Could Google's self-driving Car kill transit?

[Derek]

I noticed the Prius didn't use a turn signal when pulling into Taco Bell. They should probably address that.




Still a really awesome project put on by Google. Very impressive.

[tyroneshoelaces]

Self driving cars approved by the California legislature.

http://www.cnbc.com/id/48861376

Since a huge percentage of cars in this country are in California, the rest of us usually follow their lead.

[2oh1]

That's sort of misleading.

If signed by Governor Jerry Brown, Padilla's bill would legally allow autonomous vehicles on the road and charge the state's Department of Motor Vehicles with determining the standards for self-driving cars, rules which current do not exist under the present vehicle code.

In other words, rules are on the way. The idea of driverless cars roaming without drivers isn't likely. And that's putting it mildly.

the bill requires the cars to have a flesh-and-blood human being behind the wheel if something goes wrong.

That last sentence is poorly written. The bill requires autonomous vehicles to have a flesh-and-blood human being behind the wheel, period. The reason for the requirement is to ensure safety in the event that anything goes wrong.

We already have too many cars on the road with only one occupant. If we ever see a time when driverless cars don't require anyone to be behind the wheel, we're going to see a massive increase in traffic congestion, and that will only encourage mass transit use rather than discourage it (getting back to the initial question posed in this thread).

[llamaorama]

Clearly congestion tolling is necessary or these will cause all kinds of problems.

The cars are all computerized and could be networked together, and you wouldn't need physical barriers or gates to change their paths, right? So a city where robot cars are taking over the streets simply could tame them with a traffic management program?

[Stryker]

This could result in the rebirth of suburbs. When you factor in these cars are electric, they're the ideal way of funnelling people in suburban rapid busing.


I think it's very logical to assume people will use a combo. Think of how subways and busses are used now.


Electric petty cabs for your neighbourhood trots at slow speeds, than express bus services that are able to run at full speed.

The benefits are undeniable.

When your realize how much resources are wasted on everyone owning heir own car.

The traffic congestion caused by people unable to drive efficiently.

THe number of accidents caused by driver error, which if you know anything far out weighs the number of automotive error. It's one of things that gets overlooked when they call it car accidents. Most are caused by drivers and not cars.

[colganc]

People won't, in general, own driverless cars. It won't make sense. Companies like Car2Go and ZipCar could have fleets of them. You sign up for a monthly fee based on your commuting pattern.

Want a guaranteed five minute pickup from when you request a ride during a 7am to 8am window? 300 per month.

Want a guaranteed five minute pickup from when you request a ride during a 6am to 7am window? 200 per month.

It wouldn't be cost effective for these companies to able to service all trips at the same time so their fleets will only be designed to handle maybe 50% at peak and shift people to the two adjoining windows. This could reduce the number of cars on the road for the number of people who have signed up by half during peak times.

These same households who signed up for a driverless taxi service would go from one car to none or two cars to one or none. You would see car ownership rates cut in half in a very short period of time.

Once driverless is accepted you could create driverless lanes on highways with increased max speed limits. Imagine cars going 20% faster. That would reduce the number of driverless taxis as well, because a smaller fleet size could serve the same peak numbers. It would also reduce the window where max congestion on freeways occur.

Buses won't survive. Maybe light rail.

You may even see some companies offer automated car pooling. Imagine once you sign up for a commute time slot and destination with a company. Due to economies of scale there may only be three or so companies offering the service in each major MSA. The companies will easily identify people in your area also commuting to the same general destination at the same time. It would be trivial for them to offer a major discount to the people who do driverless car pooling.

Imagine if 75% of car commuters sign up. Imagine you only need 10% of the max vehicles due to pooling, increased speeds, and time shifting (congestion charging). For every 100 cars normally commuting you could drop that to 33. Rush hour at many cities would totally disappear. I can see in Portland, people from north Vancouver that have 1.25 hour commutes now, could have 25 minute commutes. Pollution due to vehicles would drop dramatically as well. The companies could see actual financial benefit to using the most fuel efficient and cleanest burning cars. Fuel economy numbers could rise. Think 60mpg or better hybrids.

Totally transformative.

Pool three people into one car and drop them off after 25miles of driving. A hybrid driverless car would only use half a gallon of gas, $2. In 200,000 miles you could make 4000 of those trips and that's assuming you're not doing any smart routing. In 200,000 trips you could use 4,000 gallons of gas, not inflation adjusted terms that's around $16,000. Imagine the driverless car costing $40,000 in fleet purchasing. Let's guess at $10,000 in maintenance items over 200,000miles. We're at $66,000 for 4000 revenue generating trips with occupancy at 3 if carpooling works. In other words 12,000 revenue events. That's 6 bucks per trip nearly. Double that for your trip home. Profit and overhead can fit in that amount because there are a lot of smart optimizations that can be done beyond pooling. Caching cars at locations for the trips home from work. Picking up people on reverse commuting. Getting people that work night shifts. Delivering packages. Stuff like that.

For 12 bucks per day or roughly $240 per month you can commute to work without doing the driving. Try and find a car with that kind of total costs, it will be hard to beat.

I can't wait.

[davehogan]

Originally Posted by colganc

Buses won't survive. Maybe light rail.

I agree with you in large parts, but there are too many people who can't afford to pay $200-$300/mo, or wait four hours for a car to hit a cost effective price for them (which it may never be able to) so they can afford to travel.

I use Car2Go all the time. I don't own a car, but I pay what a lot of people spend for insurance on it in any given month. It's good, but you need disposable income to decide, "I'm going to get a burrito, I'll spend $4.10 to drive it home and eat it while it's still delicious and warm" vs. "I'll get the burrito, then wait 24 minutes for the bus, and eat half of it while I hope the bus doesn't show up early, then stand in the drizzle, then ride on the bus wanting to finish the burrito, then I can get home with a cold burrito, while kind of damp, it's now an hour after I wanted my burrito."

That situation came up for me tonight, the extra $1.60 to enjoy the heck out of that burrito made the investment worthwhile. Plus I was able to enjoy a beer, which TriMet generally frowns upon if you do it on their buses. Not to mention the whole eat a burrito is against the rules as well.

As drivers trust automated vehicles, I'd guess the self driving vehicle systems will shake out like this:

Rail-based (MAX/WES/Streetcar in Portland's case) systems: Automation makes them cheaper to operate, safer, and future investments make it a good trunk system available to the masses who live/work near stations. You still can't reach the entirety of a metro area (for example outside of Manhattan without other forms of transportation. See also: Staten Island.)

Bus-based systems: Low cost systems to get to/from rail stations. Automated driving systems will make them lower cost than we can imagine right now. If TriMet keeps charging the equivalent of $5/day as they can phase in self driving trains and buses they'll have plenty of money for expansion.

Bus and rail will probably stay with a unified fare structure as long as TriMet retains it's current structure. When drivers (and the long term costs associated with them) are no longer needed and all that's required are programmers and mechanics keeping a fleet going for 24 hour service isn't nearly as unreasonable, even if buses have to sub in for trains for a few hours every week or so for rail maintenance.

Private shuttle service, similar to how airport shuttles work would probably become much more popular. Since self driving vehicles will likely have existing data uplinks you can provide WiFi, television, etc to the passengers who want to subscribe to a service that's better than a bus but not as expensive as reserving a private vehicle. Subscriptions to this and the following would probably involve some kind of bulk discount for daily users.

Then you have the Self-Driving Car2Go model. A vehicle that will drive you from point A to point B for x amount per minute. You get your own private ride, your choice of music, AC/heat, space for your burrito (or whatever you love), etc. If you want to crank up Cheap Trick and roll down the windows go right ahead. This isn't a TriMet bus, but you're paying a premium. Plus your burrito is more deliciously warm when you get home.

Self-driving cars won't be unheard of for private ownership either though. They'll just be reserved for people with commutes that are too long to affordably use another option, and have the disposable income and parking available to purchase instead of rent by the minute/hour.

That also leads to another interesting point: Self driving cars will revolutionize parking. Why build an expensive garage at PDX or downtown when the car can drop you off and go find parking somewhere with a lot of available spots? Why not require in certain zones that your vehicle re-locates itself if a commercial vehicle needs to unload there?

Your car can come get you. Why not just have it get itself out of the way and find somewhere to wait?

Human driven cars will eventually be toys of the super rich as insurance and parking gets more expensive if self-driving cars are a success. That's probably not going to happen for a while, but as computer controls get safer with more miles traveled humans will inevitably have to pay for more of the risk pool for the insurers.

I can't believe it's 2014. Welcome to the future. It's pretty awesome here. I can't wait to see what 2030 is like. Sorry for the long post, it's just a topic I'm really interested to see how it works out.

[Owlhorn]

Why does everyone think suburban areas are the main beneficiary. Driverless cars would be a huge boon for urban areas. If you don't have to own them, then you lose the need for, a) garages, b) huge streets c) city parking lots d)Inner city freeways. If they are electric, the fuel station can eliminated as a space taker as well. More space for parks, bike lanes, rapid transit. More money for proper urban design. More allowance for carless urban design. This would compact pretty much every aspect of life as an autonomous car doesn't need a freeway to get somewhere fast. Even if they took two long trips, they would do it faster, safer, more fuel efficiently and more orderly. Congestion wouldn't even mean the same thing. Perhaps there would be 3x as many cars, but they wouldn't be behaving like people drivers.

[dubu]

I'd like them for traveling long distances. Instead of a 20 hour train ride it would be 10 hours with a driverless car