14 ms·

Uber Opening Robotics Research Facility in Pittsburgh to Build Self-Driving Cars

263 points by foobarqux 12y ago

terravion 12y ago
This seems like this should be easy to source... should we call the robotics institute? Cleaned out seems a bit hyperbolic.
NittLion78 12y ago
We're one step closer to Johnny Cab, though I suspect it will work a little better than the version in Total Recall. https://www.youtube.com/watch?v=xGi6j2VrL0o https://www.youtube.com/watch?v=xGi6j2VrL0o
loceng 12y ago
This feels a bit strange, no? Shouldn't all of their drivers immediately stop using Uber?
- codezero 12y ago
  I think individuals who are living paycheck to paycheck as Uber contractors probably aren't able to respond to the possibility that their job will be replaced by automation. Very few people are able to prepare for that kind of thing, actually, even if the warning is in the foreseeable future. If they quit, what should they do to pay the bills in the mean time?
  - titanomachy 12y ago
    Uber's design inherently makes any kind of collective bargaining almost impossible. As long as money can be made by being an Uber contractor, people will do it -- right up to the moment that Uber replaces them all with self-driving cars. There is no picket line to patrol here, no way that the drivers can meaningfully coordinate with one another. Dispatch communicates with drivers, and vice versa, but there is no possibility of lateral communication between drivers. It's pure, emotionless capitalism. Labour is now a commodity like any other. Depending who you ask, this is either the harbinger of a widening wage gap and the poverty that accompanies it, or a long-overdue development which paves the way for optimum efficiency.
    - marcell 12y ago
      I have an idea for unionizing Uber drivers via a smartphone app. It would work around the problem you mentioned, which is that Uber can quickly replace workers who strike for 1 week or more at a time. Basically, do "flash strikes" for 30 minutes at at time. It would work like this: Uber drivers install an app to join the union, and when the union decides to "strike", it sends a message to all drivers in the union. The drivers cancel all rides for some short period of time, say, 30 minutes, and then they resume driving like normal. This driver outage would be disruptive to Uber in terms of user experience, but it wouldn't be long enough to replace the striking workers.
      - titanomachy 12y ago
        Part of the reason unions work is because they sign agreements with employers that prevent them from hiring from outside the union to fill jobs during a strike. I can't imagine a scenario in which Uber enters such an agreement. Even if a union exists, Uber is free to hire non-union members and pay them as much as necessary to get them to work during a strike. In the long run, economics wins out and Uber ends up paying the minimum that enough people are willing to work for, which is exactly what they were doing anyway.
      - mikeash 12y ago
        Then the people who don't join the union (or join but don't follow the instructions) get all the rides during those 30 minutes, and everybody who canceled all their rides during the strike gets kicked out of Uber, and word quickly spreads that going on strike is a good way to instantly lose your livelihood. You'd have to somehow get 100% penetration, not only among current Uber drivers but among all potential Uber drivers in the area, before doing anything. Doesn't sound feasible, especially since anyone who declines to participate will make a killing.
      - audeyisaacs 12y ago
        What would they be trying to achieve by striking? Getting Uber to stop pursuing self-driving cars? It's pointless, because someone at some point will launch the self-driving version of Uber. Also, Surge Pricing would make it difficult to get everyone to hold out for the strike.
  - anigbrowl 12y ago
    Well, they could just switch to working for Lyft instead of a company that is actively trying to automate their jobs away, if for no other reason than to put some upward pressure on Uber's driver payout percentage.
- stdgy 12y ago
  By that logic, shouldn't all of the workers in [manufacturing|food service|sales|etc] do the same? Virtually all firms are going to invest in optimal production/service methodologies in order to stay competitive and capture the largest margins. In the (somewhat) near term, that means replacing a whole lot human labor with automation. But until then, there is still a job there to be done and a lot of people that would like to have the money for doing it. With that said, I don't disagree that the long-term career prospects of taxi driving are plummeting. I think many other careers are nearing their expiration dates as well, though I may be overly pessimistic...
  - monkeyninja 12y ago
    As scientists, we work towards a better future, where human does not need to work, and machine does everything for you, the only tiny bit of work human will do is to ..... be a battery... it is paradise...
- billsimpson 12y ago
  This doesn't represent an immediate threat to Uber's drivers, and I doubt most Uber drivers see themselves as Uber drivers in 3-5 years.
- threeseed 12y ago
  Be serious. Neither Uber nor Google have ANY experience making cars. They aren't the simplest products to sell. Complex regulations that varies city by city, state by state, country by country. And even then the minute you become successful the existing car manufacturers will simply jump into the market and swallow you whole. Uber drivers have absolutely nothing to worry about.
  - orasis 12y ago
    Uber and Google don't need to sell cars. They only need to sell rides.
  - ForHackernews 12y ago
    All the car companies are working on autonomous vehicles, too. I think you're right that Uber and Google don't know enough about building cars, I think the smart money is that somebody like Volvo[0] or Volkswagen will be first to market with a commercial autonomous or semi-autonomous vehicle. [0] http://www.cnet.com/news/a-ride-in-volvos-autonomous-car-how-the-next-step-in-driver-safety-requires-replacing-the-driver/ http://www.cnet.com/news/a-ride-in-volvos-autonomous-car-how...
    - TeMPOraL 12y ago
      Self-driving cars is mostly a software / data problem, not a car problem. Google is much better positioned for that than car manufacturers.
      - ForHackernews 12y ago
        (Leaving aside privacy issues for the moment) An existing car manufacturer could build sensors/data collection into their entire fleet over a period of several years. They could monitor the environment around millions of cars and learn to model driver behavior. I don't know how many Google Maps cars Google has, but if car manufacturers got serious about data collection, there's no way Google could compete with the scale of their installed user base.
- ars 12y ago
  They have nothing to worry about. First Uber isn't going to actually ever do this - programming is not their core competency, and they don't make enough to fund it for the length of time necessary. Second it'll be decades before there are self driving cars, so the drivers can make money in the meantime. Third, once the cars do arrive they will start with long distance highway trips, which is not what Uber does.
  - omarchowdhury 12y ago
    What is Uber's core competency?
    - ars 12y ago
      It's organizing and connecting people. It is not high difficulty engineering projects. Their app and website are pretty straightforward stuff. For a self driving car you need completely different people.
      - omarchowdhury 12y ago
        But surely they can get the right people with their capital.
        ars 12y ago
        You need the right people to know who the right people are. Uber is just not well equipped for this. They could do it certainly (throw enough money at it), but their risk/reward is not in the right place. (But it I mean setup a research team and at least type, I don't expect a self driving car for decades. Maybe a sort of road train that only drives on pre-marked routes would happen first.)
krschultz 12y ago
Google opens an Uber competitor. Uber opens a Google (research) competitor. It will be interesting to see which one can commodize the other. I feel that what Google has built (self-driving cars) is harder to replicate, but Google doesn't have the killer instinct that Uber does. If Android is any guide, Google would rather spread their innovation around to partners rather than to use it to build a killer first party product. I imagine Google will license their future self driving-car stack to the existing car manufacturers, Uber, et al.
- toomuchtodo 12y ago
  > I imagine Google will license their future self driving-car stack to the existing car manufacturers, Uber, et al. Tesla perhaps?
  - seanp2k2 12y ago
    I think a self-driving Tesla would be awesome. If anyone could make that work well and sell it, it'd be them.
  - threeseed 12y ago
    Pretty much only Tesla. Almost every other car manufacturer either has cars in testing or they will have access to the technology through technology sharing.
- serf 12y ago
  >>Google opens an Uber competitor. Uber opens a Google (research) competitor. Yep, but I feel like one significantly outclasses the other as a research facility. Will be interesting to see what happens. >>but Google doesn't have the killer instinct that Uber does. Why do you say that about the most visited site in the world? Genuine question.
  - krschultz 12y ago
    I'm an Android developer, so that experience colors my perception. But compare Apple and Google in mobile. Everything Apple comes up with goes into making their product better, and they will fight with every tool they have to make the iPhone the better smart phone. Every time Google comes up with a high quality product (Google Now, Inbox, Maps), they do everything they can to ship it on iOS as well as Android. If it wasn't for Apple, Google Now & Maps would be just as good on iOS as on Android. Google doesn't use their cloud service advantages as a weapon to make Android better than iOS. They just want their technology in front of everyone, no matter what they use. I see self driving cars as a similar market. Google wants self driving cars to happen. There are lots of good reasons, but from their perspective it is both a play for even more data, and it increases the amount of time we have to use Google products. I can look at ads even longer if I don't have to drive on my commute. Thus for them doesn't it make sense to license/give away all of that tech to everybody? It doesn't matter who manufacturers the car, so long as it uses Google's car software. Thus, plenty of companies will make self driving cars using Google's software. Thus Uber will have plenty of access to self driving cars. If Google was more focused on killing Uber, they would hoard the self driving software for their 1st party initiative. I imagine if Uber invents self driving tech, no one else is getting it. They are far more interested in the core product of driving people from A to B. It's a bit of 'commodize your compliments' and the two companies respective business model. Google operates a scorched earth campaign of making everything around ad supported internet as cheap or free as possible. Uber and Apple are building businesses around charging money for the product they make directly.
seanp2k2 12y ago
I'm imagining a future where there are self-driving busses that route themselves based on where people are and where they want to go. For a dense urban area, I think this has real potential to increase route efficiency and bus utilization while decreasing crowding and unpopulated bus trips.
- sinatra 12y ago
  Bus routes/schedules influence riders' behavior (People reach a bus stop at the time when they expect their bus to arrive). If buses try to automatically adjust route and timing based on riders' needs, riders will start showing up whenever and wherever they want. That makes me feel that the buses may become less utilized, not more.
  - 1971genocide 12y ago
    I do not think so, Many times bus routes are created without thinking about traffic. An app like that that dynamically updates based on where the "blob" of riders are would be awesome.
- Symmetry 12y ago
  I think there's a lot of potential for van-sized automated transit systems. Carrying 4 people gets you most of the fuel efficiency savings of a bus but without the big inefficiencies when you've only got one passenger on an unpopular route.
- JMill 12y ago
  Bridj [1] is a Boston startup exploring this opportunity. Here's a recent article. [2] [1] http://www.bridj.com/ http://www.bridj.com/ [2] https://www.bostonherald.com/business/business_markets/2015/02/bridj_ing_gap_between_uber_costs_t_woes https://www.bostonherald.com/business/business_markets/2015/... (I have no association with the company.)
Qworg 12y ago
Actual Uber announcement: http://blog.uber.com/carnegie-mellon http://blog.uber.com/carnegie-mellon I'm not sure a "partnership" qualifies as "cleaning out".
enahs-sf 12y ago
I find it interesting that Google will probably mint a nice ROI from it's Uber investment and ostensibly dump it all right back into research for robotic cars (I get this isn't how it works, but the idea is novel). CMU has some of the most brilliant minds in robotics - It should be very interesting to see who wins this arms race.
zobzu 12y ago
Heh so thats what this uber job offer was about.. i should have replied! ;)
clientbiller 12y ago
Sniff Sniff... I smell a buyout coming.
ThomPete 12y ago
So Uber is using low wage drivers to make money so they can invest in self driving cars to completely make the drivers obsolete. Talk about a moral paradox.
- akshatpradhan 12y ago
  I'm ok with that. These drivers will move onto intelligent work where they can get trained in less menial work. The only reason a taxi driver stays a taxi driver is because he believes he has no other alternative to make money. Let's get them training in customer service or creative work and give them human jobs.
  - ThomPete 12y ago
    I am not sure what you mean. For most people this is it.
  - Ygg2 12y ago
    I don't think it will work out that way. If Uber dumps them cold they'll have no money to retrain. Also I doubt taxi driver stays taxi driver because of all alternative opportunities he has.
    - akshatpradhan 12y ago
      What were these drivers doing before Uber? Let's not pretend that any job is an ok job. These jobs are not ok for humans and with a basic income, these people can contribute in better more creatives ways. My best friends mom was a cleaner for the Hilton for 20+ years. Sure it was a job, but there was nothing meaningful about it.
      - ThomPete 12y ago
        I am not sure what you are trying to get at. The people who are drivers are drivers for many different reasons some of the are quite well educated and I don't think you can conclude what you seem to conclude. Technology killing jobs is a real problem, technology killing jobs in an increasingly faster and faster pace is an even bigger problem as it doesn't allow people to re-educate themselves. It's a much much much bigger problem than you seem to imply.
mkempe 12y ago
Back in 1987, Carnegie Mellon already had robotic vans [1] trying to drive around the campus. They were slow-moving, so students had ample time to cross the road as they approached. I probably have a picture in a box, somewhere. As I recall a major issue was how to stay on the road. [2] [1] https://www.ri.cmu.edu/pub_files/pub2/thorpe_charles_1988_1/thorpe_charles_1988_1.pdf https://www.ri.cmu.edu/pub_files/pub2/thorpe_charles_1988_1/... [2] https://www.ri.cmu.edu/pub_files/pub3/thorpe_charles_1988_1/thorpe_charles_1988_1.pdf https://www.ri.cmu.edu/pub_files/pub3/thorpe_charles_1988_1/...
- Animats 12y ago
  I was in the Navlab once. It had a crew of five - a driver, an operator, and three people in the back at rack-mounted workstations. They had to convert the thing to hydraulic motor drive so it could drive slowly enough. Their LIDAR was very advanced. They had a 2-axis scanning LIDAR, with one rotating and one nodding mirror. So they had full 3D depth images. Most of the DARPA Grand Challenge vehicles only had line scanners.
Animats 12y ago
That makes sense. CMU has been working with Cadillac on self-driving cars, and has demoed them in Washington DC traffic. The CMU/Cadillac car has nothing visible externally which marks it as a self-driving car. They may be closer to a production product than Google.
- jayjay71 12y ago
  I'm pretty sure they aren't partnering with CMU, but rather offering many senior scientists and faculty members lucrative salaries to leave academia and work for them. So the Cadillac IP still belongs to the university. In fact, that actually spun out into a separate company called Ottomatika (which has a licensing deal from CMU).
  - Animats 12y ago
    They are formally partnering with CMU. Here's the CMU press release. http://www.cmu.edu/news/stories/archives/2015/february/uber-partnership.html http://www.cmu.edu/news/stories/archives/2015/february/uber-... "Uber and Carnegie Mellon University are announcing today a strategic partnership that includes the creation of the Uber Advanced Technologies Center in Pittsburgh, near the CMU campus. ... The agreement also will include funding from Uber for faculty chairs and graduate fellowships."
    - jayjay71 12y ago
      I stand corrected.
monkeyninja 12y ago
Imaging that, we can get rid of all drivers, must be brilliant...
omarforgotpwd 12y ago
Google probably wants to run a big data center that coordinates and organizes all the world's cars. A reliable centrally coordinated self driving vehicle network would greatly canabalize air and train travel and cargo, so the winners in this space will control the world's transporation network. Google will probably win, but either way the world economy will benefit massively from the reduction in transportation costs.
- cLeEOGPw 12y ago
  Google may benefit from it, but it's not like it would be irreplaceable. Google would just have the advantage of pioneering a new market. Besides governments in other countries (since I assume US will stay "wild capitalism" for longer than others) can just call an open competition for such controlled network management services and then be semi-controlled by that government. It's not like google or anyone gonna have full actual control over who goes where even when/if such system will exist.
omarforgotpwd 12y ago
Google wants to run a big data center that coordinates and organizes all the world's cars, probably. Reliable self driving vehicles would greatly canabalize air and train travel and cargo, so the winners in this space will control the world's transporation network. Google will probably win
Zigurd 12y ago
It's interesting to see comments of the nature of "Google doesn't have the killer instinct that Uber does." This is a fight over the businesses of logistics and transportation in general, and on a global scale. You can expect Amazon to join the fight. This is not limited to driving people across town, and winning is vastly more valuable. Nobody is going to hold back.
- lkbm 12y ago
  I always felt like long-haul B2B deliveries (e.g., from warehouse to supermarket) was the most sensible starting point for actual use of self-driving vehicles on the roads. No worries about convincing customers to give up control, the extra equipment overhead is probably relatively small when you're already working with a massive truck, no stopping to sleep/eat, no tired drivers...
coldcode 12y ago
Self-driving cars in a world where all cars are self-driving is doable. Self-driving cars sharing the road with the morons I deal with every day is not going to happen any time soon, much less dealing with accidents, road closures, ice, snow, rain and the occasional angry politician.
- ahallock 12y ago
  This is irrational. If self-driving cars are overall safer than their human counterparts, why wouldn't we allow a mix? I mean, we don't ban human drivers because of all the accidents and deaths caused of human errors, so why would be not allow self-driving cars, even with some failures?
  - Lambdanaut 12y ago
    > This is irrational Humans can be pretty irrational, and law unfortunately isn't enacted in the science lab. We'll just have to wait and see who wins on the political level.
- lkbm 12y ago
  > Self-driving cars sharing the road with the morons I deal with every day is not going to happen any time soon It's already happening, and they seem to be working pretty well. The most difficult issue is going to be politians.
Intoo 12y ago
soon techcrunch headlines would be: UBER DRIVERS LAID-OFF, REPLACED BY SELF-DRIVING CARS
Intoo 12y ago
soon techcrunch headlines would be: UBER DRIVERS LAID-OFF, REPLACED BY SELF-DRIVING CARS
fspacef 12y ago
Either way, Google vs. Uber means cheaper cab rides for all of us...

263 points by foobarqux 12y ago

This seems like this should be easy to source... should we call the robotics institute? Cleaned out seems a bit hyperbolic.

NittLion78 12y ago

We're one step closer to Johnny Cab, though I suspect it will work a little better than the version in Total Recall. https://www.youtube.com/watch?v=xGi6j2VrL0o https://www.youtube.com/watch?v=xGi6j2VrL0o

loceng 12y ago

This feels a bit strange, no? Shouldn't all of their drivers immediately stop using Uber?

codezero 12y ago
I think individuals who are living paycheck to paycheck as Uber contractors probably aren't able to respond to the possibility that their job will be replaced by automation. Very few people are able to prepare for that kind of thing, actually, even if the warning is in the foreseeable future. If they quit, what should they do to pay the bills in the mean time?
- titanomachy 12y ago
  Uber's design inherently makes any kind of collective bargaining almost impossible. As long as money can be made by being an Uber contractor, people will do it -- right up to the moment that Uber replaces them all with self-driving cars. There is no picket line to patrol here, no way that the drivers can meaningfully coordinate with one another. Dispatch communicates with drivers, and vice versa, but there is no possibility of lateral communication between drivers. It's pure, emotionless capitalism. Labour is now a commodity like any other. Depending who you ask, this is either the harbinger of a widening wage gap and the poverty that accompanies it, or a long-overdue development which paves the way for optimum efficiency.
  - marcell 12y ago
    I have an idea for unionizing Uber drivers via a smartphone app. It would work around the problem you mentioned, which is that Uber can quickly replace workers who strike for 1 week or more at a time. Basically, do "flash strikes" for 30 minutes at at time. It would work like this: Uber drivers install an app to join the union, and when the union decides to "strike", it sends a message to all drivers in the union. The drivers cancel all rides for some short period of time, say, 30 minutes, and then they resume driving like normal. This driver outage would be disruptive to Uber in terms of user experience, but it wouldn't be long enough to replace the striking workers.
    - titanomachy 12y ago
      Part of the reason unions work is because they sign agreements with employers that prevent them from hiring from outside the union to fill jobs during a strike. I can't imagine a scenario in which Uber enters such an agreement. Even if a union exists, Uber is free to hire non-union members and pay them as much as necessary to get them to work during a strike. In the long run, economics wins out and Uber ends up paying the minimum that enough people are willing to work for, which is exactly what they were doing anyway.
    - mikeash 12y ago
      Then the people who don't join the union (or join but don't follow the instructions) get all the rides during those 30 minutes, and everybody who canceled all their rides during the strike gets kicked out of Uber, and word quickly spreads that going on strike is a good way to instantly lose your livelihood. You'd have to somehow get 100% penetration, not only among current Uber drivers but among all potential Uber drivers in the area, before doing anything. Doesn't sound feasible, especially since anyone who declines to participate will make a killing.
    - audeyisaacs 12y ago
      What would they be trying to achieve by striking? Getting Uber to stop pursuing self-driving cars? It's pointless, because someone at some point will launch the self-driving version of Uber. Also, Surge Pricing would make it difficult to get everyone to hold out for the strike.
- anigbrowl 12y ago
  Well, they could just switch to working for Lyft instead of a company that is actively trying to automate their jobs away, if for no other reason than to put some upward pressure on Uber's driver payout percentage.
stdgy 12y ago
By that logic, shouldn't all of the workers in [manufacturing|food service|sales|etc] do the same? Virtually all firms are going to invest in optimal production/service methodologies in order to stay competitive and capture the largest margins. In the (somewhat) near term, that means replacing a whole lot human labor with automation. But until then, there is still a job there to be done and a lot of people that would like to have the money for doing it. With that said, I don't disagree that the long-term career prospects of taxi driving are plummeting. I think many other careers are nearing their expiration dates as well, though I may be overly pessimistic...
- monkeyninja 12y ago
  As scientists, we work towards a better future, where human does not need to work, and machine does everything for you, the only tiny bit of work human will do is to ..... be a battery... it is paradise...
billsimpson 12y ago
This doesn't represent an immediate threat to Uber's drivers, and I doubt most Uber drivers see themselves as Uber drivers in 3-5 years.
threeseed 12y ago
Be serious. Neither Uber nor Google have ANY experience making cars. They aren't the simplest products to sell. Complex regulations that varies city by city, state by state, country by country. And even then the minute you become successful the existing car manufacturers will simply jump into the market and swallow you whole. Uber drivers have absolutely nothing to worry about.
- orasis 12y ago
  Uber and Google don't need to sell cars. They only need to sell rides.
- ForHackernews 12y ago
  All the car companies are working on autonomous vehicles, too. I think you're right that Uber and Google don't know enough about building cars, I think the smart money is that somebody like Volvo[0] or Volkswagen will be first to market with a commercial autonomous or semi-autonomous vehicle. [0] http://www.cnet.com/news/a-ride-in-volvos-autonomous-car-how-the-next-step-in-driver-safety-requires-replacing-the-driver/ http://www.cnet.com/news/a-ride-in-volvos-autonomous-car-how...
  - TeMPOraL 12y ago
    Self-driving cars is mostly a software / data problem, not a car problem. Google is much better positioned for that than car manufacturers.
    - ForHackernews 12y ago
      (Leaving aside privacy issues for the moment) An existing car manufacturer could build sensors/data collection into their entire fleet over a period of several years. They could monitor the environment around millions of cars and learn to model driver behavior. I don't know how many Google Maps cars Google has, but if car manufacturers got serious about data collection, there's no way Google could compete with the scale of their installed user base.
ars 12y ago
They have nothing to worry about. First Uber isn't going to actually ever do this - programming is not their core competency, and they don't make enough to fund it for the length of time necessary. Second it'll be decades before there are self driving cars, so the drivers can make money in the meantime. Third, once the cars do arrive they will start with long distance highway trips, which is not what Uber does.
- omarchowdhury 12y ago
  What is Uber's core competency?
  - ars 12y ago
    It's organizing and connecting people. It is not high difficulty engineering projects. Their app and website are pretty straightforward stuff. For a self driving car you need completely different people.
    - omarchowdhury 12y ago
      But surely they can get the right people with their capital.
      - ars 12y ago
        You need the right people to know who the right people are. Uber is just not well equipped for this. They could do it certainly (throw enough money at it), but their risk/reward is not in the right place. (But it I mean setup a research team and at least type, I don't expect a self driving car for decades. Maybe a sort of road train that only drives on pre-marked routes would happen first.)

krschultz 12y ago

Google opens an Uber competitor. Uber opens a Google (research) competitor. It will be interesting to see which one can commodize the other. I feel that what Google has built (self-driving cars) is harder to replicate, but Google doesn't have the killer instinct that Uber does. If Android is any guide, Google would rather spread their innovation around to partners rather than to use it to build a killer first party product. I imagine Google will license their future self driving-car stack to the existing car manufacturers, Uber, et al.

toomuchtodo 12y ago
> I imagine Google will license their future self driving-car stack to the existing car manufacturers, Uber, et al. Tesla perhaps?
- seanp2k2 12y ago
  I think a self-driving Tesla would be awesome. If anyone could make that work well and sell it, it'd be them.
- threeseed 12y ago
  Pretty much only Tesla. Almost every other car manufacturer either has cars in testing or they will have access to the technology through technology sharing.
serf 12y ago
>>Google opens an Uber competitor. Uber opens a Google (research) competitor. Yep, but I feel like one significantly outclasses the other as a research facility. Will be interesting to see what happens. >>but Google doesn't have the killer instinct that Uber does. Why do you say that about the most visited site in the world? Genuine question.
- krschultz 12y ago
  I'm an Android developer, so that experience colors my perception. But compare Apple and Google in mobile. Everything Apple comes up with goes into making their product better, and they will fight with every tool they have to make the iPhone the better smart phone. Every time Google comes up with a high quality product (Google Now, Inbox, Maps), they do everything they can to ship it on iOS as well as Android. If it wasn't for Apple, Google Now & Maps would be just as good on iOS as on Android. Google doesn't use their cloud service advantages as a weapon to make Android better than iOS. They just want their technology in front of everyone, no matter what they use. I see self driving cars as a similar market. Google wants self driving cars to happen. There are lots of good reasons, but from their perspective it is both a play for even more data, and it increases the amount of time we have to use Google products. I can look at ads even longer if I don't have to drive on my commute. Thus for them doesn't it make sense to license/give away all of that tech to everybody? It doesn't matter who manufacturers the car, so long as it uses Google's car software. Thus, plenty of companies will make self driving cars using Google's software. Thus Uber will have plenty of access to self driving cars. If Google was more focused on killing Uber, they would hoard the self driving software for their 1st party initiative. I imagine if Uber invents self driving tech, no one else is getting it. They are far more interested in the core product of driving people from A to B. It's a bit of 'commodize your compliments' and the two companies respective business model. Google operates a scorched earth campaign of making everything around ad supported internet as cheap or free as possible. Uber and Apple are building businesses around charging money for the product they make directly.

seanp2k2 12y ago

I'm imagining a future where there are self-driving busses that route themselves based on where people are and where they want to go. For a dense urban area, I think this has real potential to increase route efficiency and bus utilization while decreasing crowding and unpopulated bus trips.

sinatra 12y ago
Bus routes/schedules influence riders' behavior (People reach a bus stop at the time when they expect their bus to arrive). If buses try to automatically adjust route and timing based on riders' needs, riders will start showing up whenever and wherever they want. That makes me feel that the buses may become less utilized, not more.
- 1971genocide 12y ago
  I do not think so, Many times bus routes are created without thinking about traffic. An app like that that dynamically updates based on where the "blob" of riders are would be awesome.
Symmetry 12y ago
I think there's a lot of potential for van-sized automated transit systems. Carrying 4 people gets you most of the fuel efficiency savings of a bus but without the big inefficiencies when you've only got one passenger on an unpopular route.
JMill 12y ago
Bridj [1] is a Boston startup exploring this opportunity. Here's a recent article. [2] [1] http://www.bridj.com/ http://www.bridj.com/ [2] https://www.bostonherald.com/business/business_markets/2015/02/bridj_ing_gap_between_uber_costs_t_woes https://www.bostonherald.com/business/business_markets/2015/... (I have no association with the company.)

Qworg 12y ago

Actual Uber announcement: http://blog.uber.com/carnegie-mellon http://blog.uber.com/carnegie-mellon I'm not sure a "partnership" qualifies as "cleaning out".

enahs-sf 12y ago

I find it interesting that Google will probably mint a nice ROI from it's Uber investment and ostensibly dump it all right back into research for robotic cars (I get this isn't how it works, but the idea is novel). CMU has some of the most brilliant minds in robotics - It should be very interesting to see who wins this arms race.

zobzu 12y ago

Heh so thats what this uber job offer was about.. i should have replied! ;)

clientbiller 12y ago

Sniff Sniff... I smell a buyout coming.

ThomPete 12y ago

So Uber is using low wage drivers to make money so they can invest in self driving cars to completely make the drivers obsolete. Talk about a moral paradox.

akshatpradhan 12y ago
I'm ok with that. These drivers will move onto intelligent work where they can get trained in less menial work. The only reason a taxi driver stays a taxi driver is because he believes he has no other alternative to make money. Let's get them training in customer service or creative work and give them human jobs.
- ThomPete 12y ago
  I am not sure what you mean. For most people this is it.
- Ygg2 12y ago
  I don't think it will work out that way. If Uber dumps them cold they'll have no money to retrain. Also I doubt taxi driver stays taxi driver because of all alternative opportunities he has.
  - akshatpradhan 12y ago
    What were these drivers doing before Uber? Let's not pretend that any job is an ok job. These jobs are not ok for humans and with a basic income, these people can contribute in better more creatives ways. My best friends mom was a cleaner for the Hilton for 20+ years. Sure it was a job, but there was nothing meaningful about it.
    - ThomPete 12y ago
      I am not sure what you are trying to get at. The people who are drivers are drivers for many different reasons some of the are quite well educated and I don't think you can conclude what you seem to conclude. Technology killing jobs is a real problem, technology killing jobs in an increasingly faster and faster pace is an even bigger problem as it doesn't allow people to re-educate themselves. It's a much much much bigger problem than you seem to imply.

mkempe 12y ago

Back in 1987, Carnegie Mellon already had robotic vans [1] trying to drive around the campus. They were slow-moving, so students had ample time to cross the road as they approached. I probably have a picture in a box, somewhere. As I recall a major issue was how to stay on the road. [2] [1] https://www.ri.cmu.edu/pub_files/pub2/thorpe_charles_1988_1/thorpe_charles_1988_1.pdf https://www.ri.cmu.edu/pub_files/pub2/thorpe_charles_1988_1/... [2] https://www.ri.cmu.edu/pub_files/pub3/thorpe_charles_1988_1/thorpe_charles_1988_1.pdf https://www.ri.cmu.edu/pub_files/pub3/thorpe_charles_1988_1/...

Animats 12y ago
I was in the Navlab once. It had a crew of five - a driver, an operator, and three people in the back at rack-mounted workstations. They had to convert the thing to hydraulic motor drive so it could drive slowly enough. Their LIDAR was very advanced. They had a 2-axis scanning LIDAR, with one rotating and one nodding mirror. So they had full 3D depth images. Most of the DARPA Grand Challenge vehicles only had line scanners.

Animats 12y ago

That makes sense. CMU has been working with Cadillac on self-driving cars, and has demoed them in Washington DC traffic. The CMU/Cadillac car has nothing visible externally which marks it as a self-driving car. They may be closer to a production product than Google.

jayjay71 12y ago
I'm pretty sure they aren't partnering with CMU, but rather offering many senior scientists and faculty members lucrative salaries to leave academia and work for them. So the Cadillac IP still belongs to the university. In fact, that actually spun out into a separate company called Ottomatika (which has a licensing deal from CMU).
- Animats 12y ago
  They are formally partnering with CMU. Here's the CMU press release. http://www.cmu.edu/news/stories/archives/2015/february/uber-partnership.html http://www.cmu.edu/news/stories/archives/2015/february/uber-... "Uber and Carnegie Mellon University are announcing today a strategic partnership that includes the creation of the Uber Advanced Technologies Center in Pittsburgh, near the CMU campus. ... The agreement also will include funding from Uber for faculty chairs and graduate fellowships."
  - jayjay71 12y ago
    I stand corrected.

monkeyninja 12y ago

Imaging that, we can get rid of all drivers, must be brilliant...

omarforgotpwd 12y ago

Google probably wants to run a big data center that coordinates and organizes all the world's cars. A reliable centrally coordinated self driving vehicle network would greatly canabalize air and train travel and cargo, so the winners in this space will control the world's transporation network. Google will probably win, but either way the world economy will benefit massively from the reduction in transportation costs.

cLeEOGPw 12y ago
Google may benefit from it, but it's not like it would be irreplaceable. Google would just have the advantage of pioneering a new market. Besides governments in other countries (since I assume US will stay "wild capitalism" for longer than others) can just call an open competition for such controlled network management services and then be semi-controlled by that government. It's not like google or anyone gonna have full actual control over who goes where even when/if such system will exist.

omarforgotpwd 12y ago

Google wants to run a big data center that coordinates and organizes all the world's cars, probably. Reliable self driving vehicles would greatly canabalize air and train travel and cargo, so the winners in this space will control the world's transporation network. Google will probably win

Zigurd 12y ago

It's interesting to see comments of the nature of "Google doesn't have the killer instinct that Uber does." This is a fight over the businesses of logistics and transportation in general, and on a global scale. You can expect Amazon to join the fight. This is not limited to driving people across town, and winning is vastly more valuable. Nobody is going to hold back.

lkbm 12y ago
I always felt like long-haul B2B deliveries (e.g., from warehouse to supermarket) was the most sensible starting point for actual use of self-driving vehicles on the roads. No worries about convincing customers to give up control, the extra equipment overhead is probably relatively small when you're already working with a massive truck, no stopping to sleep/eat, no tired drivers...

coldcode 12y ago

Self-driving cars in a world where all cars are self-driving is doable. Self-driving cars sharing the road with the morons I deal with every day is not going to happen any time soon, much less dealing with accidents, road closures, ice, snow, rain and the occasional angry politician.

ahallock 12y ago
This is irrational. If self-driving cars are overall safer than their human counterparts, why wouldn't we allow a mix? I mean, we don't ban human drivers because of all the accidents and deaths caused of human errors, so why would be not allow self-driving cars, even with some failures?
- Lambdanaut 12y ago
  > This is irrational Humans can be pretty irrational, and law unfortunately isn't enacted in the science lab. We'll just have to wait and see who wins on the political level.
lkbm 12y ago
> Self-driving cars sharing the road with the morons I deal with every day is not going to happen any time soon It's already happening, and they seem to be working pretty well. The most difficult issue is going to be politians.

Intoo 12y ago

soon techcrunch headlines would be: UBER DRIVERS LAID-OFF, REPLACED BY SELF-DRIVING CARS

Intoo 12y ago

soon techcrunch headlines would be: UBER DRIVERS LAID-OFF, REPLACED BY SELF-DRIVING CARS

fspacef 12y ago

Either way, Google vs. Uber means cheaper cab rides for all of us...