COMMENT OF THE DAY: HOW TO EITHER ELIMINATE OR WILDLY EXACERBATE THE HOUSTON TRAFFIC JAMS OF THE FUTURE 
“Once people figure out that it’s cheaper to keep the autonomous cars on the road rather than storing them in parking garages, and that having them circle the block at a very low rate of speed in perpetuity keeps them in a ready state – ready to zoom off to pick up passengers and bring in income while the owner is at work – then, and only then, will Houston residents know the true meaning of ‘gridlock’ and immediately wish that legitimate mass transportation alternatives had been built long ago.” [justaswell, commenting on Comment of the Day: Holding Out for that Younger, Sexier Mass Transit Option We Haven’t Met Yet] Illustration: Lulu
Sign up for the Newsletter
this is ridiculous, obviously if everyone is car sharing their autonomous cars fewer people will buy cars. If there are so many cars around that there arent enough riders to fill them then they wont make money, another reason there will be fewer of them. This guy needs an economics lesson.
I’ve seen a many Strawmen, behold a first Strawcar.
Yet more light rail whining. Zzzzzzz
WAY past old. Get over it already. We’re too spread out.
i think all these companies (especially uber) are dramatically underestimating how difficult it is to solve the set of problems that are needed to have fully autonomous cars. this is pretty much in line with the tech bubble that we are in where these companies are making ridiculous technology claims to keep the investment capital flowing. a more reasonable goal that will be commonplace a lot sooner is sortof a hybrid where easy parts of the journey are turned over to autonomous control, but we are a long way off from having the entire trip reliably handled by a computer. there are just too many weird situations and road conditions that need human judgement to be handled safely and solving those problems will take a while.
.
its kinda like machine translation. people have been working on that for a really long time, but, even with newer approaches that have significantly increased performance in the last several years, the results of machine translation still tend to look a little clunky and have problems with slang and things like that. getting it 80% of the way there is definitely an achievement and can be useful for some applications, but the remaining 20% are some very tricky problems that probably won’t be solved any time soon.
@I love Heights Walmart Yes, there will be less cars in the market. But 100% of autonomous cars will be on the road vs parked. That’s the point. You only pay for it when you use it instead of paying for it when it’s parked at your home or work. Less total cars in the world but more on the road. This is an economics lesson.
.
So the number of cars on the road will be higher than now. Right now, cars on the road exactly equals demand as people are only driving when they need it. When they are all autonomous, supply will never exactly match demand and therefore will always be over supplied trying to reach the next potential demand.
I believe we’re going to have the EXACT same issues with robocars as we do with personal automobiles. Any other conclusion is blind optimism untempered by the realities of human nature and motives.
This feature has been talked about by Tesla and I see it as something as a marketing gimmick to try and drive sales. There aren’t presently enough Tesla vehicles to so completely crowd out the market for rideshares or to induce congestion on their own and that won’t change for a good long while. By the time that that gets to be a factor, I suspect that professionally-operated fleet vehicles will have so completely undermined the private contractor form of ridesharing that the incentives to participate simply won’t line up. The professional operators are bound to contract with parking garages and shopping centers for staging areas, and they’ll be regularly-spaced in accordance with the shifting geography of demand and infrastructure throughout the day.
.
The other thing that *is* going to happen is an overhaul of the transportation code. This is inevitable, and it is already something that legislators (and lobbyists, more importantly) are actively contemplating. There are bound to be misfires with unintended consequences, but that’s par for the course. On the whole, this future tech opens up wonderful possibilities for improving transportation finance so that is closer to optimal.
.
@ my2cents: I agree that there’s a lot of fluff, but this stuff is being tested in Beijing. If these things are able to make decent headway in a big Asian city and make sense of the crazy unpredictable stuff drivers do over there totally in contravention of law, politeness, or sanity…well I’m impressed.
.
Also, if you’ve ever tried to learn a vastly foreign language or teach somebody who speaks a vastly foreign language how to speak yours, then you ought to understand how poor that analogy is. Language is symbolic and so abstract that it obfuscates human perceptions and the capabilities of philosophers. Driving is spatial. Provided good enough environmental awareness and computing power, software engineers can handle that stuff really well. Where they run into problems with AI and robotics is with stuff like tactile sensation, fine motor abilities, and…language. I suspect that the implications of this could be that anybody wishing to drive over unsurfaced roads or off-road terrain will have to have privately-owned vehicles, a steering wheel, and manual overrides, but that >95% of people will be well-served without them. Where problems exist that could overwhelm programming, that there will simply need to be new minimal standards for roadway design and maintenance, error reports to identify where there are problem areas, and financing mechanisms in place to deal with the issue.
It will be a while before a true Level 5 autonomous vehicle will exist. Anything less than that will be politically unpalatable. In fact, it may be so far into the future – and companies are so desperate to introduce these things in the near term – that some type of compromise may be introduced, say, an ultra-cautious smartcar type of design relegated (as much as possible) to the side streets, and with some type of predictability and consistency in terms of picking up passengers, like an autonomous jitney service. Otherwise, add in too much risk and there will be required a human to sit in the front seat (for manual overrides, and for other safety reasons), in a lot of ways basically defeating the purpose of an autonomous vehicle.
Why do people think that the same people who aren’t willing to share rides now will begin sharing rides just because the car is driving itself?
“there will simply need to be new minimal standards for roadway design and maintenance, error reports to identify where there are problem areas, and financing mechanisms in place to deal with the issue”
.
Right, so to support the pipe dream of fully autonomous taxis replacing personal car ownership, we need a ton of new infrastructure. Why not just go with proven high-capacity technologies that already exist, like BRT and rail? It’s much cheaper to electrify a bus or train than buy up high-tech batteries for a bunch of inefficient cars to get low-carbon transportation. Hell, it’s way cheaper to just free up regulations to allow dense urban development and just let people walk/bike around.
.
I can see autonomous vehicles taking over where there already is a heavy carpooling culture in HOV lanes like the DC suburbs. And sure, probably in core downtown areas where things are more controlled. But like others have pointed out, just making an Uber driverless isn’t likely to convince most suburbanites to ditch their cars.
Soooo, no one going to bring up that the CIA can hack cars now. Just think if they are autonomous.
@ GoogleMaster: Early adopters will begin sharing rides for the same reason that carpoolers do right now, which is to reduce their transportation expenses and gain access to HOV lanes. This is not a novel concept except that the way that it is packaged. What *will* change is that if you bring a service like Uber into that space, and ultimately multiple competing fleets of vehicles, then there’s greater flexibility in terms of on-call responsiveness, reliability, and menu pricing in terms of vehicle type and the number of other passengers.
.
The menu pricing is especially important. Automotive expenses take many forms and many or most car owners have a difficult time putting them in perspective with their household budget. At a dealership, in particular, a consumer tends to make purchases that firstly fulfill their needs for establishing a positive social identity (which is ridiculously expensive), and often rationalizes that choice by shopping for a vehicle which meets >99.5% of their transportation needs…even if it might be less expensive just to buy a subcompact sedan and rent a pickup truck if and when that need arises. Then they have to keep and maintain a big expensive machine which is rarely efficient and is costly to insure; and this capital is mostly idle, only used for ~10% of the day. Most people aren’t going to add up the myriad costs or consider depreciation, and weigh the utility of those expenditures. Now compare the American road fleet that is owned by consumers with the fleets owned by car rental companies. Rental fleets are menu-priced, and those prices take into account both fixed and variable costs and are responsive to calendar-driven or event-driven fluctuations in demand. Their inventories are weighted toward the low end, with relatively few trucks, SUVs, minivans, or sports cars. Consumers do have the choice to rent the same sorts of cars that they tend to purchase, but make more pragmatic decisions when those decisions involve a shared vehicle, one that is only theirs for an ephemeral moment in life which does not weigh on their sense of social identity. Are you seeing the analogy yet?
.
As demand snowballs and more people use the service, the quality and price should improve even further.
@ Derek: No, in order to support the non-pipe-dream of facilitating self-driving jitney services, we should make the same capital expenditures that have already proven the most successful without self-driving jitney services. Pavement is a proven technology, and it handles the vast majority of people in every city whether they are riding in cars or carpooling or in local buses or express buses or are making a connection to or from a railway station. Tech doesn’t get more proven than that.
.
I am not necessarily suggesting that these ought to be electric vehicles. If you can get enough people to share space in the same vehicle then fuel costs per capita (a variable cost) become a minor consideration in terms of expense or environmental impact by comparison with the cost of supplying electricity.
.
Also, if you want to support urban development then the very best thing that can be done is to reduce the total number of vehicles in the American fleet. The reason we still have so many parking lots and parking garages in urban areas is that those continue to generate revenue for their owners. Likewise — talk to an office broker — the reason that the downtown market has difficulty competing with other office submarkets is that office tenants have to pay for the cost of employee parking. And that cost is astronomical. The fewer cars there are going into downtown and the less often they sit there as idle capital, the greater is the impetus for parking lots to be redeveloped more densely. And then…you’ll get what you wanted all along.
@TheNiche,
I think I recall from previous messages that you no longer live in Houston. Maybe where you are now, there are a lot of carpoolers, but carpooling just isn’t a big thing in Houston.
.
* Kinder Houston study 2015: “more than 80 percent of respondents said that they travel alone by car to get to work” https://kinder.rice.edu/blog/shelton062915/
* TAMU Texas study 2016: “In the last year, have you carpooled
to avoid congestion?” -> 85% No, both Texas overall and Houston region https://tti.tamu.edu/policy/texas-transportation-poll/
* TAMU HGAC study 2014: 56% always drive alone, 18% always use commute alternatives, 25% both commute alone and use alternatives https://www.ghcommutes.org/findings/
.
What we have found time and time again is that, when people say they’re in favor of more mass transportation to ease congestion, usually what they mean is more like “I’m in favor of all of y’all taking mass transportation so my single-occupancy-vehicle commute will be easier.”
.
When we want to get from point A to point B, we want the freedom to leave NOW, without waiting for other riders and without deviating from the shortest, fastest route from A to B to pick up and drop off other riders. We want to be able to go from home to work to choir rehearsal to home on Monday, home to work to home on Tuesday, home to work to gym to home on Wednesday, home to work to second job to home on Thursday, etc. Carpooling / ridesharing only works if your schedules and destinations match someone else’s.
.
I put my money where my mouth is; I live only 2 miles / 10 minutes from work. I have ridden my bike and walked the trip on occasion, but usually I drive, because I have something to do after work almost every night of the week and the activity is somewhere else in the city far away from either work or home. Sometimes I need to stop at the grocery store or the pharmacy on the way home.
@ GoogleMaster: You’re correct that about 80% of workers drive alone in private automobiles. The Census Bureau’s 2015 5-year ACS indicates that about 317,378 (or 10.7%) of workers living in the Houston MSA carpool to work by car, truck, or van. Despite substantial investments, only 65,709 (or 2.4%) of workers use public transportation.
.
The Census Bureau also estimates the time spent in commuting by mode. Of the 77 million minutes that workers spent commuting in a car, truck, or van, approximately 12.7% of those minutes are spent in a carpool. When you consider the impediments to carpooling which presently are very real and costly, I don’t see that as necessarily very discouraging. And also, when you consider the nature of congestion and that it doesn’t occur until hitting a certain level of volume and then becomes exponentially worse with each additional vehicle, I also don’t see these figures as being insubstantial.
.
Carpooling stats vary within the workforce. About 15% of workers who make less than $25,000 per year commute in a carpool, whereas only about 6.5% of workers who make over $75,000 carpool. People who carpool also tend to be fairly young.
.
The Downtown Management District did a study in 2013 about how workers access downtown and found that only 56% of workers there were commuting by a private single-occupancy vehicle. Park & Ride accounted for 25.8%, other transit for 6.2%, and carpooling was 9.3%. Carpooling was most common on a per capita basis from dense urban neighborhoods inside the loop, particularly Montrose, the Washington Avenue Corridor, and the East End; however the sheer numbers of people that commute from the suburbs were such that over 70% of carpoolers into downtown originated from areas greater than 20 miles away, and a disproportionate amount originated from areas that were poorly served by METRO, like east Harris County. In downtown, the demographics of carpoolers were heavily skewed by suburban residents in their 50s. The study indicated that the ease and availability of Park & Ride services suppressed carpooling numbers.
.
The same study indicated that out of the downtown employers that offered only parking subsidy, 68% of workers commuted by single-occupancy vehicle, whereas out of the employers that offered only transit subsidies, only 29% commuted by single-occupancy vehicle.
.
The lesson in this is that people are sensitive to cost and convenience of transportation based on their individual circumstances, and many are flexible. Yes, right now the 80% figure looks bleak and overwhelming. The point of this argument, though, is not to establish how things are, but how things might be changed; and to your point, that change is possible.