Uber deploying fleet of self driving SUV in Pittsburgh

Uber deploying fleet of self driving SUV in Pittsburgh

I am excited as I have seen these driving around for about a year and now they are "real"
Seems like more hardware than Tesla (spinning lasers on the roof). This is a BIG deal for Pittsburgh as they have hired over
500 employees and even built a test track on an old steel mill site

Good news is that although they are using Volvos XC90 they could use other auto makers as well (i.e. Tesla)
I wonder why they are not working with Tesla more closely.

Vawlkus | 18 août 2016

probably can't afford to, until the Model 3 rolls out.

rossRallen | 18 août 2016


francoisbitz | 18 août 2016

XC90 is not much cheaper than Model X base.

carlk | 18 août 2016

Tesla is a potential competitor since Elon already has in his plan autonomous driving and shared mobility. It does not sound like he wants to do it with Uber or any other companies. He does not need to either since there is not much Uber can offer that Tesla could not do on its own.

Remnant | 18 août 2016

@ francoisbitz (August 18, 2016)

<< Uber deploying fleet of self driving SUV in Pittsburgh ... I wonder why they are not working with Tesla more closely. >>

For enhanced, quasi-autonomous self-driving, the Automobile SW will have to shift from a mere Bot Mode to a full AI Mode, which would mean, among other things, that reliance on road signage will have to be supplanted by reliance on image analysis and interpretation.

HD full-color video, Enhanced Radar Data Display, and AI Steering and Collision Avoidance Algorithms would have to be embedded in the SW for autonomous self-driving. Video game SW appears to contain a good bit of ideas in this regard. Tesla has not hinted of developments in that direction.

Check: | 18 août 2016

Not self-driving exactly: "Uber’s Pittsburgh fleet, which will be supervised by humans in the driver’s seat for the time being, consists of specially modified Volvo XC90 sport-utility vehicles outfitted with dozens of sensors that use cameras, lasers, radar, and GPS receivers. Volvo Cars has so far delivered a handful of vehicles out of a total of 100 due by the end of the year. The two companies signed a pact earlier this year to spend $300 million to develop a fully autonomous car that will be ready for the road by 2021."

It's a science project. I predict that it will drag on for many years as CMU milks Uber's bankroll for its science project.

carlk | 19 août 2016


Haha. Sounds like you have worked on university projects before. I did work at and with universities, one of them happens to be with CMU in a different area. I can say I would not do it if it's my own money.

Gayatrikr | 20 août 2016

When did pa approve driverless cars ? Or this does not come under that clause ?

elguapo | 21 août 2016

Ummm. They can't work with Tesla because Tesla's driver assist tech isn't the best out there today. Mercedes and Volvo have more capable systems as it relates to more sensors, etc. Tesla has two advantages. They're overconfident, so they were willing to put their tech out there and let it run above 35 mph while companies like Audi speed limited theirs. Second, Tesla appears to have better software today.

Don't flame me. Anyone who knows me knows I am a huge fan and have a sub 500 vin X and an S. I just try to tell it like it is. I love TM, but their AP tech isn't the best out there anymore. It's two year old hardware...

carlk | 22 août 2016


It's a "driverless" car with a driver. BTW Google cars could do that many years ago and is even more advanced than that today. They don't even have steering wheel or pedal now. I can't imagine why Uber have not approached Google first. They probably had but perhaps Google instead of wanting to sell the technology wanted to have the whole banana. I think Uber is in a real tough position now. Any company got the self driving fleet out first will totally wipe them out. The sharing app is very simple to do. Uber's infrastructure is going to be totally worthless.


Sensor technology is pretty mature now and not a differentiator. It's the software that makes the difference. Human drivers have only two eyes but our brain can process large amounts of data and make super fast logistic decisions. More sensors can help some when you have weak software but it's still the software and processing power that will help you to go to the next level. Tesla's fleet learning and 3D mapping is second to none. It's also putting a lot of real efforts in AI and processor development. We'll see.

And what you said "Mercedes and Volvo have more capable systems as it relates to more sensors, etc." is of course not true. They don't have more capable systems than Tesla's, even the AP1.0. You probably have watched too many of their TV commercials. Here is a review of Tesla's Autopilot and MB's Drive Pilot.

How funny that the author said "Did anyone in Stuttgart drive a Tesla on Autopilot? Even once?"

Ross1 | 22 août 2016

Generally speaking, isnt it right that Volvo pick up responsibility for accidents, whereas Tesla doesnt?
If that extends to fleet purchases it is a big persuader.

Also, how can there be much negative press if Volvo says: Hey, its not what we expected but we will pick up the tab wholeheartedly.

elguapo | 22 août 2016

@carlk I've driven an Audi with TrafficJam Assist or whatever the lame name is and it works exactly as ASAP except it actually recognized a stopped car in front of me and went around it.

I respect your opinion, but do you disagree that the Tesla US sensors are inferior to radar used for BSM, for example -and other ways of monitoring what's around the car?

Regardless, I realize software is critical, I guess I was really commenting ore on the hardware, where TM has fallen behind - for now...

elguapo | 22 août 2016