Uber Enters Race to Perfect Driverless Vehicle Technology

[TallManNY]

I was thinking that, but then they need an infrastructure for gas stations/charging stations with pumps/plugs that can automatically insert itself into the car. Just an overall complicated scenario!

Affordable, fully electric cars are closer than fully autonomous cars. So expect this fleet, if it ever exists, to be electric. That means space to park somewhere for a significant period of time to recharge. But the cars would be programmed to space this out.

The thing is, the rental car businesses already have space to house cars and refuel them, plus an infrastructure of employees to clean the cars and check them before delivering them back to customers. It seems like Uber having an App is not that much of an advantage compared to all the stuff that rental cars already have in place.
[doublepost=1463830554][/doublepost]

 

[jlc1978]

Affordable, fully electric cars are closer than fully autonomous cars. So expect this fleet, if it ever exists, to be electric. That means space to park somewhere for a significant period of time to recharge. But the cars would be programmed to space this out.

The thing is, the rental car businesses already have space to house cars and refuel them, plus an infrastructure of employees to clean the cars and check them before delivering them back to customers. It seems like Uber having an App is not that much of an advantage compared to all the stuff that rental cars already have in place.

Self driving cars would make Uber more like a dispatchable ZipCar than a taxi. Uber could negotiate parking spaces , and if the fleet is electric build charging stations, to keep cars available to respond to requests. If it sees high demand in an area it can move cars from a low demand area there and park them until called. With no drivers to worry about keeping busy or having turn down far away fares Uber can adjust car density to meet demand. Surge pricing, in theory, be less since they do not have to convince drivers to become available so the only time it should get real high is if they have almost no available cars.

Uber could also use the vast amounts of data they collect to predict demand and adjust cars based on that. For example, the Uber app could report locations of all the installed apps to determine the potential customer base in any area and combine that with what it knows about buying habits to gauge demand. The app could even monitor the use characteristics of the phone to decide if someone appears drunk and offer a ride. It could use the phones movements or even just realize you've been in a bar for 4 hours and assume you've been drinking. Scary, huh?

 

[TallManNY]

It's sort of like the FedEX Home Delivery model. You franchise out the name and the franchisee pays for the equipment, maintenance, etc. while you provide the product to them. In FedEX' case it's packages to deliver, in Uber's it's people (and packages I guess if they wanted). Uber would provide the back office and riders which means the would not need nearly as much capital as they would if they owned everything. Plus they have no money at risk if the service fails to generate enough revenue to cover the capital and operational costs.
In Uber's case, if they develop and build the cars they can make money selling or leasing them to their franchises, if they go that route, or to independent contractors.

Great point. I've been stuck on how Uber's business model could include that much capital deployment to expand world wide and buy that many cars and maintain them. However if the Franchisee buys and maintains the car (and also uses it as their own personal vehicle when they need it) then that is the capital source.

These cars really will be transformative. Another thing they will disrupt will be housing costs. Right now housing costs are based, in general, on location and proximity, in particular proximity to jobs. But once transportation is easier and cheaper, the value of being 10 minutes closer on the turnpike will go down. That will disrupt housing prices.

 

[jlc1978]

These cars really will be transformative. Another thing they will disrupt will be housing costs. Right now housing costs are based, in general, on location and proximity, in particular proximity to jobs. But once transportation is easier and cheaper, the value of being 10 minutes closer on the turnpike will go down. That will disrupt housing prices.

I'm not sure about that unless the transportation system evolves to not only make it easier but also to take less time. The value in living near a turnpike or subway stop is it not only makes it easier to get to work but also faster. Self driving cars have the ability to do that through the use of a control system similar to the current air traffic control system.

For example, when a light turns green all the cars could start to go at once, unlike now where essentially one car starts moving and the next one in line waits a few seconds to let the one in front move away before moving. You could do away with timed lights was well since the system would flow cars through the intersection safely by controlling each cars timing, requiring it to speed up or slow down to stay in the flow. The problem with that is you can't mix self driving and human operators in the same traffic pattern since humans are far too unpredictable and would cause all kinds of problems.

It could however, be used in city centers where non-self driving cars would be banned; essentially creating a people mover system that could link with other public transit and parking spaces for driven cars. The advantage of such system is the ability to dispatch cars based on demand and not be restricted to a fixed route like a bus or subway.

In some ways it could be like DC's slug lines where riders que at signs giving destinations and get picked up by strangers; in this case the public transportation stops / parking lots could have signs for locations within the service area and riders would que there and be picked up by larger vehicle than a typical car.

 

[theluggage]

Number of people who have driven for Uber: 162,000
Number of people who planned to work for Uber for life: 0

You forgot:
Number of Uber executives who plan to stay in the taxi business for life: 0.
[doublepost=1463833915][/doublepost]

Sure why not. I need a flight to Arizona. Oh hey there's a plane sitting over there, I'll jump in. The guy says he knows how to fly this thing.

As long as he's got a pilot's license then its probably safer getting into a plane with him than it is getting into a car with someone who merely has a driving license... and you'd probably be subsidising his expensive flying hobby rather than feeding his kids, so he'd probably have a pretty good attitude.

Of course, you'd be talking about a cheaper alternative to chartering a private plane from an existing business rather than something that was going to compete with cattle-class airline tickets.

Of course, the regulatory hurdles would probably be huge, and maybe Ubers usual "ignore the rules and hope they go away" approach might not work in aviation...

 

[CFreymarc]

A decade ago, the sheer prospect of driverless cars becoming mainstream was a fantasy. A decade from now, we could be looking back and wondering how we ever did without.

A decade from now the self driving car will be dead with consumer liability lawsuits killing it off. I'm betting against it.

 

[ateslik]

This was an awesome post and exactly what I was thinking. I'm all for progress but I'm not for turning the world into 100% robots.

but what about 70%? Where do you draw the line?

I have a friend who got a roomba. Now when he sees a person vacuuming he gets indignant, "why are they doing that?!?". That's the future right there.

 

[Tycho24]

Yes. Because that's how they insulate themselves from liability in their business model.

Their app and servers allow people to call drivers to pick them up, while the car's owner pays for the vehicle, and it's mechanical upkeep, cleaning, fuel expenses, and If anything goes wrong, they just say "Well, it's not our car." And you're left to make claims against the car owner's personal insurance.

...But then you find out their insurance company has a clause stating that their (consumer vehicle) policy is not in effect when the car's owner is engaged in "a paid 'ride sharing' or 'ride hailing' service" (i.e. Uber). And now you're SOL because the car was not really insured.

I don't think this is being remotely thought through by those that think this way...
Consider: being an Uber driver requires ZERO dollars. You need a recent car & no criminal record; then you drive around, do your normal day & get an alert when you're needed.
That is the key to their current success.
They built an app & have very few other costs.
But, in you guys' new idea- all their drivers are going to have a 2nd vehicle that is a brand new driverless model & they'll just let it drive around all day serving Uber.
Ummm... lol, wut??
Where are these drivers getting all this money to buy brand new cars??

 

[TallManNY]

I'm not sure about that unless the transportation system evolves to not only make it easier but also to take less time. The value in living near a turnpike or subway stop is it not only makes it easier to get to work but also faster. Self driving cars have the ability to do that through the use of a control system similar to the current air traffic control system.

For example, when a light turns green all the cars could start to go at once, unlike now where essentially one car starts moving and the next one in line waits a few seconds to let the one in front move away before moving. You could do away with timed lights was well since the system would flow cars through the intersection safely by controlling each cars timing, requiring it to speed up or slow down to stay in the flow. The problem with that is you can't mix self driving and human operators in the same traffic pattern since humans are far too unpredictable and would cause all kinds of problems.

It could however, be used in city centers where non-self driving cars would be banned; essentially creating a people mover system that could link with other public transit and parking spaces for driven cars. The advantage of such system is the ability to dispatch cars based on demand and not be restricted to a fixed route like a bus or subway.

In some ways it could be like DC's slug lines where riders que at signs giving destinations and get picked up by strangers; in this case the public transportation stops / parking lots could have signs for locations within the service area and riders would que there and be picked up by larger vehicle than a typical car.

Being closer, and therefor faster, will always be nice. But the "cost" to driving another 15 minute goes down a lot if you are just sitting in your comfy car watching TV during the drive. And as you say, the whole traffic pattern changes when there are auto cars. The dramatic decrease in accidents will make traffic flow faster. Even the decrease in just bad driving will make the traffic flow faster. Many traffic jams start not with an accident, but just a close call that slows down traffic for a bit which then creates a decrease in traffic flow which builds on itself.

There will even be an option for folks to live far away and get a nap in during their drive. Seats in cars might be designed to fully recline so that you can do an overnight trip that you sleep during. That will seriously disrupt the airline industry. Absolutely no point to many flights if you can take your car at a cost of about $30 in electricity and the drive is really not much different than a quiet night at home in front of the TV.

 

[freepomme]

Sure why not. I need a flight to Arizona. Oh hey there's a plane sitting over there, I'll jump in. The guy says he knows how to fly this thing.

Hey, this isn't Arizona.... And why are we heading straight for..... Oh $&@"& hold on....

It's too late!

 

[xtshabi]

The future of transportation keeps getting more interesting by the minute.

 

[TallManNY]

I don't think this is being remotely thought through by those that think this way...
Consider: being an Uber driver requires ZERO dollars. You need a recent car & no criminal record; then you drive around, do your normal day & get an alert when you're needed.
That is the key to their current success.
They built an app & have very few other costs.
But, in you guys' new idea- all their drivers are going to have a 2nd vehicle that is a brand new driverless model & they'll just let it drive around all day serving Uber.
Ummm... lol, wut??
Where are these drivers getting all this money to buy brand new cars??

Yep. The genius to Uber was that it took advantage of a bunch of people who already had cars but were either under employed or even unemployed. Some were already working for car service companies, but many were just folks with a car who were only finding some part time work. So they've added in 20 hours of Uber driving into their schedule. And it cost Uber nearly nothing to get a piece of the action from that extra driver. That is the genius. Though it should be noted that Uber still struggles to make a profit.

"Overall, Uber remains massively unprofitable as of the data shown in these documents. GAAP losses (net revenue minus cost of revenue, operating expenses and other costs) totaled $671.4 million in 2014. Those losses expanded to $987.2 million in the first half of 2015."

In this new model, someone has to buy a whole heck of a lot of driverless cars. Now maybe it is Uber. Or maybe it is a franchise model. I don't know. But it isn't the same thing as benefiting off of excess car and driver capacity.

Also Uber is going to lose some customers. If I want to go out for a night on the town (or even a drink and dinner), I can't take my own car because I don't drink and drive. But if I have my own driverless car, why would I call an Uber? I have my car drive me to dinner and then drive me back home. Cost is just a bit of electricity (which, in a few years, I likely generated myself from solar panels on my roof).

 

[CarlJ]

Am I the only one who glanced at the picture and saw a car with its roof covered with sorta weaponized-looking speakers?

 

[SeaFox]

But, in you guys' new idea- all their drivers are going to have a 2nd vehicle that is a brand new driverless model & they'll just let it drive around all day serving Uber.
Ummm... lol, wut??
Where are these drivers getting all this money to buy brand new cars??

Who said it had to be a second vehicle? The only people driving all the time are (funny enough) cab drivers. Everyone else goes from place to place when they need, but their vehicles spend large amounts of time parked and doing nothing. If the car is self-driving, you could go to you normal 9-5 job and just tell your car to go do some Uber hours while you're at the office.

Also, I'm not sure what your imagination of an "Uber car" is, but they aren't old clunkers poor people are driving as makeshift taxies to make ends meet. Lots of them are recent (likely still not paid off) vehicles. I'm even aware of a Tesla owner who does Uber driving. Why would someone who makes the kind of money to buy a Tesla (someone who could buy a second car for Uber use as you ask) spend his free time driving other people around?

The business is going to shift to franchise fleets on the near term once self-driving cars are feasible -- a lot of that will be so Uber can get out of it's current labor-law related troubles, but once the self-driving car becomes a more common thing, there's no reason you wont see individuals get into the game, too.

 

[Brien]

I'm sure their drivers love to hear this. Could you imagine working for a company that was actively working towards a goal of getting rid of about 99% of their workforce? I'm all for progress, but that would make me feel really weird.

According to this quote from the Washington Post:



That's going to be a lot of people to lay off several years down the road, and that's only the tip of the iceberg. According to the website All Trucking:



So that's a lot more people who will be eventually replaced. See this story about semi-autonomous trucks driving across Europe to get a glimpse of where things are headed. Then there are other jobs to consider: Postal and package delivery drivers, bus drivers, chauffers, maybe even tow truck drivers and more. Then you have tertiary industries, such as automotive insurance, which could eventually no longer be needed or at least have a very reduced role. That means adjusters, agents, managers, call centers and more jobs would be no longer needed, or have a severe reduction in number of positions. The reduction in accidents would also mean that auto mechanics would have a reduced role in society, especially for collision repair. This also means that fewer auto parts would need to be made, which impacts other industries, and could make auto repair in general more costly. You would also need fewer patrol officers, and cities would get less income from speeding tickets, parking tickets, and more—which is, perhaps surprising to some, a big part of many city budgets.

And I'm not even touching on the job impacts of a move from oil to electric.

To reiterate, I'm all for progress, and am in no way saying that these changes are bad. I'm all for a more efficient system that makes life easier and is better for the environment. But I just hope that organizational think tanks and governing bodies are acting now to research the impacts that this will have on our economy and the employment outlook, because this is going to be a tremendous period of transformation affecting many fundamental aspects of our society. And yet again, I'm not even touching on the privacy aspects revolving around government surveillance, the law aspects for whose fault an unlikely accident would be, and more. That's a whole other can of worms. It's really incredible how much change this will bring. It's an exciting time to be alive, but change can be difficult, especially when it has the potential to be this disruptive. But I'm hopeful our society will eventually benefit greatly from these improvements. The potential to greatly empower the elderly, people with disabilities and more with easy access to transportation will be fantastic, as well as the reduction in automobile deaths and pollution from energy efficient autonomous driving.

This is an issue that extends beyond autonomous cars.

One of the big issues of the century will be figuring out how society will operate in a world where labor (across the board) is obsolete.

 

[mw360]

Ummm... lol, wut??
Where are these drivers getting all this money to buy brand new cars??

Here?

https://get.uber.com/cl/financing/

"Thanks to Uber I am driving a new Toyota Avalon Hybrid. This has allowed me to drive more and save half the money I used to spend on gas in my previous car. I don't have to worry about the payments since Uber deducts the payments out of my weekly earnings."

This is what it may say in future...

"Thanks to Uber I own a new Uber Self Drive. This has allowed me to make extra money on top of what I earn driving in my Toyota Avalon Hybrid. It's also out making money while I'm at home with my family or even when I'm asleep. I don't have to worry about payments since Uber deducts the payments out of my weekly earnings."

 

[friedmud]

But how would Uber's automatic-driving cars fuel up/charge up? Either they have to pass that responsibility onto the customer, or have the car drive back to Uber employees/volunteers to fill up in between rides.

 

[ChrisA]

Today the world wide death toll for human driven cars is about one million. Yes a full MILLION people are killed in cars every year world wide. (India, China and Africa add more than their share to this total) It is as if a small city was being blasted off the face of the earth every year.

So to be "safe", driverless cars only need to kill LESS then one million people. If they caused the death of "only" 100,000 people we'd be 10 time better off then we are today.

Human drivers have set the bar VERY LOW.

(Even if you are very US centric, human drivers kill 20,000 Americans every year. If driverless cars killed 2,000 per year we'd be 10 times better off. US drivers have set a very low bar)

So, these cars need not be perfect, just better than humans who are by and large quite poor at controlling cars
[doublepost=1463942305][/doublepost]When all cars are driverless how will this effect the design of cities?

I think the #1 thing will be the elimination of parking lots. Why does the Uber car need to park? It will drop you off at the curb (or taxi stand lane) and then zoom off to get the next ride. If you are rich and own your own driverless car it will still drop you at the curb and then drive off to who cares where and charge itself. The end effect is no more huge lots of asphalt filled with parked cars. These lots tend to be larger then the retail store. Take these huge lots out of a city and everything can be closer together or maybe more park space. But in any case parking lots are some of the latest things we have in cities and they will no be needed. it is always a change when you remove the biggest thing.

The next generation will likely be even more fat and out of shape than we are so they will appreciate not having to walk all the way across some parking lot to get to the store, being dropped at the curb will make use even more lazy. and YES every retailer will demand curb space for drop off and pickup or else the lazy customers will not shop in his store. I think this transition from parking lots to curb space will have a big impact in how cities look. Certainly it will kill the current design of shopping malls which have nearly zero curb space and huge lots. The ratio if curb space to lot area will change

 

[pat500000]

I don't care what anyone says...I will not put my life with driverless car.
Who knows....these psycho man-killing driverless cars would probably run us over if we piss the off.v
LOL.

 

[ChrisA]

Did you read

I don't care what anyone says...I will not put my life with driverless car.
Who knows....these psycho man-killing driverless cars would probably run us over if we piss the off.v
LOL.

Did you read where I wrote the human driven cars, world wide, currently kill over ONE MILLION people every year. Already it is so bad that I'm amazed that people actually go near streets.

Yes those "psycho man-killing driverless cars" will likely kill a few people but if they kill less then one million a year it will be better than the current situation.

 

[pat500000]

Did you read


Did you read where I wrote the human driven cars, world wide, currently kill over ONE MILLION people every year. Already it is so bad that I'm amazed that people actually go near streets.

Yes those "psycho man-killing driverless cars" will likely kill a few people but if they kill less then one million a year it will be better than the current situation.

Killing nobody is LOT better than killing one.

 

[ChrisA]

I find this surreal. How do they deal with things like flat tires and drive ways at houses? I mean, when the driverless pizza delivery arrives, will I need to walk 30 feet in the snow to get the pizza? ...

You are asking the question "how will driverless cars operate in a city that is designed for human drivers?" This will be a problem at first but cities get re-designed as new buildings are built. For example the shopping mall did not exist until after there was tract housing in suburbia and cheap gas. It was a reaction to those conditions.

When driverless uber cars are common who will need or want a drive way? Even if you own the car you'd tell the car to go find a place in the charging hive that is a mile down the street. And being uber-lazy the car will pick you up on your front porch NOT on the driveway. Heck it might even dock with your house and you never go outside.

Flat tires are easy. The car calls a service who sends a human mechanic or a tow truck. The tow truck might drive itself but with the mechanic as a passenger.

Maybe the square sliding door (like a mini van) opens 6 inches from your house's front door?
[doublepost=1463967692][/doublepost]

Killing nobody is LOT better than killing one.

This is an interesting fact of human perception. We feel safe in a car even knowing that thousands of people die in them every week. But some people would refuse to ride in a driverless car even if only one person a week was killed in them.

Fear is not based on reason. It is based on familiarity, we fear things we don't understand. This explains why some people feel safe in cars but not airplanes when obviously the airplane is more safe.
[doublepost=1463968334][/doublepost]

Affordable, fully electric cars are closer than fully autonomous cars. So expect this fleet, if it ever exists, to be electric. That means space to park somewhere for a significant period of time to recharge. But the cars would be programmed to space this out.

The thing is, the rental car businesses already have space to house cars and refuel them, plus an infrastructure of employees to clean the cars and check them before delivering them back to customers. It seems like Uber having an App is not that much of an advantage compared to all the stuff that rental cars already have in place.
[doublepost=1463830554][/doublepost]

That is a good observation. But a self driving uber just might put a car rental company out of business. Who needs to rent a car?

About washing, I thing the car can drive itself to the lowest priced can wash. There are many small ones around. If there is a need to clean these cars then places will open. They need not be large or centralized.

This is going to happen, when I don't know but I do know that there will be much money to be made by offering services to the companies that have these cars. My idea is to offer sensor installation at fixed locations enabling cars to in effect "see around a corner"
[doublepost=1463968904][/doublepost]

....
There will even be an option for folks to live far away and get a nap in during their drive. Seats in cars might be designed to fully recline so that you can do an overnight trip that you sleep during. That will seriously disrupt the airline industry. Absolutely no point to many flights if you can take your car at a cost of about $30 in electricity and the drive is really not much different than a quiet night at home in front of the TV.

Overnight drive? OK maybe. But what if your car can drive up a ramp to connect with a Hyperloop Car? Your sliding door opens and the door on the hyper-loop opens and the two are inches apart. You step over to the other seat and then travel at 1,000 miles per hour in a vacuum tube and then step out into another driverless car.

We should not think in terms of driverless cars operating on current roadways. All the current roads were designed with human ability in mind.

 

[Whiplash]

Here's what people do in taxis with human drivers: They puke, they pee, they poop. They drag mud and dog excrement and who knows what inside. They have sex with each other and sex with themselves. They cough and sneeze and leave bacteria everywhere. They leave uneaten food and wrappers and other trash. They do every drug on the planet and then some.

Taxi and uber drivers have to deal with this now. People are just plain filthy. Take away the human driver and you can bet this behavior will skyrocket. And you can add more examples of volatile shenanigans. On cam? A little spray paint for that. Stolen credit card? Let's break out the AK47's and party. Maybe set the car on fire, 'cuz who's going to stop you.

Yep. I can hardly wait.

 

[TallManNY]

Overnight drive? OK maybe. But what if your car can drive up a ramp to connect with a Hyperloop Car? Your sliding door opens and the door on the hyper-loop opens and the two are inches apart. You step over to the other seat and then travel at 1,000 miles per hour in a vacuum tube and then step out into another driverless car.

We should not think in terms of driverless cars operating on current roadways. All the current roads were designed with human ability in mind.

Sounds nice but I don't think they will never build a hyper loop. A hyper loop will face similar economic challenges that our current rail system faces. And having to compete with driverless, electric cars will make it even harder. Now going 1,000 miles per hour seems great, but how much will it cost? And how cheap does it have to be to give up the convenience of having your own car (with all its stuff in it) at your destination and all you pay for is electricity.

Driverless cars is going to put the passenger rail system completely out of business. And while that is going on, you are going to be hard pressed to make the case to either investors or government that someone should spend tens of billions on a hyper loop system because there are going to be tons of passengers willing to pay high enough ticket prices and those passengers are going be around for the necessary 20 to 30 years to make back the upfront investment.

And while the roads were designed for humans, it is basically too late to change them. At least in the U.S. we have little political will to do large infrastructure investments and we frankly haven't had it for 40+ years. Basically we might be able to get it together enough to keep our bridges from collapsing. But we aren't going to build out another transportation system. Driverless cars will just have to make their own place in what we've got.

 

[ChrisA]

Here's what people do in taxis with human drivers: They puke, they pee, they poop. They drag mud and dog excrement and who knows what inside. They have sex with each other and sex with themselves. They cough and sneeze and leave bacteria everywhere. They leave uneaten food and wrappers and other trash. They do every drug on the planet and then some.

Taxi and uber drivers have to deal with this now. People are just plain filthy. Take away the human driver and you can bet this behavior will skyrocket. And you can add more examples of volatile shenanigans. On cam? A little spray paint for that. Stolen credit card? Let's break out the AK47's and party. Maybe set the car on fire, 'cuz who's going to stop you.

Yep. I can hardly wait.

I saw the answer to this on a Sci Fi movie. The car locked its doors so the passenger can't get out then drives to the police station.