Commercial Driverless Taxis Have Arrived

-edited

After more than a decade, Waymo's driverless ride-hailing service is open to customers. For now, it's free.

TechCrunch discusses Hailing a Waymo Driverless Taxi.

"Congrats! This car is all yours, with no one up front,” the pop-up notification from the Waymo app reads. “This ride will be different. With no one else in the car, Waymo will do all the driving. Enjoy this free ride on us!”

It marks the beginning of a driverless ride-hailing service that is now being used by members of its early rider program and eventually the public.

Limitations

  • The company’s driverless rides are currently free and only taking place in a geofenced area that includes parts of Chandler, Mesa and Tempe.
  • Even Waymo vehicles with safety drivers don’t yet take riders to one of the most popular ride-hailing destinations: the airport.
  • The everyday interactions between a passenger and an Uber or Lyft driver, such as conversations about pick-up and drop-offs as well as sudden changes in plans, become more complex when the driver is a computer. It’s an area that Waymo’s user experience research (UXR) team admits it is still figuring out.
  • Computers and sensors may already be better than humans at specific driving capabilities, like staying in lanes or avoiding obstacles (especially over long periods of time), but they lack the human flexibility and adaptability needed to be a good mobility provider.

Pooh Poohs Coming

I expect many readers will pooh-pooh this story with the usual nonsense about weather, old men on roller skates veering into traffic, liability issues, theft, and other silliness.

The fact of the matter is this is 2019.

I expected trucks, not taxis would be far in front. And they likely still are.

City driving where riders might change their minds is far more complex that point-to-point trucking from interstate hub to hub. Taxi adoption is likely to be slower.

The primary thing holding up hub-to-hub trucking is national regulation. I expect to see that next year.

Once allowed, interstate truck driving will quickly adopt. Commercial taxis will have a longer adoption period.

Personal Anecdote

The limo service I typically take to the airport just got rid of all its limos and drivers. The drivers are now contractors and must own their own vehicles. Those who refused were fired.

With that change, the owner's business model morphed into taking a percentage of the fare for scheduling rides rather than owning any vehicles and hiring drivers.

Within five years or less, those jobs will be gone.

Driverless Has Arrived

Driverless is here. The only debate is how fast it ramps up.

I suggest that within a 2-3 years of federal regulation, the majority of truck on the interstates will be driverless.

Even faster would not be a surprise.

Mike "Mish" Shedlock

Comments (76)
No. 1-29
Tony_CA
Tony_CA

the vast majority of commercial trucks will not be driverless in 2-3 years. Maybe in 10-15 years. As a society, we are simply not ready.

Jdred
Jdred

An attorney friend of mine pointed out the massive revenue loss to cities and municipalities if driverless cars that don’t speed or break traffic laws in any way are the norm. He thinks they will be an impediment to adoption. I think by the time they wake up it will be too late...

Six000mileyear
Six000mileyear

Does the camera in the car check the rider against outstanding warrants and drive to the police station if there is a hit on the database?

Stuki
Stuki

"With that change, the owner's business model morphed into taking a percentage of the fare for scheduling rides rather than owning any vehicles and hiring drivers.

Within five years or less, those jobs will be gone."

Dude, you're an ex programmer. Which task do you reckon is most eligible for automation: 1) Driving to and from an airport where everyone is desperate, running late, and erratically breaking every rule in the traffic book; or 2) "scheduling rides.........." Whose jobs are in trouble, again...

dansilverman
dansilverman

Well the key term here is 'georeferenced' roads. The driverless revolution is not in the car or it's software but in the roads and how they will be built. BTW, would you trust your life with a company that sells ads to keep aflot.

Maximus_Minimus
Maximus_Minimus

I truly hope the navigation isn't done by Google navigator. Some of the routes with this software are on the adventurous side.

Realist
Realist

As always; every time a human driver does something stupid and endangers my life, I can’t wait for this technology to be adopted in the vast majority of all vehicles. My wife has developed significant eye problems, and can’t drive at night anymore. She would love for this tech to be commonplace.

But neither of us is as optimistic as Mish for its widespread availability and adoption. I agree with Tony. It will be at least 15 years (and more likely 20-30 years) before this is commonplace in the majority of vehicles.

One simple reason is that even if the tech worked perfectly today (which it does not), it would take 15 years of vehicle production to replace the existing fleet of human driven vehicles.

Add on to this the ongoing improvement in the tech that is still required, the infrastructure changes required, the legal changes required, etc and realistically, you are looking at close to three decades.

I would love to see it happen sooner. But I’m not an optimist. I’m a realist.

Greggg
Greggg

We are seeing driver-less cars increase over the last few months here in Michigan... in dealer's lots.

Snow_Dog
Snow_Dog

“Within five years or less, those jobs will be gone.”

What jobs WILL BE here? Or will we all have UBI as our subsistence?

The0
The0

I agree that the projections in the article are too optimistic. Not only for technical / regulatory reasons but also for a more practical reason: because rolling stock lifetime / turnaround time. The trucks purchased now will be in use for 10-ish years at least. But I agree that highway trucking is a better starting point than taxis. That problem is mostly solvable today - as opposed to urban driving which is many, many years away to be completely solved. And then a decade or two to globally penetrate the market.

We should also distinguish geofenced mobility services from full self-driving vehicles. First one is mainly a mapping problem and a simpler one too - we can expect a few demonstration programs to pop up the next 2-3 years in simpler locations like Chandler, AZ or closed communities like the way Voyage does it (I also see a lot of activities in airport and recreational area automation).

As someone who works on self-driving cars and attends several events on this topic, I can feel the frustration of the tech community about the false promises and lies some people said and the media's way of hyping things up without fact cheking or looking into technical details.

We're getting there but it's also clear it's not a single technology that will just "arrive". It will be deployed incrementally the next decade or two.

Mish
Mish

Editor

"I never said "vast majority of trucks" - I get POed when brutally misquoted - I said

" the majority of truck on the interstates will be driverless" and they will

wootendw
wootendw

"The company’s driverless rides are currently free and only taking place in a geofenced area that includes parts of Chandler, Mesa and Tempe."

My ex-wife has invited me back to Arizona (Chandler) for Thanksgiving so maybe I'll have a chance to give it a shot, assuming the demand isn't greater than the supply. If they're really driverless, my guess is that the Phoenix area will be virtually all driverless by 2033. Outlying areas, everywhere, will take much longer.

Aaaal
Aaaal

Mish, when politicians in podunk cities are willing to accept graft and in turn allow their city to be the 'playpen' to test these technologies, don't mistake this as automation going full blown. It is inevitable yes, however, not for long time. Until they can outlaw every single driver from being on the road, full automation will not happen except in 'cordoned off' areas.

Automated vehicles cannot share the road with callous, IDGAF for the rules, retard drivers whose numbers have exponentially skyrocketed in the past 3,4 years.

The only viable means of this becoming common is dedicated roads/lanes such as HOV/toll lanes being converted for this express purpose.

Quatloo
Quatloo

Driving on highways by computer does seem easier to do than driving in the cities. But the trucks still have to be able to go into the cities to load and unload. So even if 90% of their time is on the highways, they still have to work flawlessly in the cities before they will be allowed on the roads.

numike
numike

“But the worries about operatorless elevators were quite similar to the concerns we hear today about driverless cars.” ― Garry Kasparov, Deep Thinking: Where Machine Intelligence Ends and Human Creativity Begins

numike
numike

“Whoever perceives that robots and artificial intelligence are merely here to serve humanity, think again. With virtual domestic assistants and driverless cars just the latest in a growing list of applications, it is we humans who risk becoming dumbed down and ultimately subservient to machines.” ― Alex Morritt, Impromptu Scribe

Country Bob
Country Bob

Mises Institute -- a group out of Alabama that promotes Austrian economics and libertarian thinking (Mish's thinking too?) -- has an article showing how socialism has held Africa back, despite Africa having a bounty of natural resources (metal ores, crude oil, fertile soil, etc) and an abundant (and cheap) labor supply.

Colonial powers were kicked out decades ago. Billions in foreign aid has come in -- from the world bank, from foreign investors, and most recently from China's One Belt, One Road initiative...

...but due to socialism's ugly reality, Africa continues to suffer economically decade after decade after century.

KidHorn
KidHorn

I have a short drive to and from work. Maybe 10 highway miles. Even on this short stretch, driverless trucks would have a big problem. There's a lot of construction with old painted lines and new painted lines. Sometimes it's not clear where the lanes are. You have to kind of guess based on where the other cars are driving. And to make room for construction, the lanes were narrowed. Sometimes a semi can't even fit between the lines. If you want to pass, you have to wait until the lane on the other side is clear and punch it around the truck.

Arnstein
Arnstein

Once again, a commercial outfit begins a research project featuring new technology and a human in the driver's seat. This is nothing new. This is not progress.

The fact remains that the arrival of driverless vehicles cannot yet be predicted. There is no schedule, there is no plan.

PZoio
PZoio

Hello everyone! This is my first post, and hope not the last one :-) When 5G will be fully implemented, geofenced area limitation will be gone, since every car will be connected to light poles (or whatever) within city streets and with any other cars, driveless or not. 5G has huge potential due to the amount of data transmitted but mainly due to the very, very low latency.

Mish
Mish

Editor

“But the worries about operatorless elevators were quite similar to the concerns we hear today about driverless cars.” ― Garry Kasparov, Deep Thinking: Where Machine Intelligence Ends and Human Creativity Begins

Beautiful comment thanks numike

dansilverman
dansilverman

We were promised driverless taxis in San Francisco in 2016, 3 years ago. And then this happened.

"The 2017 Volvo SUV was traveling at roughly 40 miles an hour, and it did not appear that the car slowed down as it approached the woman, said Tempe sergeant Ronald Elcock."

Now, 10s of billions of dollars of investor money later plowed into Uber, and now up in smoke, we still don't have driverless taxis. Oh but wait, Waymo to the rescue.

Mish
Mish

Editor

"So even if 90% of their time is on the highways, they still have to work flawlessly in the cities before they will be allowed on the roads."

Wrong - Xil had an excellent comment on the "last mile" - a point I have made many times

Maximus_Minimus
Maximus_Minimus

I am sure the Waymos will have the soft object detector installed, unlike this driverless Volvo: https://www.youtube.com/watch?v=_47utWAoupo

vardaman
vardaman

Never gonna happen.

SleemoG
SleemoG

It is endlessly amusing to watch all the naysayers and Luddites be wrong day after day, year after year.

Herkie
Herkie

These cars are NOT driverless, they are computer driven, they are NOT autonomous, they are connected to the cloud, loss of signal will get you dead. The trillions spent on 5G and other technologies that this will require is the largest single misallocation of capital since the invention of war.

I will never get in one, even if that means living in a nation where they can't afford the technology. You are all going to regret this if you live 15-20 years because right about then they will admit that these monstrosities and human piloted vehicles will never be able to coexist, and after misallocating trillions they will simply ban human driving not the computer driven.

Until then these will be slowing traffic because they really cannot cope with human driving, and they never will, there are some things computers can do better, and there are some things they will never do half as well.

Driving a car is one of those things where you must have a human conscience.

themonosynaptic
themonosynaptic

Let's test out a possible scenario:

  1. A self-driving car that can handle 95% of driving conditions (let's assume that there are restrictions on use during blizzards or other intense weather or other situations) becomes available. The testing has proven that this vehicle is as good as a very alert, experienced driver and, unlike a human driver, can maintain that standard of driving 100% of the time. The car costs the equivalent of a regular car plus $50,000 (for the extra h/w and s/w).

  2. The car gets adopted in wealthy communities and is a status symbol. It gets traction in L.A., S.F., and the wealthier parts of Florida. Celebrities in their self-driving car post photos and videos on social media. They are a "thing".

  3. Despite some inevitable, and highly publicized, fatalities, the driving stats hold up and the cars are far safer than human operated vehicles.

  4. Status seeking individuals in every other community in the U.S. and Europe want to be the first in their sphere to have the cool car.

  5. The cars gather momentum, and "executive" limo companies start offering service in them - achieving excellent profit margins from the premium fees and driver savings.

  6. They start to become boring because they are just commonplace - like the iPhone after the 3G model when many regular people had them.

  7. Ride sharing companies see the financial liabilities drop and start using them for their profit potential.

  8. Regular car sales drop precipitously as the second hand market is flooded with discarded newer model inventory. Regular cars are seen as a liability (think tube TVs about 10-15 years ago).

  9. Most households now have an "old" regular car they keep for unusual needs, and either a self driving car or a car service like Uber for most of their day-to-day needs.

There are only two time periods that are important:

a/ When we will get to step 1 - my best guess 2022 b/ How long from step 1 to step 9 - my best guess three years

Herkie
Herkie

Uber's Self-Driving Cars Have a Huge Flaw When It Comes to Jaywalkers

Uber shut off the emergency breaking system and the computer identified the dead woman as A) a motor vehicle B) a bicycle and C) finally as a person just one second before impact.

"The company's system is now able to handle scenarios such as jaywalking in which people or cyclists are not following road rules, she added, though human drivers may still need to intervene at times."

Right, you are paying ZERO attention to the road because the car is driven by cloud computing, a child runs out in front of the car, the reflex response time of a real human driver might save that kid or dog most of the time, but the response time of a human to intervene when they are not even paying attention to what is outside the car is never EVER going to be quick enough to stop in time if they have to intervene. And if you have to remain in position foot hovering over the pedals hands on the wheel and alert and looking for potential problems at all times then WHAT IS THE F&%$ point of a self driving car?