Support BikePortland - Journalism that Matters

Could driving get catastrophically cheap? One smart Portlander speculates

Posted by on January 15th, 2016 at 11:55 am

car2go in the wild

Still pretty expensive.
(Photo: J.Maus/BikePortland)

Here’s an interesting argument about the near future from a Portland entrepreneur who’s also been a longtime advocate for biking and mass transit.

It’s a useful counterpoint to yesterday’s news that the U.S. Department of Transportation will kick off a $4 billion program to write national rules for self-driving cars.

The futurism here comes from Steve Gutmann, a Southeast Portland resident who has worked at various transportation-related startups since becoming an early employee at the local company that later merged with Zipcar. He’s also on the board of the anti-sprawl group 1000 Friends of Oregon.

gutmann

Steve Gutmann.
(Photo: M.Andersen)

In an email this week sent to BikePortland and a few others, Gutmann described a future where self-driving taxis will function like car2go, except cheaper because the cars will spend less of their time out of service. With that level of price and convenience, he predicted, large numbers of people would switch from public transit. Here’s our exchange, one lightly edited email at a time, starting with Gutmann:

Uber, Lyft, Tesla, Daimler, Ford and many others are all piling in. Massive change is coming to the transportation sector very, very quickly.

The key policy response, IMO, is to ensure that transit, walking and electric bikes are cheaper to operate than driverless taxis, because a car will otherwise always win out over other modes, and the final prediction of the video — that congestion won’t go away — seems likely to come true.

Which would suck.

Hopefully we’ll have the foresight to tax taxi fares (which will likely be very low, sans drivers…) to the point where everyone has an incentive to choose other, less congestion-and obesity-inducing modes whenever possible.

But regardless, if driverless taxis eliminate most personal cars, one nice co-benefit is that we’ll be able to stop arguing about parking congestion and minimum parking requirements! : )

Fasten your seat belts, this is going to be interesting.

Advertisement

I wrote back to ask how cheap he thought autonomous taxis could get compared to car2go, and to get his take on whether transit agencies are likely to autonomous buses to lower their costs, too.

Gutmann said:

If they can come to the next customer, they’re almost never idle, and I’ll bet a single car could be twice as productive.

In terms of competing, cars already out-compete transit for most people. If they become way cheaper to use than they are now, while transit agencies continue to drift toward becoming transit driver retirement funds that run a few busses, they’re going to lose. The cost of a transit driver (and another one who has already retired) is a huge competitive burden for transit agencies.

So yes, I think the transit agencies need to adopt driverless vehicles asap, or they risk losing ridership very quickly.

I replied that buses and trains will always be more space-efficient than cars during congested rush hours, but only if they have their own dedicated lanes:

Seems to me that this scenario leads to big drops in off-peak transit trips as more and more poor-ish people can switch from bus-oriented lives to automous-vehicle-oriented lives. But with peak-hour trips the time economics are way different and transit still wins, at least to the extent that we can continue improving its access to right-of-way.

Transit would then have a big problem with its assets being underused in off-peak hours, sorta like carsharing companies do today.

Gutmann:

Yup. So maybe we need traffic cordons (pay a $5 toll to take a Car2go inside the cordon…) so that they’re only used to drive people TO transit (vs. competing with transit), or for trips that don’t end downtown…

The other possibility is that buses just focus on trunk lines, running more frequently along main corridors, and they stop trying to serve unprofitable corridors. They could simply have fewer (and driverless) buses going up and down the main corridors all day long. Higher frequency at rush hour, but still being used off-peak, but only along certain high-demand routes.

All of this has been discussed elsewhere. But if you’ve been following this week’s tech news, you know that Gutmann may well be right that self-driving cars could spread rapidly. (In another email, he speculated that automaker Tesla could equip its upcoming Model 3 with “a ‘share’ feature that turns it into a peer-to-peer carshare vehicle, available for one-way or round trip (driverless) taxi rides.”) And it’s something that seems almost completely absent from our local transportation debates.

— Michael Andersen, (503) 333-7824 – michael@bikeportland.org

NOTE: Thanks for sharing and reading our comments. To ensure this is a welcoming and productive space, all comments are manually approved by staff. BikePortland is an inclusive company with no tolerance for meanness, discrimination or harassment. Comments with expressions of racism, sexism, homophobia, or xenophobia will be deleted and authors will be banned.

132
Leave a Reply

avatar
25 Comment threads
107 Thread replies
0 Followers
 
Most reacted comment
Hottest comment thread
37 Comment authors
DaveBDan SturgesAdam H.PeteTrain Engineer Recent comment authors
  Subscribe  
newest oldest most voted
Notify of
Todd Boulanger
Guest
Todd Boulanger

Yes – and the big problem of successful transition to driverless cars (EV cars too) would be the same land use problems we have now per sprawl: large unwalkable superblocks of land types, big fast roads, and parking lots as access barriers / density reducers. Maybe the parking lots would get smaller…but our roads might get bigger to handle longer peak periods as all these driverless car zip around for trips vs. being stored 90% of the time.

Todd Boulanger
Guest
Todd Boulanger

Driverless vehicles could have one small silver lining for transit agencies…if these vehicles can serve the paratransit market (either as a public transit service or as a private asset)…right now the cost to provide paratransit (per the ADA) is generally 6x to 8x the cost of a passenger for fixed routes in this region ($35 to $40 per trip)…and is becoming a cost that may bankrupt our transit providers unless something changes (ADA law, employment contracts for paratransit operators, higher supply of fixed route trips to subsidize paratransit, or reduced demand).

Adam
Subscriber

It’s insulting that the car companies are getting a $4 billion paycheck while people who walk, ride bikes, or take public transport still fight for pennies. The feds have again shown their incredible short-sightedness here, by touting low gas prices and putting all their eggs in one driverless basket.

SD
Guest
SD

Please tell me if I am wrong, but one concern I would have is an increase in the number of zero occupancy vehicles. Personal car ownership may decrease the total number of cars while increasing the numbers of cars on the road (traffic) at any given moment. I could imagine that the market may continue to favor the convenience of having a car immediately available over environmental sustainability or quality of life, e.g. less time in car or bike/ped accessible streets.
Essentially, free parking may become continuous driving, or the cost of continuously driving is less than parking.
With our current appetite for building and widening roads, wouldn’t we just continue to make the argument that we need to build greater capacity to fit all of the ZOVs?

Scott H
Guest
Scott H

$5-day/$100-month is already a ridiculously high price for a trimet pass, the day I can use my phone to call a Tesla to my house and take me to work for $2 will be the nail in the coffin if we can’t figure out how to tax electric/automated taxis to subsidize public transit. A street fee isn’t going to do it, a gas tax isn’t going to do it. Whoever replaces Novick needs to be able to see more than two years into the future.

soren
Guest
soren

Mass transit will also become more flexible and agile. For example, driverless vans and small buses could serve flexible routes that are optimized algorithmically to shorten travel time for users.

In meso- and south america “combis” are the predominant form of mass transit and they work very well indeed. I think that mass transit in the USA will come to resemble the “combi” model — except with EV vans.

mike
Guest
mike

Well, they won’t “still use oil” but yes they will perhaps still cause problems. I watched this recently, interesting thoughts about how it might HELP with congestion.

http://tonyseba.com/portfolio-item/clean-disruption-of-energy-transportation/

Champs
Guest
Champs

Social engineering is certainly possible with fleets of autonomous commercial vehicles. It also seems unnecessary if livable communities are a natural state.

Utopian scenario as alternative to dystopia: demand for low-car lifestyles and creative transportation pooling will *increase*. Running meters make the savings quite evident.

Brian E
Guest
Brian E

I see more potential with driverless cars delivering packages and freight. Get one to deliver your groceries, etc.

Peter W
Guest

Planning for a future of/with self-driving cars was a topic brought up by members of the Citizens Advisory Committee during the last Washington County TSP update.

Unfortunately, staff did not spend much time (or ink) diving into that topic (they decided that the plan update would be a ‘minor update’, regardless of the future looking to bring major updates).

Chris I
Guest
Chris I

MAX will be driverless long before we see more than 1% modal share of driverless cars on the roads. Fixed guideway transit is relatively simple to automate. Several cities already have driverless transit vehicles.

The biggest issue with the driverless car pipe dream is the fact that the majority of Americans want to drive themselves, or don’t trust the technology. Most of the benefits of driverless cars (reduced following distances, better coordination of traffic, etc) require that 100% of the vehicles on the road are driverless. Throw one “old” 2015 F150 with a driver from Estacada into the mix, and delays are created. This will require one of two things:

1. Separated infrastructure for driverless cars and traditional cars. Obviously this will be politically and physically difficult.
2. Wait until tradition cars are all off of the road. This would take decades. We still have people driving classic cars from the 1940s on our roads.

Anthony
Guest
Anthony

“So yes, I think the transit agencies need to adopt driverless vehicles asap, or they risk losing ridership very quickly.”

I would argue that they lose either way. While transitioning to driverless vehicles might help them save money due to not having to pay and provide benefits for employees, how many people are going to want to ride on a driverless bus? Studies have shown that people are already hesitant to cede total control to a personal vehicle, and I would guess that it would be even worse for a large transit vehicle. So I don’t really see this working out for transit agencies either way.

Todd Boulanger
Guest
Todd Boulanger

One policy to consider for ZOVs would be to segregate them to the far right lane and keep the HOV lane open for POVs (passenger occupied vehicles) as ZOVs would be a second lower class of road user with less concern for lost time…unless robots are enfranchised or corporations owning ZOVs fight this future law.

SteveG
Guest
SteveG

Anthony-

I wouldn’t be so sure that people won’t accept driverless transit. Check this out:

http://www.techinsider.io/driverless-buses-will-hit-the-road-in-switzerland-2015-11

http://www.telegraph.co.uk/news/uknews/road-and-rail-transport/11180429/Driverless-buses-on-the-way.html

http://www.popsci.com/driverless-buses-go-with-traffic-flow

http://gizmodo.com/the-first-driverless-transit-system-in-the-u-s-starts-1635908058

It’s just a matter of time before this hits Portland. The main drivers will be safety (many driverless vehicles are already far safer than human-operated vehicles) and costs. Professional drivers are expensive.

Charley
Guest
Charley

It’s very simple: no car should be allowed to operate without a passenger or driver in the vehicle. Move a foot without a human onboard, and get a ticket. If that were the law, car companies would simply have to program their cars that way. I suppose someone could hack it, but then they’d be at risk of a ticket. It wouldn’t be impossible to circumvent the law, but such a law would be a great, simple way to cut down on some of these nightmare ZOV scenarios.

Hello, Kitty
Subscriber
Hello, Kitty

It seems to me that a cheap driverless vehicle could provide a far higher level of service to an individual than bus service does currently. I do suspect that if vehicles are available on a per-ride basis, and the price is low, it may put Trimet out of business.

I’ve thought about this a lot, and my current thinking is that this will be a good thing.

Mark
Guest
Mark

I don’t buy the hype. This list of “Self-Driving Car Doomsday Scenarios” sums up my thoughts pretty well: http://transportblog.co.nz/2016/01/07/doomsday-driverless-scenarios/

Joseph E
Guest
Joseph E

Re: “except cheaper because the cars will spend less of their time out of service.”
I strongly disagree with the contention that driverless taxis will be cheaper than Car2Go or similar services in cities. Taxis lose value per mile, not per year, because they are driven into the ground before they can become obsolete. More frequent use will make the cost of parking a little cheaper, but this will be offset by the higher cost of vehicles due to the need for expensive electronics and sensors. A Car2Go is a simple vehicle, with mainly mechanical systems. A Google car has very complicated electronics; they are not going to be cheap to maintain.
The cost per mile for a driverless taxi will be the same as for a carshare or Car2Go vehicle: about $0.35 to $0.40 a mile for long distances in rural areas, or $0.60 to $0.70 a mile in cities. It will be very difficult to get costs lower, even if gas prices stay low.

John Liu
Guest
John Liu

Imagine if it were very cheap, very comfortable, and very convenient to get driven around by a taxi. Many people would take taxis instead of driving and parking their personal cars. The amount of traffic on the roads would not be reduced, but the amount of parking needed would be much less. With less demand for parking, today’s curbside parking could be converted to bicycle lanes. Today’s parking garages could be converted to bicycle parking and other uses.

Now, suppose these taxis were all driven by drivers with perfect vision and awareness and control, instant reaction time, always courteous and responsible, strictly obeying all laws. This would mean they would adhere to posted speed limits, not hit other cars, not hit pedestrians, not cut off cyclists. This would mean cyclists could ride in bike lanes, or indeed in any lane, at whatever speed, without fear of being hit by these taxis.

Replace “taxi” with “autonomous car”.

These autonomous cars will, of course, consume energy to get around. However, they could be electric. And they could be very small, hence energy-efficient. Private cars, because most people only have one, are chosen to meet any need the person ever has, however infrequent. Even if the person usually drives alone and for just 10 miles, he’ll buy a car that can carry the whole family plus luggage for the occasional long road trip. But when he summons the autonomous car for his 5 mile commute, he’ll happily ride in a small vehicle (with no driver, even a tiny vehicle can be very spacious for the passenger) and doesn’t care what the car’s range is (as long as it is at least 5 miles).

There will also be fewer autonomous cars, compared to today’s conventional cars. Since each autonomous car is carrying people 24/7, rather than sitting idle 23 hours/day, there don’t have to be as many cars per capita. Manufacturing fewer cars reduces energy consumption, hugely.

So, imagine a future Portland where there are just as many cars on the road as today, maybe even more, but all those cars are the size of Smart Cars, are driving 20 mph, never break traffic laws or hit pedestrians or cyclists, and are silent and electric, and all the previous curbside parking lanes are wide bicycle lanes, into which no car ever ventures.

That doesn’t sound so bad, does it?

Will we ever get there? Don’t know. It’ll take many decades to get all the way there. Maybe a few autonomous cars will be in our roads by 2020. I think the first ones might well be luxury cars that you can switch to self-driving mode as an alternative to driving yourself. Those wouldn’t bring much societal benefit, but I don’t see them making things worse either. To really get to the “future Portland” described above, we need, basically, autonomous taxis.

Joseph E
Guest
Joseph E

The one place where driverless will make a huge difference is in small towns and rural areas. Right now a taxi with driver can’t make any money in a rural area with infrequent demand for taxis. But since the cost of a driverless car will mainly be per mile, it can still be profitable making infrequent trips. This will be particularly helpful in low-density areas, to get access to intercity bus or trains, or for trips to the airport from small towns. It will also make it possible to have a driverless taxi waiting at a train station or bus stop out in the suburbs, to get people the last 1 mile to home. Someone could pay $1.50 a day for a 1 mile round trip to the transit stop, instead of paying for parking their own car, and then take the bus or train to the city. This will make it much easier to provide transit service to lower-density areas without building huge parking lots.

gutterbunnybikes
Guest

Not sure how anyone can really be opposed to driverless cars.

They obey the law, they aren’t drunk or stoned, they aren’t distracted, they are predictable, and I’m going to go out on a limb and say they will increase volume yet lessen congestion because easily 90% of traffic hold ups are people caught in intersections blocking opposing traffic, collisions from human error, red/yellow left turns, and they’ll be able to tailgate and draft each other.

though there is no data to really back it up, I’ve long suspected that human error is as much a contributor to congestion as traffic volume.

B. Carfree
Guest
B. Carfree

We’re all assuming that autonomous cars will not crash into us. I think part of that has to do with Google’s management of perceptions. However, they were recently compelled to hand over their “human intervention records” and those show that, had the “driver” not intervened, Google cars would have crashed into things eleven times in 424k miles. That’s a rate of about once every other year, based on current average miles driven. Autonomous cars have a ways to go to reach our expectations.

SteveG
Guest
SteveG

Here’s a long list of driverless transit systems:

https://en.m.wikipedia.org/wiki/List_of_automated_urban_metro_subway_systems

Seems To me that TriMet should start by exploring technology that would allow for the elimination of drivers for MAX trains. Then BRT and then busses in general. This would keep TriMet solvent without having to constantly increase payroll taxes, which punishes job creation.

Dan Sturges
Guest

I hope Steve runs for Mayor, as we are going to need mobility-intelligent public leaders as we progress into this future.

I have to laugh, we look at the cute Google self-driving car as our big future. Actually we should be thinking about Generation 8 of the concept. As the Google “car” is like the first Apple MacIntosh with 128K memory standard.

All this back and forth about Cars vs Transit. If you are riding in it, not driving, and it’s taking you somewhere, well, it’s “Transit”. In this AUTOMATED mobility future (yeah, forget “autonomous” – our vehicles will not be autonomous in the deeper future – your city won’t want them to be, and neither will you) any 1,2,3,4,6,8,12,18,24,48,200 person surface vehicle that is self-driving is a TRANSIT vehicle! 8 little 2 seaters that are full and traveling in a PLATOON – that means they are all traveling within 4 inches of each other, well that takes the same space of a 16 person can or mini-bus. Do you get the idea? It’s a complete new game.

Of course we want people using active modes over powered ones. Of course we want the smallest footprint of Potland’s coming Public Mobility System. And we need to mak sure every person can get to nearby commerce (hub) and not rely on corporate automated mobility system (Don’t want to be dependent on Big Sister)…

I think the thing I mostly never hear is how this future is about the shape and function of our future cities. We are always talking about transportation. I for one expect many aspects of our lives to change in the time me adopt AVs. Virtual Reality is one. Who has googles on when traveling through Portland? Does it look like their office? Are they crouched over in the back in some kind of truck hardly using much space and traveling way cheap?

I also expect more people to use Airbnb and live in a more distributed manner. Yeah the first AV I expect to see go by me is probably Steve’s neighbor’s “CLOSET”.

Instead of saying what the AV will be and how it will be bad, why don’t us people just make it right, how it should be. As if aliens were in charge!

Mayor Gutman! Like Architect-Mayor Jaimie Lerner in Braxil created Bus Rapid Transit. Here we are in a President election cycle and I sure wish we had fewer LAWYERS running for office and more ARCHITECTS and real CITY SHAPERS!!

DaveB
Guest

Academia has not been asleep about autonomous cars. There have been several modeling studies done about what would happen if all private trips were replaced by autonomous cars that could reposition themselves. One that looked at shared autonomous vehicles (SAV) serving a 10 square mile area of Austin Texas (twice the size of Car2Go’s current operating area in Austin). Each vehicle would serve 30-40 travelers per day, very few waiting more than 5 minutes throughout the entire area. Preliminary results indicate that each SAV would replace around eleven conventional vehicles, but adds up to 10% more travel distance than comparable non-SAV trips, resulting in overall beneficial emissions impacts, once fleet-efficiency changes and embodied versus in-use emissions are assessed.

(From: The travel and environmental implications of shared autonomous vehicles, using agent-based model scenarios by Daniel J. Fagnant, Kara M. Kockelman, 2009, published Transportation Research Part C 40 (2014) http://www.elsevier.com/locate/trc (not free)

There was another report from Europe in the past few years that modeled what would happen if all personal vehicle trips AND buses and train trips were replaced with shared autonomous vehicles. As I recall the results indicated it would result in more traffic than presently (although with much faster travel times for everyone) but concluded that the additional traffic would be mitigated by continuing to provide buses & trains on main routes. (Sorry, I’m not finding a copy of the report to quote it more precisely.)