Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
×
Transportation China

Baidu's Supercheap Robotaxis Should Scare the Hell Out of the US (theverge.com) 93

Baidu's new Apollo Go robotaxi brings significant advances in affordability and scalability that should make U.S. competitors like Waymo a bit nervous, according to The Verge's Andrew J. Hawkins. From the report: The RT6 is the sixth generation of Apollo Go's driverless vehicle, which made its official debut in May 2024. It's a purpose-built, Level 4 autonomous vehicle, meaning it's built without the need for a human driver. And here's the thing that should make US competitors nervous: adopting a battery-swapping solution, the price for one individual RT6 is "under $30,000," Baidu CEO Robin Li said in an earnings call. "All the strengths just mentioned above are driving us forward, paving the way to validate our business model," Li added. [...]

We still don't know the net effect of Baidu's cost improvements. But bringing down the upfront cost of each individual vehicle to below $30,000 will go a long way toward improving the company's unit economics, in which each vehicle brings in more money than it costs. There are still a lot of outstanding costs to consider, such as hardware depreciation and fleet maintenance, but from what Baidu is signaling, things are on the right track. From the looks of it, the company is passing those savings along to its customers. Base fares start as low as 4 yuan (around 55 cents), compared with 18 yuan (around $2.48) for a taxi driven by a human, according to state media outlet Global Times. Apollo Go said it has provided 988,000 rides across all of China in Q3 2024 -- a year-over-year growth of 20 percent. And cumulative public rides reached 8 million in October.

This discussion has been archived. No new comments can be posted.

Baidu's Supercheap Robotaxis Should Scare the Hell Out of the US

Comments Filter:
  • than being asked to ride in a "supercheap robotaxi".
  • Baidu's going to have to make them in the US for them to be economical.

    • They won't be economical then. Labor costs too much. Chinese personnel are a bit more efficient if you only look at the numbers...
  • Things work differently in China, obviously. Government-run companies don't have to worry about private lawsuits when their robotaxis cause accidents and kill people. It's considered acceptable losses for the sake of market dominance.
    • One of the guys hired to drive me in China told me that if you have a collision with a city bus you are deemed to be at fault regardless of circumstances. Even if the bus runs into you.

      I didn't think to ask if "fault" here means financial liability, but if the EV robotaxi operator can get similar status to city busses (as a public service or whatever) then yes your legal exposure is vastly better than if you were working under the rules we have here.

      • by AmiMoJo ( 196126 )

        That's not quite correct. It's more a case of you will find it hard to provide you were not at fault, because busses have priority and are covered in cameras.

        It's actually a pretty similar situation in the UK.

    • by cstacy ( 534252 )

      Things work differently in China, obviously. Government-run companies don't have to worry about private lawsuits when their robotaxis cause accidents and kill people. It's considered acceptable losses for the sake of market dominance.

      Cue the Ford Pinto.

  • No, not really. (Score:2, Insightful)

    by sarren1901 ( 5415506 )

    American auto-firms could lower the prices of their vehicles but that would lower profits. Can't have that... If competition from China scares them, tough cookie.

    Sadly, they will just petition the US government to "protect" them from competition. Instead of being innovative or offering more options people want, they double down on bigger, less efficient vehicles and in a lot of ways, feels like big auto just wants to add more and more premium features to the vehicles that I don't need and certainly don't wa

    • Re:No, not really. (Score:5, Interesting)

      by Ogive17 ( 691899 ) on Friday November 22, 2024 @06:17PM (#64965693)
      First - is there really high demand for driver-less taxis?

      Second - enough with the protectionism talk. CCCP has been propping up their domestic manufacturing for decades while limiting foreign investment and foreign firms from selling. They are also quick to spread FUD in their controlled media against many foreign countries to encourage their population to reject those brands. I do not have an issue with protectionist tariffs on Chinese imports.

      Third - wouldn't light rail be even better than a bunch of battery powered taxis?
      • Re: (Score:2, Insightful)

        by AmiMoJo ( 196126 )

        Look in the mirror. The view from Europe is that you guys are doing pretty much everything you accuse them of these days.

        It doesn't matter. They got there first, again. There are plenty of other places they can sell them to. We aren't going to get through this with tariffs, we need to get competitive.

        • by Ogive17 ( 691899 )
          Did my post have anything to do with Europe? Nope.

          I'm not naive, I know the US is not always a great trade partner. My comment was specific to US-China trade.

          Getting competitive means heavy subsidies in many cases.
      • Why wouldn't there be demand for such a thing? Providing transportation with low labor cost sounds good for the consumers. Having a capital cost that is very incremental instead of one humungous monolithic cost (like rail) is also flexible and beneficial, because you can start realizing the benefits much sooner - although yes there are some inefficiencies due to having more vehicles. Individual vehicles is also much more robust to disruption than something that damages a section of track.

        There's a lot to co

      • while limiting foreign investment and foreign firms from selling

        Not really. They have been propping up local domestic firms, but the reality is a large part of the domestic manufacturing in China *is* foreign investment. The largest car company in China a few years ago was Volkswagen, both in terms of sales as well as domestic manufacturing. If you limit the market to ICE vehicles only, it is *still* VW.

        The biggest issue is those foreign companies have been just as averse to EVs in China as everywhere else, so when by policy China moved in the EV direction they were mas

      • To your first question - if it lowers the cost of a ride, absolutely. Maintaining a livable wage for the meatbag operating the machine is almost always the most significant cost. It is not labour that really brings any value to our society.
    • Instead of being innovative or offering more options people want, they double down on bigger, less efficient vehicles

      Actually it looks a lot like that is what people want.

      "swing voters did not believe Harris when she said she had modified her previous stands: Large majorities said she really did want to ban fracking and require all U.S.-made cars to be electric by 2035, according to polling by the Democratic strategy group Blueprint. Of these skeptical swing voters, more than three-quarters voted for Trump."

      https://www.washingtonpost.com... [washingtonpost.com]

      Sorry bout the paywall.

    • What about reducing labor costs? That also seems like a huge discrepancy in manufacturing costs. Getting innovative here in America seems to fly in the face of union's policies.
    • The "Scare" part in the title demonstrates a lack economic perspective. Cheap products available for my purchase and utilization should scare the hell out of me? A large price difference waiting to be exploited by US industry is bad? Nobody should be scared - we should see the opportunity and make the best of it. And yeah, China is totalitarian communistic state with hegemonic ambitions – I get all that. My point is nobody needs to scared of this.

  • Most of the cost of an EV is the cost of the battery. If you are willing to accept a range of 90 miles or so you can make an EV really cheap.

    It seems that Baidu made the calculation that the low-range of a cheap EV doesn't matter if a) it is operating in a limited area, and b) it can swap out the battery quickly at accessible stations. So the operation model is it that it will do 5 to a dozen or more rides depending on the trip length, zips into a change station, gets a new battery. Repeat.

    Not a ba

    • The driverless hardware like the LIDAR is also kind of expensive, although maybe they found a way to make it cheaper, or are able to get good results from a lower quality version.
      • by vlad30 ( 44644 )

        The driverless hardware like the LIDAR is also kind of expensive, although maybe they found a way to make it cheaper, or are able to get good results from a lower quality version.

        They made it cheaper in a factory in china. And there is the problem the USA and western countries outsourced production and those things are cheaper for Chinese companies not western companies

    • It also means you can have fewer battery safety features if your smaller, cheaper battery is less of a toxic chemical/electrical fire hazard than a full-sized battery. Then you get rid of some other safety features (by never incorporating them in the first place) aimed at passenger safety, and you can bring costs down even more.

      Regardless, $30k per EV is still hella expensive compared to the $10k and less EVs on the Chinese market right now.

  • by Fishbone ( 115047 ) on Friday November 22, 2024 @06:25PM (#64965713)

    As someone who has worked on the technical side of a rollout of an Apollo platform on US roads, no, there's not a lot to be terrified about (yet, if at all).

    While there are a lot of situations where Apollo is able to drive safely (mostly in situations of light-to-no traffic), it still struggles with:
    - navigating tight spaces
    - sensor fusion
    - false positive collision avoidance ("sees" dust or water spray as a possible collision object)
    - inclement weather in general
    - requiring a HD map for navigating the world (which is expensive to build and even more expensive to maintain over time)
    - requiring realtime kinematics corrections for localization, which isn't widely available

    I love the fact that it's more or less an open source project, but after operating the platform for a good while I can safely say that I wouldn't worry too much about it being the death of manually operated cars. At very least, not in the near term.

    • What is sensor fusion?
      • The capability for different sensor packages (lidar, radar, cameras, etc) to work in concert for better localization and object detection. The ability to perceive the world around a vehicle as a cohesive whole, instead of several different individual sensor domains that don't really "talk" to one another. In a perfect world, this would also provide a level of redundancy for a failed system, such as an obstructed sensor.

        • What is the difficulty? I'm just trying to understand here. It seems like there are already a lot of standard algorithms with potential for solving this problem.
          • You'd probably be surprised at how little of this is actually present on modern platforms (I know I was). Most platforms, regardless of how many types of sensors that a marketing team will say a vehicle has, lean heavily on one type of sensor. For Apollo, it's lidar. If the lidar unit(s) have problems, even if there are additional types of sensors in your suite (ours additionally had both multiple radars and cameras), the system either fully stops working, or working so badly that it becomes necessary for a

            • Interesting. If only I had lots of time I'd like to apply myself to that problem and see what kind of improvement I could get.
  • by Anonymous Coward

    They will just enact protectionist measures, so there's no competition.

  • Really I think we should be celebrating the fact that autonomous passenger vehicles are nearly ready to go mainstream. My parents would find this liberating as they age. I woudl find it convenient, cities may find that ubiquitous cheap taxis cut down on much of the car ownership that has clogged roads, hogged space for parking and made holes in household budgets.
    • I agree, it's quite a statement on where we are that our response is just assumed to be fear. Self-driving cars are going to be the greatest labor-saving invention of the last 80 years.
  • There goes that $30,000 price tag...
  • Does anyone seriously think Chinese robo-taxis are going to be allowed to compete in the United States? The real problem is that US robo-taxis won't be able to compete in the rest of the world. Which means that American manufacturers are only going to be supplying the limited domestic market That means a much lower volume and likely much higher prices. But that is where everything is headed.
  • Are non-subsidized Chinese cars competitive? That's the real question. Certainly labor costs are cheaper in China, but labor typically constitutes only a small part of the total costs, typically only around 7% for US car makers [cnn.com]. So, Chinese car prices are indeed lower, but the suspicion is that the Chinese government is subsidizing those cars and furthermore that prices are deflated for export. There's no practical way to estimate the true unsubsidized cost of Chinese cars and how much or whether Chines

  • I wouldn't even go inside a Chinese building. You think I'm going to trust their vehicles, ai, or self driving tech?

It is better to travel hopefully than to fly Continental.

Working...