Only large corporations running fleets will have the necessary deep pockets to fund and then absorb the losses with these Robotaxi services. There's little to no room for individual owner/operators and long term viability will be a real concern even for fleet operators.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Waymo
- Thread starter Daniel in SD
- Start date
EVNow
Well-Known Member
Depends on whether the Hilton model or the Air BNB model succeeds. I think they can live side-by-side.
There are interesting parallels. We'll see how things go when these robotaxi services really start rolling-out.
Mobility options have greatly increased the last few years. I see bicycles, electric bicycles, and scooters at a lot of places. Ride sharing services seem to have pushed-out taxis to the margins judging from what I see at mass transit centers like airports and train stations.
I think it's going to hinge on how individual owner/operators will get access to consumer $. In order to thrive (vs survive) I think they'll need to join a service (pay $) and need to spend more $$ (eg, advertise, get preference) or stay independent and park nearby and hope someone comes by.
Mobility options have greatly increased the last few years. I see bicycles, electric bicycles, and scooters at a lot of places. Ride sharing services seem to have pushed-out taxis to the margins judging from what I see at mass transit centers like airports and train stations.
I think it's going to hinge on how individual owner/operators will get access to consumer $. In order to thrive (vs survive) I think they'll need to join a service (pay $) and need to spend more $$ (eg, advertise, get preference) or stay independent and park nearby and hope someone comes by.
diplomat33
Average guy who loves autonomous vehicles
The ride had a safety driver out of an abundance of caution because of rain. In good weather, there is no safety driver.
We also see there is a driver facing camera system to make sure the safety driver is paying attention.
We do see that at one point, the rain picks up and an alert about weather appears on the screen. Rider support calls to check in and the safety driver disengages. Later, we see a check mark on the screen which the safety driver says means the car is able to re-engage autonomous mode but the safety driver finishes the ride in manual.
I wonder why they didn't just stay in autonomous mode and let the safety driver disengage only if the car had run into a problem. I mean, why disengage completely and finish the entire ride in manual even after the autonomous mode was ready to re-engage? Perhaps that was the choice of the safety driver? Maybe in case something unexpected happen that the safety driver is not able to intervene in time?
I think because of the customer, Waymo does not want to take any risks. So if the system detects even the smallest degradation in perception because of rain, they pull over or disengage to play it safe. I do understand why Waymo wants to play it safe. In autonomous mode, Waymo is liable. So if something had happened, Waymo could get sued hard, especially if the customer was injured even in the slightest. Sadly, it's the world we live in.
But the screen indicated that the system was ready to re-engage autonomous mode and it seemed to show good perception and path finding in the rain. I get that they want to be safe with riding customers, but I think Waymo's FSD could handle the rain just fine IMO. I don't think they needed to disengage in this scenario.
I trust when the system is even better and they have data that it meets their safety criteria, they will leave the autonomous mode on in bad weather.
Last edited:
EVNow
Well-Known Member
Haha ... now I know why Waymo doesn't offer their service in Seattle area. 300 days a year they would be shut down !
diplomat33
Average guy who loves autonomous vehicles
I am sure when they get the system confidence higher in rain, they will offer autonomous rides in Seattle.
I think we should keep in mind that the autonomous driving did seem to be able to handle the rain just fine. The disengagement was not because the autonomous driving failed but because of Waymo's safety rules.
EVNow
Well-Known Member
Yes - the question is whether we'll get "FSD" in Tesla before that or not. I think I have a bet with someone that FSD in Tesla will come before Waymo service in our area.
Waymo has to handle very hilly terrain and rain (and snow too, though only a few days a year). Likely we'll see the service in San Jose before Seattle and San Francisco.
bd7349
Member
Looks like Waymo really can't handle rain all that well. Disengages completely at around 7:22 in what I'd call light to moderate rain (the wipers weren't going fast at all).
diplomat33
Average guy who loves autonomous vehicles
I disagree. I think Waymo handled the rain well. If you watch the whole video, the safety driver disengages but I did not see anything wrong from the car. The car's autonomous mode never made any mistake. It never lost track of the lanes or anything. The car had good perception and path-finding the whole time and the screen actually indicated that autonomous mode was ready to engage.
Driver doesn't move but the car says "Needs to stop" then goes to manual.. Tesla AP handles same type of rain with no disengagements..
EVNow
Well-Known Member
I can confidently say, Tesla EAP (not even FSD) has better perception than I do - at night, when its raining with glare from cars driving in the opposite direction. In Seattle area this is probably 50% of my driving in a year. A lot of people with glasses probably feel that way.
EVNow
Well-Known Member
mikes_fsd
Banned
What is your point?
Dashcam image quality is compressed for storage.
The FSD computer gets the camera feed in (raw) full resolution - this has been "discussed" plenty of times before.
diplomat33
Average guy who loves autonomous vehicles
Sure, it's not bad. I've also seen videos of FSD Beta missing turns and almost hitting incoming cars, hitting curbs, almost hitting pedestrians and driving in the wrong lane to make a turn. Tesla still has a lot of work to do.
mikes_fsd
Banned
Tesla is the only one with such advanced features in end-user's hands.
Yet, the Waymo Preacher can't seem to see the improvements in these 2 months.
Right, even with the lower quality I can see that the lower image has an obstacle (I can even recognize what it is!).
These are images captured from the vehicle involved in the fatal collision in March 2019 (Final Reports for 2 Advanced Driver Assistance System Crash Investigations Published)
I think it's silly to say that Tesla's perception is better than a human. It's very good at seeing lane lines but I still don't think it's better than a human even at that.
Similar threads
- Article
