Connected/Automated vehicles and infrastructure in Boston

[shmessy]

^ I agree. The business infrastructure and the prevailing car-culture will take decades to be ready for a general switch from personal car ownership to automated taxi cabs.

There is no "car culture" in the central core of major cities.

Once again, I am talking about something that would be introduced over time through many concentric stages and would eventually end up being in place for no more than 5% of America's land mass.

 

[DuckVillen]

What a nice idea! I share your vision, but it ain't gonna happen that way.

"Bubble cars only" rules would be a pain in the ass for AV mobility providers. It would be a logistical hurdle and major operational expense to switch riders over to some micro vehicle once they're in town, and a big hassle to ask of customers who were promised that they could get from their suburban location to anywhere, seamlessly.

No, providers will optimize for the 95% case that handles American suburbs and "suburban cities", where you need to get onto high speed arterials and highways to get anywhere--and that means heavy, standard cars. You only get to evade the NHTSA rules if you stick to 25 mph and below (or if you're a three wheeler like an Arcimoto, but expect that loophole to be closed down if it becomes overused.) [*]

There will be enormous, top-down pressure to preempt any local regulations that would prevent a one seat ride into town, and there'll be bipartisan consensus for it. The "patchwork of local regulations" have to be harmonized as a matter of national industrial policy, after all--gotta beat those Chinese!--and it will be assumed that billions of dollars of profit are on the line, not just for AV manufacturers or operators, but for ancillary industries that expect to profit. And I'm sure some of those preempted local regulations would have looked like "you can only send a one-to-four passenger AV down this set of streets if it's electric, weighs less than 1000 pounds, is covered with a rubberized safety material, and scores 10 percentage points better than the national standard in the pedestrian-and-cyclist detection test--and empty cars are banned unless they are making a pickup reachable only from this street."

This has already started (see: the AV START act; various TNC attempts to preempt local regulation in statehouses) but I wouldn't expect them to stop at just preemption. I'm sure there are many ways to lean on cities to ensure providers' vehicles can get around, and they won't even be meaning to crush rules put in place to support enlightened urbanism--they'll take aim at the "patchwork" of "alarmist" and "luddite" safety concerns and accidentally hit the urbanist concerns.

For that matter, I worry somewhat about what this means for active transportation (such a pain having all those pedestrians and cyclists free to roam, occasionally getting themselves under our wheels and upsetting our customers by walking in front and forcing our cars to stop; could you make that problem go away for us? Maybe we could give you some money to put gates at crosswalks? Or we could send you videos of jaywalkers [**]--we'll even do the facial recognition lookup for free!)

The good news is that everyone seems to have overestimated just how soon wide AV deployment would be practical. For myself, I continue to think that there will need to be a lot more active infrastructure, like radio beacons and V2V, than the initial "we're not like those IVHS defense contractors trying to get your sweet, sweet highway dollars" PR position suggested, and real downtown areas will continue to be hard for AVs for a long time.

Now, the nice case is that providers might be persuaded to choose to encourage their customers to switch to a micro vehicle in town. I think there are various ways that could be made to happen, preemption or no, since the basic street-space economics support it. But you are fighting the big cost of switching vehicles and running a unique fleet in the few center cities that would support them.

[*] But, see the Talking Headways podcast interview with the policy head at Nuro Robotics (AV grocery delivery); they're apparently debuting a cargo-only microcar for grocery delivery in Houston, they're 25 mph only for now but hoping to get approved for 45 mph someday.

[**] True or false, for the ArchBoston crowd: the "jaywalking" fine is still $1 in Boston, right?

 

[ra84970]

Concentrically introduced from inner core in stages. It won't be a switch flip. Manually driven cars would be prohibited. There would be no mix.

Delivery trucks, etc. would be allowed only during off-peak times such as Midnight to 7am, or such.

And we are not talking about this stretching out to Newton or anything. Society can walk and chew gum at the same time. We can have a different transpo system in urban cores than we have in suburbs/rural. Parking lots/depots around the circumference.

Sigh.

Once again, the manually driven car as we know it is not going to go away in 95% of America..

I'm talking about inner core of cities. (where every inch of space/real estate is precious).

Don't fret, people in Weston will still be able to drive their reinforced tanks.

There is no "car culture" in the central core of major cities.

Once again, I am talking about something that would be introduced over time through many concentric stages and would eventually end up being in place for no more than 5% of America's land mass.

This spatial sequestering is not a tried-and-true path for transportation change. Change depends on overlays of older and newer technology - esp. in transportation. Clean breaks like you're hypothesizing are not realistic.

There are so many unresolved items of how autonomous vehicles (at Level 5 automation) could even function on streets in the urban core of metro regions. Some of the issues that we can already see are:

* Legal - Liability for AV-L5s is unclear in the case of a crash, error, or fault of the AV system. || Today, liability is on the operator. If a crash happens, liability is generally assigned after an investigation about operator error. If other faults happen, e.g. the hardware is ill maintained, does that mean it's on the AV-systems' manufacturer forever? Today, a warrantee continues for X number of years, but, after that time all the systems are the responsbility of the owner of the car. Do we continue with this model of maintenance liability when AV-L5s systems are now more responsible for safe operations?

* Financial - Depending on how liability is determined, then theres a question of how a autonomous vehicle operator is insured. Is there a requirement to insure and license the people in the vehicle? If the vehicle's AV systems have a fault, have an error - does one need to carry insurance to manually "drive" the AV? What about minors that maybe in an AV without a licensed "manual" driver? What about a single passenger who is not licensed for good reason? Like disability, youth.

* Infrastructure (public owned) - Many AV-L5 systems appear to depend on vehicle-to-infrastructure/infrastructure-to-vehicle communications to support AV-L5 systems on board vehicles- communication standards are not identified, who is working on this in a way that works? I don't see any pilots that are going on locally. Most road authorities can barely maintain a state of good repair, How do we expect them to successfully maintain the "I" of the V2I/I2V?

* Infrastructure (personal) - Some AV-L5 systems appear to also require all other road users to purchase/acquire technology to communicate with on-board AV-L5 systems. That is an incredible burden being placed upon other road users that other technological changes did not require.

Mass cycling is perhaps a useful foil -- mass cycling didn't require paving, but, it was enhanced by it. Paved paths also enhanced the transportation systems for other vehicles and other users. AVs require a lot of things from other trnaspo systems or a lot of people, without enhancing those other systems or other users.

 

[tysmith95]

The first fully automated vehicles will function like Ubers, with no personal ownership. Due to the economics of not having to pay a driver, it'll make car ownership obsolete for a large chunk of the population.

 

[bakgwailo]

The first fully automated vehicles will function like Ubers, with no personal ownership. Due to the economics of not having to pay a driver, it'll make car ownership obsolete for a large chunk of the population.

So.... maybe? Traditionally Uber heavily subsidized ride fares to bring costs down via VC/capital, and self-driving cars were basically their path forward to being profitable at the price point that consumers now expect. If anything, I would guess rates to the same, or go up.

As far as car culture in the core of the city, well, everyone I know who grew up in Chinatown had a car (and needs it down to commute out of the city for work). You have the examples of North Enders having multiple cars and abusing resident stickers, everyone I know from Mission Hill had cars, etc. It is certainly less than say West Roxbury, but, people still like cars and I doubt that changes in a decade. Well, unless we just skip right to transporters or something.

 

[JumboBuc]

Sigh.

Once again, the manually driven car as we know it is not going to go away in 95% of America..

I'm talking about inner core of cities. (where every inch of space/real estate is precious).

Don't fret, people in Weston will still be able to drive their reinforced tanks.

I fully understand you vision, shmessy. I just don't think it's at all practical or realistic.

If we wanted a system of totally closed bubbles, where we have one system of light, plastic, slow vehicles in the inner core and another fully segregated system of heavy, metal, fast vehicles in the suburbs and exurbs, we could have that today! There's no reason that such a system hinges on autonomous vehicle technology. It's not technology that is preventing that system from being a reality, it is organization and regulation. Obviously we don't have that system today, because such a model doesn't fit our society's vision for transportation. If we can't even erect a congestion charge to separate urban versus non-urban travel, how can we expect to erect a system of fully segregated urban versus non-urban roadways?

In this envisioned "two bubble" world there is no (or minimal) overlap between "inner core" vehicle travel and "outside of core" vehicle travel? That's just not how people move. It is entirely possible (and perhaps even likely) that both of the following are true: 1) the majority of miles travelled on Boston roads are driven by cars that live outside of Boston; and 2) the majority of miles driven by cars that live in Boston occur outside of Boston. People in "inner cores" need to leave those cores and people outside of cores need to enter those cores. In the "two bubble" vision, anybody passing from inside a core to outside of one or from outside a core to inside of one would have to stop at the border and transfer vehicles? That would be a huge inconvenient pain in the ass for everyone, including urban core residents (many of whom own cars today expressly for the purpose of getting out of their inner core neighborhoods). This "switch vehicles at the urban / non-urban border" model already stands for park-and-ride / transit. But still we have a ton of people passing on that option and driving between the urban core and non-urban core road networks.

The first fully automated vehicles will function like Ubers, with no personal ownership. Due to the economics of not having to pay a driver, it'll make car ownership obsolete for a large chunk of the population.

I'm a bit sympathetic to this argument (and used to be more so) but Uber and Lyft's key to success since day one has been offloading practically all of their expenses to unrelated third parties (their independent contractor drivers), and paying those third parties relatively little in return. For Uber and Lyft to bring all vehicle capital costs -- as well as the substantial liabilities of using unproven autonomous driver technology -- in-house would be a 100% reversal of their business model. Yes, Uber has been investing in AV. Part of this is being done under the hopes that it will one day pan out for them, but a lot of it is also being done because they're a VC-flush (and now IPO-flush) "startup" that needs to keep up the appearance of a "technological innovator" in order to keep pulling in investor money. But the cost of providing Uber rides has always been far greater than the price charged for those rides, and I see little likelihood that AV tech could reverse this in the next decade at least. And even though we already have capital-subsidized cheap and abundant Ubers, people still own cars.

-----------------------------

Everyone who has ever driven a car knows that some roads are easier to drive on than others. Driving 100 miles on an Interstate is a breeze that requires very little thinking. Driving 100 miles on "country roads" takes some more focus (you can't just set the cruise control) but is still easy. Driving 100 miles on suburban roads is also fairly low stress, but requires a bit more thinking. Driving 100 miles in an "inner core of a city" is a daunting, exhausting undertaking. And if you want to think in terms of time (e.g., 1 hour) instead of miles (e.g., 100 miles) the same dynamic still applies. This is how people interpret driving conditions and difficulty, and this is how computers / AV tech interprets driving conditions and difficulty as well. It's the exact same thought process.

Computers will be able to handle highways (to the extent that they aren't already able to handle them today) in the near future. Then they'll be able to handle "country roads." Then suburbs. And then cities will be last.

A very large percentage of total VMT and VHT are highway miles, and those are the easiest to automate. So it makes sense that those will be the first to be automated. First you'll see better and better "adaptive cruise control" technology to the point your car will drive itself from onramp to offramp (we already have some of this today!). This is especially applicable to long-haul commercial trucking, which is a much bigger industry than many people give it credit for.

AV / CV long-haul heavy trucks can deliver very large productivity gains without ever having to get into the difficult tech of complex urban street grids and unpredictable driver / pedestrian / cyclist behavior. We already have a major established model for gaining energy and labor efficiency by partially automating the movement of cargo across land: freight rail. Additional AV / CV tech that makes cargo trucks look just a little bit more like cargo trains would deliver enormous productivity gains. Current trucks can move only about one container per worker, while trains can move as many as hundreds. If AV / CV tech is able to decrease overall long-haul trucking labor intensity to an average of 2 containers per worker, that would be an enormous improvement. And this can be done with much less required disruption to our long-established transportation network and systems.

 

[shmessy]

The first fully automated vehicles will function like Ubers, with no personal ownership. Due to the economics of not having to pay a driver, it'll make car ownership obsolete for a large chunk of the population.

+1. This. I don't see individual ownership of these vehicles at all.

One's cellphone will be used to order up the ride to pick you up and you then dial in the destination. It will be a subscription app (who knows if there will even be cellphones by 2035, anyhow? Probably something else).

Remember, this is a real estate/economy issue. The amount of premium downtown space that will be freed up (think a comfortable future Boston population of 2+million) is what will drive this.

 

[George_Apley]

None of what you're saying is outlandish except the date. 2035 is fifteen years off. As much as the world has changed since 2005, our regulatory and insurance systems have not. I completely understand all of you points, I am just much more bearish on the sort of necessary changes being made on the timeframe you're using.

 

[real_EthanHunt]

The first fully automated vehicles will function like Ubers, with no personal ownership. Due to the economics of not having to pay a driver, it'll make car ownership obsolete for a large chunk of the population.

Only one edit, not just first, but all.
Transportation would become a utility. Used on demand without any ownership by the user.

everyone I know who grew up in Chinatown had a car (and needs it down to commute out of the city for work). You have the examples of North Enders having multiple cars and abusing resident stickers, everyone I know from Mission Hill had cars, etc. It is certainly less than say West Roxbury, but, people still like cars and I doubt that changes in a decade.

Do people like cars or do they like the independence of getting from place to place on their own terms? Give them an option with as much flexibility without the costs and issues of ownership and they'll switch in a heart beat. We've already seen it happen on a smaller scale with zipcar and TNCs

 

[shmessy]

None of what you're saying is outlandish except the date. 2035 is fifteen years off. As much as the world has changed since 2005, our regulatory and insurance systems have not. I completely understand all of you points, I am just much more bearish on the sort of necessary changes being made on the timeframe you're using.

I hear ya, George. And it may not happen by 2035 in BOSTON (although, I do think Boston is RAPIDLY changing - - maybe more so culturally/business-wise than any other city in US).

But, as I mentioned earlier, I believe it WILL happen by 2035 in places like Singapore/Tokyo/Portland, Oregon, etc. It will be out there and it will spread. And Boston isn't as old-timey and flat-footed as it used to be.

 

[shmessy]

Only one edit, not just first, but all.
Transportation would become a utility. Used on demand without any ownership by the user.


Do people like cars or do they like the independence of getting from place to place on their own terms? Give them an option with as much flexibility without the costs and issues of ownership and they'll switch in a heart beat. We've already seen it happen on a smaller scale with zipcar and TNCs

.......or the costs/headaches of urban parking.

 

[ra84970]

(Snip)
A very large percentage of total VMT and VHT are highway miles, and those are the easiest to automate. So it makes sense that those will be the first to be automated. First you'll see better and better "adaptive cruise control" technology to the point your car will drive itself from onramp to offramp (we already have some of this today!). This is especially applicable to long-haul commercial trucking, which is a much bigger industry than many people give it credit for.

AV / CV long-haul heavy trucks can deliver very large productivity gains without ever having to get into the difficult tech of complex urban street grids and unpredictable driver / pedestrian / cyclist behavior. We already have a major established model for gaining energy and labor efficiency by partially automating the movement of cargo across land: freight rail. Additional AV / CV tech that makes cargo trucks look just a little bit more like cargo trains would deliver enormous productivity gains. Current trucks can move only about one container per worker, while trains can move as many as hundreds. If AV / CV tech is able to decrease overall long-haul trucking labor intensity to an average of 2 containers per worker, that would be an enormous improvement. And this can be done with much less required disruption to our long-established transportation network and systems.

I think logistics and commercial freight is the sector in which automation makes the most sense for the transporation network -- for several reasons:

* LABOR - pre-pandemic, the logistics/freight sector was struggling to attract and retain labor at the prevaling wages. Wages were creeping up the past few years, but, are still awfully low for the particularly rough work that it is to drive long-haul fright. Thankfully, during the pandemic, demand for long and medium haul drivers is still strong. I struggle to see that the Teamsters are strong enough to fight off management's pursuit of automation, or, if it comes down to it, if some of the really bad parts of long-haul are alleviated and ameliorated by automation, if some factions may actually support some level of automation.

* TECHNOLOGICAL "EASE" - Interstate highways are as close to a "closed" environment that you'll get on the motoring network outside of actual test tracks. For Volvo, Ford, Toyota - developing a system that would allow for partial vehicle automation, infrastructure-to-vehicle, and vehicle-to-vehicle systems to get as close to L5 automation seems the easiest. There is so much fewer variables to consider when it comes to developing those technologies *and* most states/countries invest heavily in the state of good repair for their highway networks.

I could see the launch of an AV system that works for commercial freight then expand to intercity passenger transport as well. Those Boltbuses could be quickly automated to drive most of the way from Bos to NYC.

 

[George_Apley]

I hear ya, George. And it may not happen by 2035 in BOSTON (although, I do think Boston is RAPIDLY changing - - maybe more so culturally/business-wise than any other city in US).

But, as I mentioned earlier, I believe it WILL happen by 2035 in places like Singapore/Tokyo/Portland, Oregon, etc. It will be out there and it will spread. And Boston isn't as old-timey and flat-footed as it used to be.

Authoritarian technocratic city-states like Singapore, sure. Cities in countries that have a more collectivist social more like Japan, maybe. I just don't see any metro in the USA being in a position to have CBD-AV zones in the foreseeable future.

 

[ra84970]

Only one edit, not just first, but all.
Transportation would become a utility. Used on demand without any ownership by the user.


Do people like cars or do they like the independence of getting from place to place on their own terms? Give them an option with as much flexibility without the costs and issues of ownership and they'll switch in a heart beat. We've already seen it happen on a smaller scale with zipcar and TNCs

Let's be real transportation hasn't been treated like a utility in this country for a long time.

And the analogy to Uber/Lyft or other TNCs is inapt. The cost of "ownership" of the uber/lyft vehicle is born by the driver of your uber/lyft vehicle as an independent contractor. Uber/lyft, if they continue to frame themselves as a technology platform, will try to find themselves a contractual relationship that allows for the same level of flexibility, but, the only way that their model works is by relying on the ability to exploit independent contractor labor for bargain rates on labor and maintenance. In most cases, automation doesn't do away with labor and maintenance costs -- it shifts it up to higher-skill and more efficient labor. For automated vehicle fleets it moves from a large number of vehicle operators to smaller, but higher-skill (and wage/salary) workers who maintain AV vehicles, roadside assistance, and emergency remote operator teams who can handle drive-by-video systems.

Timeline is one issue for the vehicle systems, but, operations and maintenance of a fleet of vehicles won't be going away soon. Just imagine if Uber/Lfyt had to contract with Hertz for the provision of AV vehicles in good condition -- even at corporate rates, I'm not sure that they'res a "there" there for economic efficiencies that won't be overwhelmed by increased skilled labor costs.

 

[JeffDowntown]

Let's be real transportation hasn't been treated like a utility in this country for a long time.

And the analogy to Uber/Lyft or other TNCs is inapt. The cost of "ownership" of the uber/lyft vehicle is born by the driver of your uber/lyft vehicle as an independent contractor. Uber/lyft, if they continue to frame themselves as a technology platform, will try to find themselves a contractual relationship that allows for the same level of flexibility, but, the only way that their model works is by relying on the ability to exploit independent contractor labor for bargain rates on labor and maintenance. In most cases, automation doesn't do away with labor and maintenance costs -- it shifts it up to higher-skill and more efficient labor. For automated vehicle fleets it moves from a large number of vehicle operators to smaller, but higher-skill (and wage/salary) workers who maintain AV vehicles, roadside assistance, and emergency remote operator teams who can handle drive-by-video systems.

Timeline is one issue for the vehicle systems, but, operations and maintenance of a fleet of vehicles won't be going away soon. Just imagine if Uber/Lfyt had to contract with Hertz for the provision of AV vehicles in good condition -- even at corporate rates, I'm not sure that they'res a "there" there for economic efficiencies that won't be overwhelmed by increased skilled labor costs.

This is actually a huge potential disruption of the TNC model.

While Uber/Lyft make a big deal about their future being AV, in fact their core business model requires them to NOT own the vehicles. And there really isn't a logical owner of Uber/Lyft AV fleets who wouldn't be in the position to just operate the fleet themselves. The AV transition creates a huge disconnect with Uber/Lyft's financial models that will be very challenging for them to cross.

In many ways it would be much more logical/seamless for Zipcar to deploy AVs rather than Uber/Lyft. Much less business model disconnect.

 

[ra84970]

This is actually a huge potential disruption of the TNC model.

While Uber/Lyft make a big deal about their future being AV, in fact their core business model requires them to NOT own the vehicles. And there really isn't a logical owner of Uber/Lyft AV fleets who wouldn't be in the position to just operate the fleet themselves. The AV transition creates a huge disconnect with Uber/Lyft's financial models that will be very challenging for them to cross.

In many ways it would be much more logical/seamless for Zipcar to deploy AVs rather than Uber/Lyft. Much less business model disconnect.

I had been thinking about the fleet owner being in the position to take over uber and lyft's ridematching platform. I don't think you'd have the cachet of Uber/Lyft's name, however, I could see like a Hertz AV division, or GM's "Maven"-like service to be able to step-in and deliver rides very quickly.

 

[shmessy]

Can Self-Driving Cars Stop the Urban Mobility Meltdown?

Autonomous vehicles will be a major option in future metro transportation, but their optimal role in the mobile mix varies across five city archetypes.

www.bcg.com

".......Countries and cities that innovate to solve these issues are likely to become the leaders in AVs. Cities that introduce pilot projects—so they can solve potential teething problems and gain public support early on—will improve their chances of experiencing a smooth transition to AVs, according to most of our experts. Amsterdam, Singapore, and Boston have all taken this advice and are testing autonomous buses, drafting regulations and multimodal transportation strategies, and opening public roads to vehicle testing......."

 

[Balerion]

Q: Can Self-Driving Cars Stop the Urban Mobility Meltdown?

A: No.

 

[JeffDowntown]

Q: Can Self-Driving Cars Stop the Urban Mobility Meltdown?

A: No.

Unless you can break the pattern of rush hour volumes of SOV, then self-driving vehicles have zero impact on the urban mobility meltdown. Self-driven SOV road congestion is still road congestion.

 

[shmessy]

Unless you can break the pattern of rush hour volumes of SOV, then self-driving vehicles have zero impact on the urban mobility meltdown. Self-driven SOV road congestion is still road congestion.

"Zero impact"? Really??

Cellphone app order-on-demand pod service which will eliminate on-street parking, parking garages, individual driving techniques, etc. will have "zero impact"? Interesting........

Just think of the benefit of 100% street utilization for flow, instead of the current 50-70% utilization due to on-street parking. Given the effects throughout the system from the current 30-50% area utilization of the windy streets downtown and you can imagine the downstream benefits.

Stents that clear arteries make stronger, more effective hearts. Individually owned and cared-for vehicles that clog the sides of the streets are the plaque.

I'm not saying there won't be times of slower mobility with 2+ million living in Boston, but "zero impact" is a clearly hyperbole.