driving technology on a congested city street. With dedicated lanes or with no congestion, I can imagine them doing fine.
When it comes to negotiating with humans, I've already seen Waymos flash their lights at bicyclists and pedestrians to cross.
For other maneuvers, I've been in one that backed out of the crosswalk when it realized it crossed too far.
It's difficult to define what they "can't" do, even in the medium term.
A bus on the usual fixed routes? Not really.
The healthier option for the masses is expanding mass transit networks, not luxe taxis. Light rail is infinitely more efficient and safer than individual vehicles, be they owner-operated or self-driving. It moves more people, further, faster, and cheaper than any potential self-driving automobile. When you include the whole of networks - ferries, bus routes, commuter rail - they’re the single most effective solution to everything self-driving car proponents claim their products can do, if only they had another round of funding.
I’m not saying we should stop investing in the technology, mind. I just think it’s a better use of our limited resources to invest in improving, expanding, and modernizing mass transit than to build armies of privately-owned taxis, self-driving or otherwise.
Not by themselves. But combine them with a congestion tax, unless the taxi is full, and you've got an infinitely adaptable, cheap, and scalable bus network.
I like public transit, and I Know he public can get a lot more for its money if we leave behind our old thinking.
NYC only has 6k buses for 2 million weekday passengers. Make them more convenient by say having 20k of them and you get more traffic throughput.
Classic school bus costs ~10k/year in depreciation and get 10 miles to the gallon making the driver at an average 18$/hour the largest cost even outside of cities.
The majority of bus drivers are semi-retired folk looking for something to keep them active and earn a bit on the side.
I'm speaking from experience.
I would guess a driver costs $60-90k year including taxes and benefits.
Drivers aren't a majority of the cost of operating a bus, but I'd bet it's at least 20%.
Fuel is also cheap as they are only averaging 7MPH even at 2-4MPG it’s just not touching 27+$/hour for the drivers.
Per utilized seat, handmade diesel busses are way more expensive to buy and operate than commodity vehicles like Jaguar iPace or the Zeekr Mix.
Further you need HR and managers to handle having a large driver workforce. So while drivers aren’t responsible for all of that 70%, they are a high percentage of it.
By moving from specialty vehicles like large diesel busses to vehicles based on commodity/consumer minivans, it will be possible to multiple external services. This gives the taxpayer the benefit of scale and competition.
There's no need for a monopoly in city transit. In a future state, we could have multiple transit providers, similar to how airlines operate today.
Adding buses to already congested streets might just make the situation worse.
Israel ran into this problem, and they are desperately trying to recover by building light rail in big cities.
That seems counterintuitive, but bus utilization and congestion occur at the same time. So one bus is replacing 20-60 cars in busy periods. At 2AM an empty bus may be taking up a lot of road, but the streets are practically empty.
It might depend on many things, like whether the bus actually goes anywhere useful, or the would-be passengers would still have to travel far from their stop (and would prefer to drive a car, as bus doesn't solve their commute problem). Would be especially problematic in areas with single-family houses spaced far apart. Very infrequent bus service may also be a factor that would sway would-be passengers to drive cars instead, but it would be hard to justify more buses on the line.
Well, I'm not a city planner, just trying to imagine the potential problems. Bet people who deal with this would know better.
Each individual improvement was usually small and insignificant.
The issues with busses are lack of coverage by the network and the shortage of drivers because it's an awful job that pays shit wages, both of which are down in large part due to lack of funding and the political will to allocate it. We don't need a computer to solve these issues.
Unless this is an obscure reference about drug use in your neighborhood or something that's completely going over my head.
My original comment is not about “the poors” and I am not a tech worker nor do I make “the big bucks”. Lower the pitchforks, please.
Or to be able to wait exactly at the point you're departing from instead of an arbitrary point outside.
https://www.powerprogress.com/news/mercedesbenz-delivers-cit...
The driver's area is entirely separated by an internal door, with an optional hole in the glass for operators which take cash. These separating doors became most widely fitted after the COVID-19 outbreak (for the drivers' safety), but they were already in use in many European cities years before that.
Bus fare is already pretty expensive when you consider the quality of service, speed, and convenience outside major metro areas.
> due to lack of funding
If there's a lack of funding, and there's tech that will make buses cheaper, it's a no-brainer to adopt it.
> the political will to allocate it
Willpower and political capital are limited resources. Why squander them when there are alternatives?
This logic only holds if you assume public services must make money and this is not a universal rule. Tons of public services hemorrhage money. It's called "spending money to get things" and we as a society have been doing it for basically as long as money has existed.
> If there's a lack of funding, and there's a tech that will make buses cheaper, it's a no-brainer to adopt it.
A regular car retrofitted to operate "autonomously" costs upwards of 100k by a quick google, and we're not talking a Mercedes here, we're talking a bog standard crossover. Busses cost near 6 times what that "autonomous" car costs for a boring one that a person drives, how much do you think a robo one will cost?
Also worth noting, the prototype pictured in the article looks to move WAY fewer people than a standard city bus.
I don't assume that. But I'm not the electorate.
> Tons of public services hemorrhage money
And that makes them easy targets for cuts. I personally prefer having public services stick around to serve people. That means stiffening them against political attacks as much as possible, and taking easy wins if they're available.
> how much do you think a robo one will cost?
So now we're arguing the actual cost-benefit of the technology. That's a much better way to approach it than reflexively dismissing it.
Awful jobs that pay bad wages seem exactly what we want to solve with machines.