Future Cars: Electric and Autonomous Vehicles

[fehler]

Also, the main expense of short-bus, local transit services is the driver. Take that out of the equation, and transit can reach out to many more areas.

[twincitizen]

Yeah it's weird when people go down that road of "driverless cars will make transit less relevant", that they don't assume transit vehicles will also be driverless. Take driver salaries out of the equation and transit becomes several magnitudes cheaper to operate.

[xandrex]

The real issue with getting driverless buses will almost certainly not be the technology, but instead the unions that represent the drivers.

[talindsay]

There are automated ("driverless") subway systems all over the world and a fair number of automated surface-running rail systems. Each of those systems had to "downsize" their driver pools when they automated, and those transit operators have decades of experience with this model. I think a transition for mass transit is likely much easier than for private drivers.

[FISHMANPET]

On the other hand, it's a fight to get OPTO in NYC or get conductors off of commuter railroads in the NE. So it's not like, totally outlandish that unions could pose a problem in this case.

WMATA is automated but they keep a driver in the train "just in case." WMATA sucks so they probably need it, but still.

[RailBaronYarr]

Had many a twitter conversations about driverless transit. One argument against is the role drivers play in making transit actually function: helping disabled get on/secured, answering questions, keeping unruly people in line, etc. There's merit to those examples. But level boarding, high quality stations with information, and a hopefully smoother ride thanks to computer driving (and bulbouts and such) make most of those things less necessary. Roaming (or even dedicated) transit police can keep unruly behavior down.

It's going to be a tough conversation to have. I like that government agencies typically pay a living wage (plus benefits, etc) to hourly jobs as an example for society. But reducing operating budgets also needs to be an essential function moving forward to make transit as viable as possible. If we could double systemwide frequencies tomorrow by cutting driver salaries, is that a net win for society?

[mattaudio]

Vancouver SkyTrain is automated. It allows them to run trains at high frequency 24/7. Transformative. And we've been doing it, albeit on fully separated fixed guideways, for decades.

Driverless cars? They still take up the same space as a car with a driver.

[Snelbian]

I don't see how sticking an AI in a car changes purchasing patterns. Automate the entire 2016 Ford lineup and you'll still have people who want a jacked up pickup to drive from a cul-de-sac to an office job downtown. It's just going to be a robotic jacked up pickup. Automation has NOTHING to do with fuel efficiency or carpooling or anything at all other than (hypothetically) accident rates. The reason people like me want reform in " land use, transportation and public policy" is because these are things that need lots of improvement. And there are ways to do it that affect many aspects of life. Automated cars are emphatically not one of those ways. Envisioning a robotic chauffeur is just sexier than tweaking zoning codes and lobbying for arcane sources of transportation funding.

Just bought a pickup (not jacked up) to drive from my suburban house to my office in Minneapolis, and I couldnt be happier. No amount of technology will get me to downsize. A pickup makes my life a lot easier. A small car would be a nightmare. You need to remember that everyone has different needs.

Your lifestyle and my opinions about it aside, I was responding to nbode's post about efficiency. And you demonstrate my point - vehicle size is a choice and one that I don't think will be predicated on automation. There are a lot of factors that go into how big of a car people buy, from the desire to haul lots of things, to the desire to parallel park easily, to the desire to demonstrate some sort of virility or ecoconsciousness or tribal brand adherence or size of genitalia or disregard for the safety of others and on and on. It's not "oh, I don't have to actually steer? Great, I'll take the tiny one over there!"

[MNdible]

Not that it really changes many of the points that have been raised here, but there's real evidence that driverless cars would increase capacity of our existing freeways in ways beyond reducing crashes. By maintaining appropriate following distances and anticipating traffic slowdowns, driverless cars should be able to avoid the waves of braking that we see on our roads today -- slow downs for no particularly good reason. Freeways should be able to handle more traffic more smoothly by eliminating the human tendency to tailgate and the subsequent need to slam on the brakes.

[EOst]

You probably need a relatively high percentage of driverless cars on any given highway to achieve that, though. Realistically, how long is it going to be before driverless cars have even 25% penetration?

[Snelbian]

Which also means that the automated driving needs to be capable of accounting for the overwhelming number of drivers early on who behave stupidly. Since I don't pay much attention to Google or Tesla's efforts here, I don't know if they're at that point (though I very much doubt it). Anyone know?

[xandrex]

I mean, Google's automated cars are out driving right now, even on some highways. I remember on some NPR program they were talking to some engineer who was able to use one for a while and he said his wife preferred when he switched the car to automated mode on the highway because it felt safer.

Whether or not we can achieve more efficiency at this point given low penetration is another question.

[xanadu]

With the pervasiveness of cellular modems, I think the inevitable step is self-driving cars talking to a city-wide traffic network, where decisions are not only made ad-hoc via a camera system and AI, but also where actions are dictated to the car's AI by a central dispatch that is aware of real-time traffic patterns, road hazards and the most efficient route. Imagine if your car knew instantaneously that the car in front of it stopped due to a pedestrian running out in front of traffic, as did all the cars behind yours; they could safely ease into a stop and resume immediately. Traffic would move at a level of efficiency, safety, and potentially at a speed incomprehensible given human reaction times. This would allow us to greatly reduce the number of lanes needed for traffic, as each vehicle would move at optimum efficiency. We're a ways away from this, but give it 50 years... or 100

[seanrichardryan]

[Silophant]

Hm... what's the latency on cell networks? Computers react much more quickly to sudden obstacles than people do, no doubt, but if you add a round-trip via cell modem, is that still true? Additionally, you need to keep a fair bit of intelligence in the vehicles, or else an outage at the control center brings the whole city grinding to a halt. You could set up failsafes such that a loss of connection brings everything to a halt, but keeping some autonomy is preferable.

I'm not disagreeing that that's the direction things are headed, just pointing out some reasons that it absolutely is decades away. Several decades, not just one or two.

[mister.shoes]

I would expect some sort of ad-hoc WiFi communication—a la Apple's Bonjour—to be more effective and fail-safe than something cellular based, but LTE would certainly win the distance competition.

[Anondson]

Self-driving cars are in crashes twice as often as human-driven cars. And not once has the self-driving car been at fault. The disparity is explained that self-driving cars are programmed to perfectly follow the law. Humans have come to expect a given amount of law breaking while driving, like running red lights and speeding, that when humans come up the "perfect driver" humans keep crashing into them.

http://www.bloomberg.com/news/articles/ ... a-key-flaw

[VacantLuxuries]

It'll be interesting to see what the solution for this will be. I'm in favor of phasing out human driving altogether, but I imagine a decision like that would be met with some passion.

[Archiapolis]

I wonder what would happen if the global community put half of the energy currently dedicated to driverless cars, hyper loops, tubes, teleportation etc into boring stuff like zoning, public transportation (and it's infrastructure), biking and walking.
I'm sure it's all available via google (irony?) and wikipedia, etc but I wonder what the raw dollars dedicated to these efforts would do for things that aren't idiotic and the net benefit to the world that would result. I'll see myself out...

[moda253]

Driverless cars isn't about cars making their own decisions on the road. The notion that driverless cars would be just as poor in terms of efficiency as human driven cars is sort of crazy to me. The transit system is a network and humans are very bad at managing that network. We get distracted, angry, bored, sleepy and all sorts of other things that effect our ability to manage the system. But cars that are capable of networking with one and other can handle managing the network MUCH better. When all 250 cars ahead of you in the 1/4 mile before you have to exit understand that your vehicle needs to be in line to exit the freeway then it becomes a math problem to get your car where it needs to be in the most efficient manner. A pretty simple math problem.

Of course not all cars on the road are going to be networked to start off with, but that can change in a big hurry. There can even be after market add ons for even "dumb cars" so that they can interact with the networked vehicles. There will always be obstacles for the networked vehicles to work around a human driven car would just be another obstacle on the road.

For me I envision this being most applicable in urban settings. I have absolutely no doubt that networked smart cars would definitely process how to move vehicles more efficiently than thousands of humans.