Page 1 of 4

Google developing competitor to Uber

Posted: 2015-02-04 12:33pm
by Borgholio
Using their self-driving robot cars, no less.

http://www.bloomberg.com/news/articles/ ... over-taxis

Uber faces an ever-growing cast of adversaries that includes dubious regulators, litigious drivers, hostile members of the press, and some well-funded rivals. But the most significant threat to the app-based transportation company may be much closer to home: one of its biggest investors, Google.

Google Ventures, the search giant's venture capital arm, invested $258 million in Uber in August 2013. It was Google Ventures' largest investment deal ever, and the company put more money into Uber's next funding round less than a year later. Back then, it was easy for observers to imagine Google teaming closely with Uber, or even one day acquiring it. David Drummond, Google’s chief legal officer and senior vice president of corporate development, joined the Uber board of directors in 2013 and has served on it ever since.

Now there are signs that the companies are more likely to be ferocious competitors than allies. Google is preparing to offer its own ride-hailing service, most likely in conjunction with its long-in-development driverless car project. Drummond has informed Uber's board of this possibility, according to a person close to the Uber board, and Uber executives have seen screenshots of what appears to be a Google ride-sharing app that is currently being used by Google employees. This person, who requested not to be named because the talks are private, said the Uber board is now weighing whether to ask Drummond to resign his position as an Uber board member.

Uber is also teaming up with Carnegie Mellon University for a research facility in Pittsburgh, Pa., to develop its own autonomous vehicle technology, the company announced on Monday. (The news was reported earlier by TechCrunch.)

Google has made no secret of its ambitions to revolutionize transportation with autonomous vehicles. Chief Executive Officer Larry Page is said to be personally fascinated by the challenge of making cities operate more efficiently. The company recently said the driverless car technology in development within its Google X research lab is from two to five years from being ready for widespread use. At the Detroit auto show last month, Chris Urmson, the Google executive in charge of the project, articulated one possible scenario in which autonomous vehicles are patrolling neighborhoods to pick up and drop off passengers. “We're thinking a lot about how in the long-term, this might become useful in people's lives, and there are a lot of ways we can imagine this going,” Urmson said in a conference call with reporters on Jan. 14. “One is in the direction of the shared vehicle. The technology would be such that you can call up the vehicle and tell it where to go and then have it take you there.”

Those comments, according to the person familiar with deliberations of the Uber's board, have left executives at Uber deeply concerned—for good reason. Google is a deep-pocketed, technically sophisticated competitor, and Uber’s dependence on the search giant goes far beyond capital. Uber’s smartphone applications for drivers and riders are based on Google Maps, which gives Google a fire hose of data about transportation patterns within cities. Uber would be crippled if it lost access to the industry-leading mapping application, and alternatives— such as AOL's MapQuest, Apple Maps, and a host of regional players—are widely seen as inferior.

Google’s entrance into the ride-sharing market would also leave Uber without a partner in the suddenly plausible future in which cars without steering wheels roam the streets. Uber will either have to develop the technology itself or form an alliance with a company that can if it wants to offer autonomous vehicles within its fleet. Mercedes, Audi, Tesla, and other carmakers have said they are developing driverless cars, though it's not clear that any is as advanced as Google's.

An Uber spokesperson declined to offer a comment for this article. A Google spokesperson also declined to comment, although the company issued a cryptic tweet.

Travis Kalanick, Uber’s CEO, has publicly discussed what he sees as the inevitability of autonomous taxis, saying they could offer cheaper rides and a true alternative to vehicle ownership. “The Uber experience is expensive because it’s not just the car but the other dude in the car,” he said at a technology conference in 2014, referring to the expense of paying human drivers. “When there’s no other dude in the car, the cost [of taking an Uber] gets cheaper than owning a vehicle.”

There's already an additional sign of a rift between the companies. Last week Google announced it would start presenting data from third party applications inside Google Now, a service that displays useful information prominently on the screen of Android smartphones. Google said it had struck deals to draw data from such apps as Pandora, AirBnb, Zillow, and the ride-sharing service Lyft. The company most obviously missing from that list? Google’s old and possibly former friend, Uber.

Re: Google developing competitor to Uber

Posted: 2015-02-04 12:42pm
by Broomstick
Assuming the driverless car is reliable, I personally would be more comfortable getting into such a "taxi" than into a car driven by a complete stranger.

Re: Google developing competitor to Uber

Posted: 2015-02-04 12:45pm
by Borgholio
Image

Re: Google developing competitor to Uber

Posted: 2015-02-04 08:39pm
by Raw Shark
I'm not worried until they build a version that can detect and remove vomit.

Re: Google developing competitor to Uber

Posted: 2015-02-04 08:43pm
by Borgholio
Raw Shark wrote:I'm not worried until they build a version that can detect and remove vomit.

Image

Re: Google developing competitor to Uber

Posted: 2015-02-04 11:06pm
by Simon_Jester
It's probably doable to build a linoleum box with a robot cleaner that can detect and remove urine, vomit, and so on from the box. Problem is, nobody wants to ride around in a taxi that has a linoleum box for a backseat.

It doesn't take many drunks hiring a robot taxi, puking in the back seat, staggering out, and leaving the next fare to go "OH MY GOD NO" to ruin the reputation of your robot taxi service. And how do you prevent that from happening, build a robot taxi that refuses to accept drunk passengers? That half defeats the purpose of the service.

Re: Google developing competitor to Uber

Posted: 2015-02-04 11:14pm
by Soontir C'boath
I would just install a detector and have it tell the car to go back to the depot for cleaning.

Re: Google developing competitor to Uber

Posted: 2015-02-04 11:46pm
by Alyrium Denryle
Broomstick wrote:Assuming the driverless car is reliable, I personally would be more comfortable getting into such a "taxi" than into a car driven by a complete stranger.

There are some issues. They cannot yet handle heavy rain or snow, and there are some problems differentiating small objects on road surfaces (trash vs rock etc), and they have issues with temporary traffic signals. However, with a human operator able to take manual control at any time, those should not be significant problems.

For a completely driverless taxi system... well.... they will need to coordinate with law enforcement etc with regard to temporary traffic signals (construction etc) and add a subroutine to route themselves around such obstacles, and they would be unable to operate in bad weather. For now. Google is working on those issues.

Other than the above caveats, they are safer than people.

Re: Google developing competitor to Uber

Posted: 2015-02-05 03:17am
by salm
Simon_Jester wrote: It doesn't take many drunks hiring a robot taxi, puking in the back seat, staggering out, and leaving the next fare to go "OH MY GOD NO" to ruin the reputation of your robot taxi service. And how do you prevent that from happening, build a robot taxi that refuses to accept drunk passengers? That half defeats the purpose of the service.
Yeah, kind of like nobody uses subways and trams. :)
Really, you´d just flag the taxi as dirty and the central would send a new one. It would probably be possible to sort out a lot of dirty cabs by remotly monitoring them and the help of sensors.
Puke doesn´t seem like an unsolvable problem.

Re: Google developing competitor to Uber

Posted: 2015-02-05 04:56am
by Jub
The dirty cab issue could be solved by having one or more cameras per cab watched by a human dispatcher. At the end of each trip the cab calls home and gets inspected to be sure it's in good shape to make the next run. If it is it goes right back out on the next call, if it's nasty it can come back to a maintenance bay and get cleaned up before hitting the streets again.

The larger issue will be ensuring people pay their fares, but even that could be solved in several ways. One such way might be that the fare is calculated up front and paid before the cab even starts rolling, this would probably lead to fares being calculated by distance and estimated arrival times rather than the current way things are done. You could also require a deposit up front that's calculated to be larger than the projected fare and then the difference is refunded when the trip is complete. In any case none of this is insurmountable once self driving cabs reach mature status.

Re: Google developing competitor to Uber

Posted: 2015-02-05 06:36am
by Purple
Jub wrote:The larger issue will be ensuring people pay their fares, but even that could be solved in several ways. One such way might be that the fare is calculated up front and paid before the cab even starts rolling, this would probably lead to fares being calculated by distance and estimated arrival times rather than the current way things are done. You could also require a deposit up front that's calculated to be larger than the projected fare and then the difference is refunded when the trip is complete. In any case none of this is insurmountable once self driving cabs reach mature status.
Why not just make it so that once he journey ends the doors stay locked until the correct fair has been deposited? Refusal to pay leads to an operator being called to look at the situation remotely (like your puke check, only for a different purpose). And the offender might eventually get a free ride to the local police station.

Re: Google developing competitor to Uber

Posted: 2015-02-05 06:44am
by Jub
Purple wrote:
Jub wrote:The larger issue will be ensuring people pay their fares, but even that could be solved in several ways. One such way might be that the fare is calculated up front and paid before the cab even starts rolling, this would probably lead to fares being calculated by distance and estimated arrival times rather than the current way things are done. You could also require a deposit up front that's calculated to be larger than the projected fare and then the difference is refunded when the trip is complete. In any case none of this is insurmountable once self driving cabs reach mature status.
Why not just make it so that once he journey ends the doors stay locked until the correct fair has been deposited? Refusal to pay leads to an operator being called to look at the situation remotely (like your puke check, only for a different purpose). And the offender might eventually get a free ride to the local police station.
You don't do that because it leads to people breaking your cab trying to get out and it could also lead to people taking legal action against the cab company. It's honestly not worth it and there's no reason that you can't just charge before the ride starts.

Re: Google developing competitor to Uber

Posted: 2015-02-05 07:07am
by Zixinus
The traditional option is blacklist: no other computer-taxi service will take that costumer who has not paid until he pays his due. This forces the use of human-driven taxis but those may be more expensive.

I'm sure some other legally binding way will be thought out. I doubt Taxi drivers have to beat their fare out of their costumers every day.

The other option is that if you want to use the service, you make and register an account with the company. For the costumer it is convenient because they just install an app on their phone and call a computer-taxi anytime they need to. But this would also allow a legally binding commitment to pay dues. It might even be possible that instead of paying as you ride, you put money on your account and thus all travel is pre-paid.

Re: Google developing competitor to Uber

Posted: 2015-02-05 07:14am
by Purple
@Jub

There is truth in that.

Re: Google developing competitor to Uber

Posted: 2015-02-05 09:08am
by Borgholio
Why not just make it so that once he journey ends the doors stay locked until the correct fair has been deposited?
I think Jub's idea of pre-charging based on distance is actually a more fare idea (heh) than current systems where they charge by time. It's a hard number vs a fluid number (distance vs time), easy to verify by a customer using Google Maps or Waze, and you can't accuse of someone taking the long way to rack up time if you see on a small LCD screen in the cab the exact route they're taking and see for yourself that it's pretty much the most direct route.

Re: Google developing competitor to Uber

Posted: 2015-02-05 09:20am
by Jub
Borgholio wrote:
Why not just make it so that once he journey ends the doors stay locked until the correct fair has been deposited?
I think Jub's idea of pre-charging based on distance is actually a more fare idea (heh) than current systems where they charge by time. It's a hard number vs a fluid number (distance vs time), easy to verify by a customer using Google Maps or Waze, and you can't accuse of someone taking the long way to rack up time if you see on a small LCD screen in the cab the exact route they're taking and see for yourself that it's pretty much the most direct route.
That was my thought exactly. The current method makes sense when you need to pay a cab driver for his services and a significant delay could cost him dearly in terms missed fares, but with a robot time becomes somewhat less relevant. Plus going to a flat rate cheaper fare will likely be needed to get people over the initial distrust of a car with no driver.

Re: Google developing competitor to Uber

Posted: 2015-02-05 09:24am
by Zixinus
Borgholio wrote:
I think Jub's idea of pre-charging based on distance is actually a more fare idea (heh) than current systems where they charge by time. It's a hard number vs a fluid number (distance vs time), easy to verify by a customer using Google Maps or Waze, and you can't accuse of someone taking the long way to rack up time if you see on a small LCD screen in the cab the exact route they're taking and see for yourself that it's pretty much the most direct route.
Yes, but in a city it may not work that simply. In a city you may not want the most direct route but the route that avoids enough traffic to get you there quickly. It may be that the fastest route is genuinely a longer route, plus there are things like traffic jams. Or even go trough a route with less stop-lights. This would probably involve very complex algorithms or even outright some level of AI to do efficiently.
This is not even considering things like actual time or if the computer-taxi has to change route during driving.

What the costumer really gains is more documentation of the route. If the programming is set out to cheat the costumer, it can still be done.

Re: Google developing competitor to Uber

Posted: 2015-02-05 09:30am
by Borgholio
What the costumer really gains is more documentation of the route.
A simple solution would be to have an LCD screen in the cab where it lists the routes like navigation apps do today. Let the customer choose if they want the shortest, fastest, or least expensive route.

Re: Google developing competitor to Uber

Posted: 2015-02-05 09:46am
by Lagmonster
Why not just make it so that once he journey ends the doors stay locked until the correct fair has been deposited?
Google could quite easily issue a form of Google Transit card to anyone who wanted to use their service. The car could be engineered to require swiping the card prior to driving anywhere, which would serve as a way to identify and bill the client. The client would then either receive a bill in the mail later, or could opt to be debited directly via credit card or automated bank payment. This would also allow Google to add surcharges to a bill where surveillance determined that a client had defaced, overly dirtied or damaged the vehicle somehow.

Re: Google developing competitor to Uber

Posted: 2015-02-05 02:18pm
by salm
It is Google. You can be sure that Google knows everything about you if you use their taxi. It will be trivial for them to get the money or send the police after you if you leave without paying.

Re: Google developing competitor to Uber

Posted: 2015-02-05 02:27pm
by andrewgpaul
Do taxis in your areas charge by time? The black cabs here (and in London, I think), charge mainly by distance, with an additional time-based component when the taxi is stationary.

Re: Google developing competitor to Uber

Posted: 2015-02-05 05:11pm
by Gandalf
If one isn't paying drivers, it might be simpler to calculate based on distance at the start of the journey. Then they can just get the passenger to preauthorise the fare when they get in. Upon arrival, the fare is charged.

Re: Google developing competitor to Uber

Posted: 2015-02-06 01:28am
by RogueIce
andrewgpaul wrote:Do taxis in your areas charge by time? The black cabs here (and in London, I think), charge mainly by distance, with an additional time-based component when the taxi is stationary.
The NYC Taxis are not the only ones, of course, but they are the most famous so I'll use them as an example.

http://www.nyc.gov/html/tlc/html/passen ... rate.shtml

And it would appear that they charge mostly by distance, only using time when the cab is stationary or in "slow traffic" whatever that means.

Re: Google developing competitor to Uber

Posted: 2015-02-06 01:50am
by Soontir C'boath
RogueIce wrote:
andrewgpaul wrote:Do taxis in your areas charge by time? The black cabs here (and in London, I think), charge mainly by distance, with an additional time-based component when the taxi is stationary.
The NYC Taxis are not the only ones, of course, but they are the most famous so I'll use them as an example.
http://www.nyc.gov/html/tlc/html/passen ... rate.shtml
And it would appear that they charge mostly by distance, only using time when the cab is stationary or in "slow traffic" whatever that means.
Basically means when the cab cannot do the speed limit.
Jub wrote:
Purple wrote:
Jub wrote:The larger issue will be ensuring people pay their fares, but even that could be solved in several ways. One such way might be that the fare is calculated up front and paid before the cab even starts rolling, this would probably lead to fares being calculated by distance and estimated arrival times rather than the current way things are done. You could also require a deposit up front that's calculated to be larger than the projected fare and then the difference is refunded when the trip is complete. In any case none of this is insurmountable once self driving cabs reach mature status.
Why not just make it so that once he journey ends the doors stay locked until the correct fair has been deposited? Refusal to pay leads to an operator being called to look at the situation remotely (like your puke check, only for a different purpose). And the offender might eventually get a free ride to the local police station.
You don't do that because it leads to people breaking your cab trying to get out and it could also lead to people taking legal action against the cab company. It's honestly not worth it and there's no reason that you can't just charge before the ride starts.
I cannot seem to find anything about it, but if it is anything like shoplifting detainment, they will just be charged with vandalism and theft of services.
__________________
All this talk about a digital payment system and we are leaving out the basic form. Cash. Not everyone has a credit card or even a bank account.

Re: Google developing competitor to Uber

Posted: 2015-02-06 03:45am
by Darmalus
Soontir C'boath wrote: All this talk about a digital payment system and we are leaving out the basic form. Cash. Not everyone has a credit card or even a bank account.
If calling the g-cab requires using an app of some sort (I can see it happening easily), then such people won't even be considered part of the potential customer base. If you can't pre-load your payment data, you can't even call a g-cab in the first place.