Wouldn't Baltimore be the first Waymo market that actually gets snow? I don't think they've cracked driving in a real Midwest/northeast winter.

They’re being tested in Philly right now too

Waymo is a sort of public transit. It's just an vastly more inefficient than any other form of public transit, but an order of magnitude more efficient than private passenger cars.

reminds me of an old joke:

"When I go I want to die in my sleep like my grandfather. Not screaming and afraid like the passengers in his car"

In a Canadian context, on a two lane highway, sometimes doing the absolutely safe/totally cautious speed in a moderate snowstorm will result in a very large collection of vehicles behind you, with angry drivers. In particular if the persons collecting behind you are some combination of not very risk averse, commute on the same road every day, and are very confident in themselves because they have dedicated winter purpose studded snow/ice tires on.

Even if you also have good winter tires on, if your level of "caution" could be best measured as normal to high, sometimes it's a judgment call on when you want to pull off to the shoulder for 45 seconds to let a bunch of vehicles behind you pass. I'm not sure this is something any automated driver has been configured for. Or just generally to deal with driving when the road condition could best be described as "two only partially visible ruts in the snow where the tires of previous vehicles have driven, with snow in the centre".

Same thing in somewhere with a climate like upper Michigan or in Maine.

I drove up there in the AM Thursday, Feb 18th, during the snowstorm, about an hour before they closed the pass for the rest of the day.

You couldn't see anything. As soon as there wasn't a car 20 yards in front of you, it was a complete whiteout. Ice built up on the wiper as quickly as you could possibly reach out of your window and clear it. Radar would probably be nice, but I don't think it'd be enough to keep driving. The cameras and lidar would be an absolute wreck.

I'm sure we'll get there eventually, but that is really the final frontier for AI driving I think. Waymos aren't even allowed to drive in a snowstorm right now. I suspect that you'll be dealing with Caltrans closing the pass for the rest of your life.

Its not always about speed, This winter I was on interstate 93 in a 4WD with winter tyres. I was doing 25-35mph because the roads weren't treated. I still spun out, like many others. The road was an ice rink.

Humans and Control System Models need feedback to operate, and worse still... when any input into the vehicle's controls produce zero results, you will spin out.

My concern with a model in these conditions is that it wouldn't recoginize the fact that other cars were in the ditch and that it should probably slow down

After skiing in Utah, I wonder why the driving conditions around Tahoe get so bad. In comparison, for most places around Salt Lake/Park City, you never need chains or 4-wheel drive.

The entire city shuts down and loses their mind with just a millimeter or two of snow here. Last time we got 0.25 of an inch there were ~9 accidents within a 2-mile span on the highway in the morning, and we just ended up shutting the highway down for the day.

I love Waymo in other cities, but it'd be especially helpful here during the 1 day every other year that we actually get any snow ... if we ever get snow here again.

I don't see why one of those second-order effects wouldn't be the death of car ownership, with everyone using a rideshare service instead. Hell, that's the business model for Waymo and almost everyone other than Tesla in the autonomous-vehicle industry. It just doesn't make sense to own your own vehicle, use it for ~2 hours/day, and have to worry about parking/storing/fueling/maintaining it when you could have a service do all of that for you. Plus self-driving cars fix several issues with human rideshares, eg. you can drive it out to the boonies without worrying about how it's going to get back; you don't need to worry about getting assaulted by the robot driver; when they wait for you you only need to pay the opportunity cost of another ride rather than the opportunity cost of the driver's time. It's feasible to take a Waymo out to a state park, though you wouldn't usually do that with an Uber.

The second-order effects of that could be pretty wild. If people stopped owning their own cars, we wouldn't need houses with garages and driveways. It'd favor dense development with loading zones rather than parking spaces. It'd also be a big boon for EV adoption since the cars are all owned by one corporate owner and all go home to a centralized depot to charge at night rather than needing to retrofit EV chargers onto everyone's living situation. (Indeed, Waymo runs an all-electric fleet.). There'd be a premium on very reliable powertrains, since the cars might easily put 60-70K miles/year on them instead of the 10-15K that is typical of passenger vehicles. I dunno why Waymo went with Jaguar instead of Toyota, but perhaps "EV" is the explanation. Cars would wear out in 3-5 years instead of lasting for 15-20, and so you'd always have the latest hardware and technology on the car.

All the money we spend on traffic enforcement would become pointless, with audits of the software becoming a more effective use of dollars instead. But that blows a hole in many small local PD's budgets, many of which use speeding and parking tickets to raise revenue. Municipalities would likely find themselves powerless at regulating Big Self-Driving Rideshares.

The third-order effects are interesting as well. Once all cars on the road are self-driving, why not have them draft each other and physically link up to improve power efficiency and safety? You might even call such an arrangement a "train", blurring the line between road and rail transportation. But then, if you've got docking and linkage mechanisms, why not put the boundary between the electronics & powertrain and the passenger compartment, like the Rivian "skateboard" platform? You could return to private ownership of the passenger compartment - where, after all, some people like to store all their junk - and then have the rideshare own only the means of locomotion. Then you could extend this to other forms of locomotion like elevators, airplanes and ferries, so that your passenger compartment could just drop down an elevator shoot, onto a waiting self-driving car, which links up with others to become a train, takes you to the airport where you're loaded onto a plane without ever having to board, and then your pod deplanes and a self-driving car takes you straight to your hotel, where you now have transportation to wherever you want to go.

The future looks an awful lot like intermodal containers for people.

Off-roading aspirations and 3rd row legroom (S1) seem to be major differentiators from Rivian.

As for autonomy, Waymos have LIDARs which at least provides more redundancy.

I see these as different design tradeoffs so no judgment implied.

I, independently, made almost exactly the same comment before seeing yours lol. I already do 20+ hour cross-country trips in my Y without a break to sleep, which is only possible because I'm not meaningfully fatigued driving. it's still technically supervised but I think that's beyond the point OP is making

Its the ultra independence mindset. I don't think trains work for the commenter you talked to.

I want to move on my schedule and convenience, I don't want to have to warp my day to day around someone else's departure schedule.

there are effectively no passenger trains in America and effectively no political will to expand them. Busses take multiple times as long for the same trip compared to a car. This doesnt even get into the anti-social behavior present on both. Given these facts, it is not wild at all to prefer cars (self driving or not) vs alternate transportation methods

Unfortunately, until something big happens in the US, autonomous vehicles will be more accessible to working class americans than good and reliable mass transit, especially outside of major population centers.

You're right, but in the US a government providing any sort of public service is an immediate target for the right (and an unfortunately significant portion of the "left"). We insist on paying more for less rather than ever allowing a poor person to benefit in a way they don't "deserve". So public transit hardly exists or is woefully inadequate in most places.

I appreciate what you're saying and am a big fan of long distance train and bus journeys myself and have done a lot of both, sleeping and not.

But one huge factor that you have to contend with is the randomness of the tragedy of the commons problem on public transport / shared transport. A train journey can be blissful to sleep on right until a loud group gets on and sits across from you and there's no seats available to move.

I think this is something that can't be overlooked, especially if you're talking about something like a short trip where if you don't sleep well en route, quite a large proportion of the trip time is going to be affected. Having a private vehicle where you can guarantee control of your environment is a really huge plus.

In the US it's often not the last mile, but the last 10 or 100 miles. I'm saying this as someone enjoying fantastic public transport in Budapest.

I agree with you, however in the US the “last mile” is often the “last 50 miles” when goal is outdoor recreation.

> At least 80% of what you’re describing would be satisfied by trains and buses. It’s wild that Americans are so obsessed with self-driving cars while ignoring public transit that solves most of the problems. It’s reliable, more efficient, better for the environment, and less stressful for you.

As an American, it's far easier to imagine autonomous robot driven road trips than it is to imagine a government that is competent enough to built passenger rail networks.

> At least 80% of what you’re describing would be satisfied by trains and buses. It’s wild that Americans are so obsessed with self-driving cars while ignoring public transit that solves most of the problems. It’s reliable, more efficient, better for the environment, and less stressful for you.

So, let's say you take public transport from SF to Yosemite/Los Angeles. Now, how do I cover the last mile (or even multiple points)? Take more public transport? Hitchhike?

The reason long-distance public transport works well in Europe is that there is good local public transport in both the source and the destination cities. When that does not exist, you are better off driving.

It would also be satisfied by magic flying carpets. Between flying carpets, functional public transport, and self-driving cars, only one of these three things is not utter fantasy in the near-ish future in the United States.

Connecting two Waymo geos with a train would be an interesting company idea. You could lease freight track the way Amtrak already does it in the American West but try to negotiate a contact more favorable than Amtrak's. You could try to work with Waymo to work on bundles.

Amtrak could do the same thing but because of how Amtrak is organized in not sure that it would be possible. Most of the current Waymo geos are not connected by Amtrak directly and require transfers.

It's a train or bus that is exclusively yours, goes exactly where you want it to go, when you want to go. Sounds objectively better than a train to me.

I am a huge proponent of increased public transit (I'm of the opinion that every city should have a massive congestion tax with large swaths only accessible on foot or by public transit), but trains and buses would be wildly inconvenient for what op is describing.

Trying to take something like a windsurf board on a train, and then having to navigate multiple train changes along with whatever other baggage you have makes it a non-starter.

The "last mile problem" you mention is unresolved when it comes to getting from the closest public transit stop to the actual destination (frequently in a park or even off road).

And finally, the final cost to the rider would be significantly higher, as sleeper trains are not cheap.

I think America could do quite well if it focused on public transit in and between densely populated areas. Fewer cars in cities could make for denser cities, which in turn could allow for even more public transit. But outside of population centers, America is much more spread out than Europe, meaning that trains are less economical, and often wouldn't get the ridership that would allow them to make sense.

People just do not understand how big and spread out the US is compared to other countries. "Last mile" dramatically underestimates how much heavy lifting the personal transportation part would need to do. More like "last 50 miles".

The European mind does not comprehend how big and sparsely populated the American West is. You can't even pitch a tent in most places in the Alps, and why would you, when you just stop at a hut that has a staff and you can get fed and sleep in bunks with 20 other people? Meanwhile I can drive to numerous places where there isn't a structure or even another person in a 20km radius. No one is going to run a train to a place like that.

I looked at taking the train from my town to Glacier National Park along with my bike. The route goes from Portland and Seattle to Chicago, and has a stop at south glacier.

Step 1, get to the local train station in my town. There are 6 trains daily between me and Portland. Also, amtrak on the cross country trains requires the bikes to be in a box, in storage cars.

So I gotta get a large bike box, and get myself, my bike, the box, and some tools to break it down to our local amtrak station. Then partially dissasemble the bike, and box it. (of course, our train station has room in it for 5-10 people, and most sit outside, uncovered, which is fun in spring.)

Then, get to the main Portland Train station, with my bike box, and backpack with my stuff and tools. Wait up to 9 hours for the hawaitha train. (its often many hours late, and only leaves once per day).

Load Bike in cargo car, and then board train late at night.

Wake up around 5am, (or later, if train is behind schedule) and disembark at Glacier, re-assemble my bike. Figure out how to get it, and the box (i'll need it for the return trip) to a hotel or AirBnB.

For the return trip, its about the same, 1 daily westbound train, that is usually hours late, then hope you get to portland before the last train for the day leaves for my town, or else find a place to stay with a bike, backpack, and bike box in the sketchy area around the trainstation...

Or, hop in a car with a bike rack, and drive 10 hours. Which is easier, and MUCH cheaper if I split the cost of gas with someone else. So 2 extra travel days back for vacation, and much less stress.

The DC rollout is mired in regulatory red tape and is most likely dead until the mayoral election goes through, and if the new mayor is anti-Waymo unlikely to go through in the near future.

Is that supposed to be good?

once we've refactored humans out of driving, the speed limits can go way up

Pick-up, drop off, and routing remain a challenge for Waymo. I hate having to walk a few minutes to get to the Waymo. Not that big a deal usually, but it became a problem when I was on crutches after spraining my ankle. Same for drop-offs, with the caveat that a human driver is going to see that I'm being dropped off in a bad neighborhood and not have me walk a couple blocks and is going to drop me off right outside my destination. Finally, routing. Waymo's take the weirdest routes sometimes. There was one trip I'd swear the Waymo Driver was the digital equivalent of drunk, the route it took was so convoluted. Which is kind of interesting. It means the system can reroute on the fly based on traffic conditions elsewhere and avoid getting jammed up. It's like when Google maps has you take a weird route to somewhere you're familiar, and then you look at traffic and there's an accident it's taking you around. Still a bad experience when a 10 minute ride turns into a 20 minute ride because the Waymo decides to go a weird way.

I report these issues in the app whenever I do take a Waymo, so hopefully they'll get better.

The one to ride is Zoox though. They have limited deployment but their vehicles have no steering wheel, it's like a gondola ride to your destination.

Depends on the region, I think. Lyft and Uber partner with them in certain cities, so you transparently are charged the same as a similar ride with a human driver. It's only a better experience than a human driver, though, in my view. No chance of yapping, more privacy, no chance of your driver being a psycho, cars are better maintained.

It's hard to measure "cheaper" as an end user consumer, the price you pay for the service, because it's very likely they're operating at a loss to gain market share and growth.

Exact same reason why Uber and Lyft were considerably cheaper than taxis in many big cities when they first launched (eg: Lyft in Seattle in 2013/2014), running at a loss, and the pricing has now incrementally grown to become the same as, or even more expensive than traditional meter taxis in some places.

Waymo inflates their prices to be above that of Uber/Lyft because they don't have enough vehicles to meet demand. But their operating costs / mile are lower than that of Uber/Lyft. I'd estimate their internal cost per mile is approx. half that of Uber/Lyft. They pocket the rest because they need to recoup decades of expensive R&D.

There is also no reason to compete with Uber/Lyft on price because they are just leaving money on the table. When Waymo first launched, we saw them try to undercut (Waymo was about 20% cheaper than Uber/Lyft) but now it's about 20% more expensive. People are willing to pay extra for Waymo, so why would they charge less?

The margin on each Waymo ride is currently very, very high. I don't expect Waymo to cut prices until real competition arrives.

During peak hours Waymo is more expensive than standard uber/lyft - I don't pay attention to black/premium pricing. Off-peak the price can be comparable. I mainly check because my wife prefers it.

IIRC, in SF they're slightly more expensive before tip, but having ridden in them in SF, LA, and AZ I've always felt they were cheaper. Over the long run, they will probably end up being cheaper from the wholesale perspective since eventually the parts and technology cost will come down with time and scale while human wages will continue to rise.

That said, it doesn't really matter if they're cheaper as long as they're comparable.

The cars are newer and nicer (for now), they're almost always cleaner since they can rotating one car out for cleaning doesn't mean the driver is losing earnings, they're better drivers than the average ride-share driver, you don't feel the need to tip, and I've multiple of my friends who are women call out that they feel safer in them because there's no risk of the driver being creepy (or worse).

I don't think Waymo is trying to win on price right now. I think as long as they just stay somewhat competitive on that front the other benefits will continue to draw in customers.

If a waymo costs $200k (car+sensors+install labor) and drives 200k miles, then amortizing up-front costs alone are about $1/mile. We don't really know what the TCO of a waymo is, and it's possible it could go down with economies of scale. Rideshare drivers can get paid $1-2/mile although it varies a lot.

Yes exactly. I loved that when i opened up the app to go some where the first thing I saw before I even put in where I'm going, was how long until ill be picked up. the cars don't care about the trips, not trying to tip/ride max.

The amount of trips I would suddenly be interested in taking would skyrocket.

Same here, as someone who doesn't drive much, and is generally a "vulnerable road user". I've seen Waymos drive. When they screw up, it's by stopping dead under an abundance of caution. They never speed. They can spot a ped or cyclist from blocks away. Every time I take an Uber home, the driver is guaranteed to drive 40+ on the 20mph road in front of my house while blasting through crosswalks with people waiting to cross. The data is not really in yet (still not enough miles to really say if they are safer), but they pass the eye test.

The rain will be a real test though!

> The problem is that multiple red lights were blinking in view of the waymos not in sequence with each other, so the waymos interpreted it as a alternating-blinking red railroad signal crossing, and all of them refused to proceed, even when it was their "turn" in a 4-way-stop arrangement.

What's the hot fix for this? Are they just stuck until a tech can physically go out and reset and move them? Or can someone in a office somewhere remotely get alerted, look at the video feed/data, and override it with instruction on how to proceed?

Silly stuff like this happens all the time even with human drivers, I feel like the important piece when hearing that the technology encountered an issue is how long did it take to resolve?

This is a false blanket statement. Portland has very short (walkable) blocks, many one way streets, and it is true that most often than not cars actively stop for pedestrians to cross the street

I've personally reported a taxi driver for parking in a bike lane, and I hope he lost his cab license for it because it was really egregious. PBOT actually asked me for official testimony.

How would you know what diseases the previous passenger had?

The solution is to get MORE people onto the trains, not fewer.