Is parking too cheap?

[Shepard]

Do some or any of these towns have feeder bus service to the station - timed with the trains? Does that even exist anywhere in MA?

 

[Finn]

Raising prices is the short (immediate) term solution to making people happier (trading a small $ amount for the sizeable convenience of always finding a spot during your commute). And then places with high prices and full lots should rocket to the top of the parking-expansion list.

I'm glad to see the solid discussion here, and very much agree -- of course -- that the MBTA either doesn't understand matters of parking supply/demand or simply doesn't care. I think there's reason to believe that, at least in the short term, demand at this lot is relatively price elastic, since the ridiculous overcrowding is something that's really only started in the past 6 months. Folks have been shifting their habits away from whatever they were doing prior to the new lots being constructed.

There used to be additional parking across Foster Street, prior to construction of the new station, but that area seems to have been given over to signal housing or some such. There's a light industrial facility directly across the tracks which until about a year ago was occupied by Oldcastle Eggrock, which manufactured prefab bathrooms and things of that nature. There was some talk that the T might raze that building and convert it into an additional lot, but for whatever reason it didn't get done. Now Stoneyard Inc. occupies it (though they already had a facility in town, so the net economic benefit to the town is flat).

FWIW, the two new parking lots were built, for reasons beyond my comprehension, without a means of pedestrian egress between them. This means that anyone who parks in the second lot must either walk entirely around the first lot or, as a great many people do, forge a desire path by climbing over the two low-slung wood fences and traipsing up the steep grassy incline which separates the lots. It drives me crazy that nobody cared to fix this obvious design flaw.

 

[Arlington]

Why do people park? It is worth asking because the motivations are going to be changed by self-driving cars.

Today, people park because:

1) Owning a car seems the cheapest, fastest, most predictable way to get access to a car. You keep one at home (and like to have one at work) because so many trips start/end there. But if an on-call car is nearby it could be faster for you to "step into it" than for you to un-garage or scrape or start it.

2) Driving oneself seems the cheapest, fastest way to get access to a driver that you trust. "You" have always assumed that you knew the best way to get places, particularly your tricks for "the back way". But now Waze "sees" that just as easily.

3) Parking nearby (if it can be assumed...or even if you're crazy for thinking you can) shortens the "terminal walk" at the start/end of trips. But if there's a curb you could be dropped off at, that's much better than parking even very close.

Self driving cars are going to attack all 3 of these directly over the next 10 years.

Hopefully, by being cheap, fast, predictable, and trustworthy, they'll make "me driving myself" preposterous in ever-more commuting and leisure trips, particularly inside Route 128.

Raising the price of street parking would really really help speed this transition, and free up a whole lot of very valuable curb space and very valuable lane miles.

Cities are kind of at the same moment as when the Hansom Cab was introduced as the cheaper (single-horse) version of a two or 4-horse coach (with 4 wheels), and it suddenly became fast/easy/cheap to get through traffic in a cab (mid 1800s...until gas-powered taxis came in c.1910)

 

[Matthew]

Don't even need "self-driving" cars for that. New Yorkers have been living like that for decades with cheaper cabs and good-enough transit. With prices coming down on taxi-like service people in other cities are getting into that habit too.

Of course computer-driving makes it cheap enough to happen just about anywhere. But I'm not so sure it's so close. It's a very challenging problem in the general case. Most "self-driving" vehicles have operated under very controlled circumstances: either highway-driving, or finely-mapped and scouted neighborhood routes.

I think we might see computer-guided buses first. That simplifies the problem.

I worry that we might see another round of 'urban renewal'-style destruction in order to make cities compatible with self-driving cars... they don't like surprises such as 'pedestrians'. Though I guess they handle it better than some human drivers.

 

[Justin7]

^ Ten years is a long time. Ten years ago smartphones as we now know them were in their infancy. Think of everything your phone can do today. We are in a similar position now with self driving cars. I think it is reasonable to assume that in a decade a car driven by itself will better be able to handle pedestrians than a car driven by even the most capable human.

 

[Matthew]

Ten years is not really a long time It goes by waaay too fast. Sigh.

Anyway, regarding 'innovation', I think we're pretty good at benefiting from Moore's law, but that doesn't necessarily translate into the real world. As a computer scientist, I have a rather healthy skepticism for technological 'solutions' to everything, even while I seek them. I think it's rather telling that while we've become very good at making gadgets, and communications are amazing these days, we are still dealing with many of the same problems that plagued us in the 19th century: housing, transportation, segregation, education. Details have changed, but the overall questions remain the same. We have safer homes now, but affordability is still a crisis. With regard to transportation, we actually went backwards for a long time: we were able to move faster and faster but distances grew even worse. And speaking of 'technological utopianism': crazy people propose nonsense like the Hyperloop, while we can't even run the trains on time (or at all).

So yeah, we've seen a lot of improvement with regard to packing a lot of computing power in a small piece of hardware. That can result in some pretty cool stuff. But it's not necessarily attacking the truly hard problems. Remember, in the 1950s, computer scientists estimated that we'd have the AI thing tackled in ten years or so. It's 2015, we're not any closer, it seems. But we can build powerful supercomputers that can kinda, sorta match the pattern recognition ability of a two year old dog.

So count me as skeptical that we can 'solve' the fully-general self-driving car problem within ten years. If 'real intelligence' driving a car follows GPS and drives a car onto train tracks every month, what makes you think 'artificial intelligence' is going to do any better? But I do think that if you define the problem down, and scope it carefully, then we might be able to get something. So we'll see automatic protection systems that stop your car before a crash, for instance, which is already happening. Or perhaps on the highly regulated environment of the highway, you'll have the option of autopilot. And perhaps some vehicles can move autonomously along carefully studied corridors, with automatic protection systems in place to guard against crashes.

But I think we're a long way off from the point where you just type in an address and an autonomous vehicle shows up within 2 minutes to pick you up and drop you off anywhere in the world. I think that's quite possibly an AI-hard problem, and we know what happens with those.

 

[jass]

^ Ten years is a long time. Ten years ago smartphones as we now know them were in their infancy. Think of everything your phone can do today. We are in a similar position now with self driving cars. I think it is reasonable to assume that in a decade a car driven by itself will better be able to handle pedestrians than a car driven by even the most capable human.

Lol, no.

Hybrid cars hit the market 15 years ago. Today they make up around 3% of sales.

Rear view cameras have been around for over a decade. Only because of a federal law are they slowly finding their way into all cars.

You do realize that all the current self driving cars require a fully mapped route right? As in, they cant even navigate a parking lot on their own?

And no current model can drive in rain or snow.

Think 50 years

 

[sm89]

Why do people park? It is worth asking because the motivations are going to be changed by self-driving cars.

Today, people park because:

1) Owning a car seems the cheapest, fastest, most predictable way to get access to a car. You keep one at home (and like to have one at work) because so many trips start/end there. But if an on-call car is nearby it could be faster for you to "step into it" than for you to un-garage or scrape or start it.

2) Driving oneself seems the cheapest, fastest way to get access to a driver that you trust. "You" have always assumed that you knew the best way to get places, particularly your tricks for "the back way". But now Waze "sees" that just as easily.

3) Parking nearby (if it can be assumed...or even if you're crazy for thinking you can) shortens the "terminal walk" at the start/end of trips. But if there's a curb you could be dropped off at, that's much better than parking even very close.

I live in the City and because of that I don't want or need a car. It's not like everything is in walking distance to my apartment either. I take transit or bike almost every trip. I have this mentality because I'm from the suburbs and when I moved here I chose my housing knowing my mobility options. I think the biggest hurdle we face in Boston is that for many long time residents, the issue of parking hasn't been an issue at all until the past decade or so. They didn't all of the sudden decide to have a car. They've always driven and always been able to find parking. So for me it's easy to say who needs a car in the City, but for these people, they are being forced to conform to a new lifestyle that they never asked for.

 

[Matthew]

A lot of people grew up in a Boston (and a NY, and a Chicago, etc) where the city's population was shrinking due to white flight, 'urban renewal', highways etc.

Perhaps some people have become accustomed to a lifestyle based on the depopulation of the city: development can be blocked indefinitely through insane zoning rules, and free parking is taken for granted as a 'right'.

I know there's plenty of people who have stayed long term in Boston who don't feel that way, though, as well. This isn't Atlanta. I have talked to many long-time residents who deeply believe in the idea of Boston as a walking city, who feel strongly about the importance of transit.

 

[datadyne007]

The reason we owned a car on Mission Hill and had a resident parking sticker (and now do in Chelsea) is because my husband works in Needham and the Commuter Rail schedule doesn't work with his schedule. I work in DTX though and need the direct transit connections. It puts us in a situation where we're forced to own a car for his work, but I never use it and take the T everywhere. When we go out somewhere downtown, we just park at Wonderland. We don't drive to destinations downtown.

 

[dwash59]

Boston may be embracing car2go, the difficulty will be in getting enough spaces to make access predictable. If you rely on it for your commute and there aren't any left, you have to find some alternative. I ran into this semi-frequently when I tried to replace my bicycle with hubway. Knowing a car will be there when you leave is extremely convenient.

http://www.bostonglobe.com/business...rking-spots/usuZ2lxBBqIV7sHWQ5SOgK/story.html

 

[George_Apley]

The reason we owned a car on Mission Hill and had a resident parking sticker (and now do in Chelsea) is because my husband works in Needham and the Commuter Rail schedule doesn't work with his schedule. I work in DTX though and need the direct transit connections. It puts us in a situation where we're forced to own a car for his work, but I never use it and take the T everywhere. When we go out somewhere downtown, we just park at Wonderland. We don't drive to destinations downtown.

It's the same with my boyfriend and I. I work in Lexington and need the car just about every day. He has worked downtown/been in graduate school at the Charles River Campus and takes transit everywhere. There are a lot of split households like that, especially with recent, younger arrivals who grew up in suburbia. The "immediate suburbs" of old, like Cambridge, Somerville, Brighton, JP, Dorchester, etc., are the natural catching ground for these sorts of households.

 

[Matthew]

Car2go is a good idea but I don't think of it being as useful for commutes. It's more for utility.

It has an advantage over Hubway in that you can leave the cars in any legal parking spot. There's a bike share system with a similar idea out there, but we don't have it.

BTW, that Boston Globe article is absolutely horrendous. What a ridiculous lede. Oh no! 0.2% of parking spaces might be used for a convenient service that benefits everyone! Let's write over-the-top rhetoric and get everyone worked up over nothing, making Marty's life harder than it needs to be.

Idiotic articles like that, which are supposedly 'reporting' and therefore 'balanced', are one of the reasons I refuse to subscribe to the Globe. They don't deserve my money.

 

[Justin7]

Ten years is not really a long time It goes by waaay too fast. Sigh.

Anyway, regarding 'innovation', I think we're pretty good at benefiting from Moore's law, but that doesn't necessarily translate into the real world. As a computer scientist, I have a rather healthy skepticism for technological 'solutions' to everything, even while I seek them. I think it's rather telling that while we've become very good at making gadgets, and communications are amazing these days, we are still dealing with many of the same problems that plagued us in the 19th century: housing, transportation, segregation, education. Details have changed, but the overall questions remain the same. We have safer homes now, but affordability is still a crisis. With regard to transportation, we actually went backwards for a long time: we were able to move faster and faster but distances grew even worse. And speaking of 'technological utopianism': crazy people propose nonsense like the Hyperloop, while we can't even run the trains on time (or at all).

So yeah, we've seen a lot of improvement with regard to packing a lot of computing power in a small piece of hardware. That can result in some pretty cool stuff. But it's not necessarily attacking the truly hard problems. Remember, in the 1950s, computer scientists estimated that we'd have the AI thing tackled in ten years or so. It's 2015, we're not any closer, it seems. But we can build powerful supercomputers that can kinda, sorta match the pattern recognition ability of a two year old dog.

So count me as skeptical that we can 'solve' the fully-general self-driving car problem within ten years. If 'real intelligence' driving a car follows GPS and drives a car onto train tracks every month, what makes you think 'artificial intelligence' is going to do any better? But I do think that if you define the problem down, and scope it carefully, then we might be able to get something. So we'll see automatic protection systems that stop your car before a crash, for instance, which is already happening. Or perhaps on the highly regulated environment of the highway, you'll have the option of autopilot. And perhaps some vehicles can move autonomously along carefully studied corridors, with automatic protection systems in place to guard against crashes.

But I think we're a long way off from the point where you just type in an address and an autonomous vehicle shows up within 2 minutes to pick you up and drop you off anywhere in the world. I think that's quite possibly an AI-hard problem, and we know what happens with those.

Most of the issues you raise are not technological, but economic/political.

As far as AI, I hesitate to even call it that. The computer intelligence needed to navigate a vehicle is simplistic compared to something like a (legitimate) Turing test or even a personal assistant. A self driving car is already safer and more efficiant than a human in most instances simply due to reaction time.

We already have the technical capability. The computers, cameras, and sensors are already there. It's really just dev time at this point and I think ten years is plenty.

To be clear, I don't think everyone (or even a majority) is going to have a self-driving car in a decade. I just think they will be commercially available.

Lol, no.

Hybrid cars hit the market 15 years ago. Today they make up around 3% of sales.

Rear view cameras have been around for over a decade. Only because of a federal law are they slowly finding their way into all cars.

You do realize that all the current self driving cars require a fully mapped route right? As in, they cant even navigate a parking lot on their own?

And no current model can drive in rain or snow.

Think 50 years

Hybrid cars and rear-view cameras were not adopted more quickly because there is little demand and producers see little benefit in building them.

Google, Uber, etc. stand to make billions by eliminating the need for drivers.

The market forces here are in no way comparable.

I'm not sure what you mean by "require a fully mapped route."

 

[FenwayResident]

You do realize that all the current self driving cars require a fully mapped route right? As in, they cant even navigate a parking lot on their own?

And no current model can drive in rain or snow.

Think 50 years

This is so completely and utterly wrong it's hilarious that you included snark in your post. Self-driving cars can easily drive in rain/snow. I believe Google did their first successful snow test a few years ago. What you're citing is extremely adverse conditions. Self driving cars can't drive in stuff like Boston experienced recently - extreme white outs or torrential downpours - where it isn't safe for humans to drive either.

The parking lot issue is not really a big issue at all. The core mechanics of self-driving have all been figured out. It will just take some engineers a few months to figure out how the car should behave in a parking lot vs. on the road.

That being said, I'm not denying that there are still issues that need to be worked out before public use. But they're not huge issues, merely minor kinks, that need to be worked out.

On that note, recently six automakers at CES announced they will have self-driving cars on the market by 2020. Of course, for the first few years self-driving will be a luxury. Just like rear-view cameras, automatic breaking, collision avoidance, etc. were luxuries when they first came out. However I can easily see self-driving becoming mainstream by 2025, if not earlier, when regulators and insurance companies see the reduction in collision frequency brought by self-driving cars.

 

[Matthew]

A fully mapped route means that the computer needs to know every single detail of the streets and the potential obstacles, and it needs to be able to detect and react to ephemeral obstacles as well.

We do all this in our brains so easily that you don't really think about it. But the fact that you can look at a scene, pick out all of the objects, recognize them, know how far away they are, what shape they are, what their behavioral characteristics are, what the likely future trajectory of each object is, and to be able to do this under nearly all conditions, is an amazing feat that supercomputers struggle to even begin replicating.

Under highly controlled conditions you can simplify the problem quite a bit. That's where the "fully mapped routes" part comes in. And by fully, I mean, high quality pictures of every scene along the route, complete with preprocessing of the data to help speed up on-the-fly recognition. Don't forget to do this in a variety of weather conditions so that the learning algorithms have a chance to adapt to each one. You wouldn't want a pole to appear like a lane marking to the machine just because it happened to be raining...

I'm not sure I can get across just how difficult this kind of dynamic pattern recognition is to someone who hasn't tried to program computers at a high level. Rah rah, yeah, Google hires geniuses I'm sure, but they're smart people, not magicians.

We'll see applications of the technology, like highway autopilot, but the full replacement of human beings under any and all of the crazy situations that you can come across as a driver? I'm not gonna hold my breath.

 

[jass]

This is so completely and utterly wrong it's hilarious that you included snark in your post. Self-driving cars can easily drive in rain/snow. I believe Google did their first successful snow test a few years ago. What you're citing is extremely adverse conditions. Self driving cars can't drive in stuff like Boston experienced recently - extreme white outs or torrential downpours - where it isn't safe for humans to drive either.

The parking lot issue is not really a big issue at all. The core mechanics of self-driving have all been figured out. It will just take some engineers a few months to figure out how the car should behave in a parking lot vs. on the road.

That being said, I'm not denying that there are still issues that need to be worked out before public use. But they're not huge issues, merely minor kinks, that need to be worked out.

On that note, recently six automakers at CES announced they will have self-driving cars on the market by 2020. Of course, for the first few years self-driving will be a luxury. Just like rear-view cameras, automatic breaking, collision avoidance, etc. were luxuries when they first came out. However I can easily see self-driving becoming mainstream by 2025, if not earlier, when regulators and insurance companies see the reduction in collision frequency brought by self-driving cars.

This is so completely and utterly wrong it's hilarious

Heres a nice dose of reality for you.

Chief among these questions is how Google plans to collect information that will be programmed into its cars. Technology Review writes that each time a Google self-driving car gets ready for a trip; engineers make a handful of preparations, including extensively mapping the car's exact route. It is not unlike taking a Street View map of a road for Google Maps, but with the self-driving car, the mapping is much more detailed.

"Data from multiple passes by a special sensor vehicle must later be pored over, meter by meter, by both computers and humans," writes Technology Review. "It's vastly more effort than what's needed for Google Maps."

Heres the fun part:

Moreover, MIT notes that the cars are not 100 percent equipped to handle mapping omissions, such as road constructions, temporary stop signs or lane diversions. Although the vehicles are designed to recognize physical obstacles, they do not know how to respond to certain obstacles. For example, if new traffic lights were put up, the car would slow down as it approaches the stoplight but will not know how to obey it. The cars also would not be able to spot things such as potholes or an uncovered manhole.

Woopsie.

The cars also cant distinguish between something you can drive over (like an empty mcdonalds cup) or a rock. Imagine coming to a sudden stop because cardboard blows into the roadway?


And then you say

I believe Google did their first successful snow test a few years ago.

Reality says:

Another common obstacle the self-driving cars still have not learned to overcome is the weather. While human drivers have something called a brain sitting between their ears to help them navigate their vehicles through rain, fog or snow, Google's engineers still have to figure out how to teach their driverless vehicles to do things such as go slow in wet pavement or estimate the lanes when the road is covered in snow. In fact, Chris Urmson, the head of Google's driverless car team himself, admits the cars have yet to be tested in wet conditions because of concerns posed on the safety of the testing engineers.

Speaking of safety...

"The car wouldn't be able to spot a police officer at the side of the road frantically waving for traffic to stop," says MIT.

http://www.techtimes.com/articles/1...-quite-ready-to-hit-the-roads-researchers.htm


You also claim:

The parking lot issue is not really a big issue at all.

Reality says:

On the top floor of a Las Vegas casino garage, the company demonstrated how a modified version of its BMW i3 electric car can autonomously park itself and then can come pick you up when you’re ready to go–BMW calls it the Remote Valet Parking Assistant.

“You can send your car away to look for a parking spot, go shopping and then comes back again when you’re finished,” said Huber.

wait for it...wait for it....

To get readings of its environment, the car is equipped with four laser scanners on each side of the car. BMW also needs a map of the garage. From there, the car’s algorithm can sift through the map and the information coming off the sensors to find a parking spot.

Having to get a map of every single parking garage seems rather impractical for this feature to take off. Werner Huber, head of the research group for driver assistance at BMW, said it can start retrieving the maps straight from garage operators. For the demo, the BMW team created the map of the garage themselves.

http://www.forbes.com/sites/aaronti...self-and-picks-you-up-when-youre-ready-to-go/

A detailed map of EVERY PARKING LOT IN AMERICA?

AHAHAHAHAHAHAHAHAHAHA

In summary:

Google’s cars have safely driven more than 700,000 miles. As a result, “the public seems to think that all of the technology issues are solved,” says Steven Shladover, a researcher at the University of California, Berkeley’s Institute of Transportation Studies. “But that is simply not the case.”

Sounds like theyre calling you out.

10 years. Lol.

Self-driving on the highways (in good weather) in $50,000 cars is likely. Anything beyond that? See you in 2050.

 

[George_Apley]

Yeah, I figure we'll get "smart cruise control" that handles highway driving pretty soon, plus more advanced auto-parking, but then it will stagnate as far as actually responsive driving goes while software engineers battle with AI over true driving skills.

 

[jass]

Yeah, I figure we'll get "smart cruise control" that handles highway driving pretty soon, plus more advanced auto-parking, but then it will stagnate as far as actually responsive driving goes while software engineers battle with AI over true driving skills.

Yeah highways are easy enough. Most drivers are on mental autopilot anyway, and a computer could do that just as well, plus react better at braking.

Except when a plastic bag flies by....

Ahem.

Again I point to my hybrid/electric car example. The basics are there. The product launched 15 years ago. But its a miniscule portion of the market because price and range havent been solved yet. And those are pretty damn serious problems, even though tons of money and thousands of very smart people are working on them every day.

And as great as Tesla is, 200 mile range wont cut it for most people, and even the $30,000 they keep promising is still way out of reach for most people (and keeps getting delayed...)

Even if we get a 500 mile electric car at $20,000 in 10 years, it would have been 25 years of real commercial experimentation (never mind another 2 decades of lab work) to even get the car to become a majority of new sales, and another 10 years to become a majority of the fleet.

These things take time.

Why not start with giving the MBTA self-driving trains?

 

[FenwayResident]

Jass, all I'm reading in your giant post are examples that the technology currently is not ready. Which is exactly what I said. So thanks for agreeing with me on that one. I think where we disagree is how quickly the technology will become ready.

The fact that six automakers announced self-driving cars within five years is huge. They would never do that unless they were confident they could produce the technology in time. Clearly there is tons of money and effort being poured into this from multiple angles. I'm confident that these problems mentioned earlier will be figured out.

On the technical side - I'm aware that invariant pattern recognition is extremely difficult. My job is to implement it in finance. But when a ton of money is poured into a problem, solutions can be found. Look at the improvement in facial recognition over the last twenty years. First the money came from the government for security purposes, then it came from the private sector for advertising (i.e. Facebook). Facial recognition algorithms have improved by two orders of magnitude in speed and predictive capability.

Now that there's a ton of money being poured into self-driving cars, I believe the same will happen.