-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Autonomous Vehicle Progress
- Thread starter Tam
- Start date
-
- Tags
- Autonomous Vehicles
Your mistake is making it a waymo vs Tesla argument. It’s weak thinking at best.
“If you don’t believe waymo is the best then you must believe tesla is best”
Which of course leads to “If Tesla isn’t best then waymo is best”
And of course the grand daddy of internet arguments “the litany of articles that when read all say the same thing”. Which in this case is “waymo is working hard on this and has racked up some wins but they aren’t there yet”
Where Tesla is or isn’t has nothing to do with waymo and vice versa.
“If you don’t believe waymo is the best then you must believe tesla is best”
Which of course leads to “If Tesla isn’t best then waymo is best”
And of course the grand daddy of internet arguments “the litany of articles that when read all say the same thing”. Which in this case is “waymo is working hard on this and has racked up some wins but they aren’t there yet”
Where Tesla is or isn’t has nothing to do with waymo and vice versa.
Yep.
Waymo and GM still lead the pack in California’s new self-driving report cards has a summary and a pie chart of autonomous miles driven on CA public roads. For those who haven't looked at the reports, I'll summarize Tesla's for you. Again, this all applies only to autonomous driving on CA public roads.
Tesla:
2017: 0 autonomous miles driven (https://www.dmv.ca.gov/portal/wcm/connect/f965670d-6c03-46a9-9109-0c187adebbf2/Tesla.pdf?MOD=AJPERES)
2016: 550 miles autonomous miles driven all happening in 2 months out of 12, over 150 disengagements (https://www.dmv.ca.gov/portal/wcm/c...a/Tesla_disengage_report_2016.pdf?MOD=AJPERES)
2015: 0 disengagements but also no report in # of CA autonomous miles driven either (https://www.dmv.ca.gov/portal/wcm/c...67/tesla_disengagement_report.pdf?MOD=AJPERES)
Waymo/Google:
2017: 352,544 autonomous driven, 63 disengagements (https://www.dmv.ca.gov/portal/wcm/c...5-a72a-97f6f24b23cc/Waymofull.pdf?MOD=AJPERES, page 4)
2016: 635,868 miles, 124 disengagements https://www.dmv.ca.gov/portal/wcm/c...toWaymo_disengage_report_2016.pdf?MOD=AJPERES, page 3
2015: 424,331 miles, 272 disengagements https://www.dmv.ca.gov/portal/wcm/c...5/google_disengagement_report.pdf?MOD=AJPERES, page 4
Google's been at this for a long time.
Since Tesla is a company about hype, if they had something that was up to snuff to do that above, wouldn't they have done this by now?
Last edited:
What is there to disagree with? Waymo isn't there yet as evidenced by their complete lack of broad consumer release and they aren't anywhere near one. Neither is Tesla.
This isn't a zero sum game. Their independent progress has no bearing on each other. Continuing to make it a Tesla vs Waymo argument is a complete waste of time.
This isn't a zero sum game. Their independent progress has no bearing on each other. Continuing to make it a Tesla vs Waymo argument is a complete waste of time.
They'd better get a move on, before I start slipping up the other side of the bathtub curve...
xkcd: Fatal Crash Rate
xkcd: Fatal Crash Rate
"Complete lack of broad consumer release" is neither a necessary nor sufficient condition for proof of having a working product. You're conjuring that requirement out of thin air.
Waymo Filings Give New Details on Its Driverless Taxis
Your bizarre argument amounts to nothing more than "if I can't buy it, then they aren't close to having a useful product, because to be useful means I can buy it".
Is it a threat to your stock holdings? Is that your motivation?
mspohr
Well-Known Member
Top Toyota expert throws cold water on the driverless car hype
Toyota believes truly driverless cars — autonomous vehicles with nobody behind the wheel — are a long way away, possibly decades.
They seem to base their argument on the supposition that AVs need to be "several orders of magnitude" better than human drivers.
Is this just NIH?
Toyota believes truly driverless cars — autonomous vehicles with nobody behind the wheel — are a long way away, possibly decades.
They seem to base their argument on the supposition that AVs need to be "several orders of magnitude" better than human drivers.
Is this just NIH?
voyager
Active Member
Is there a consensus among these companies regarding the technology that needs to be deployed??
In other words, would these companies put roughly the same sort of detecting stuff on a given vehicle, on the same spots?
-GPS for vehicle positioning
-Lidar (and radars) for 3D 360 degree imaging/mapping the vehicle's surroundings
-Cameras for seeing traffic signals and to 'complete the picture'
-Sensors to sense nearby objects and own car measurements
-Computer for processing all information
Correct me if I am wrong.
In other words, would these companies put roughly the same sort of detecting stuff on a given vehicle, on the same spots?
-GPS for vehicle positioning
-Lidar (and radars) for 3D 360 degree imaging/mapping the vehicle's surroundings
-Cameras for seeing traffic signals and to 'complete the picture'
-Sensors to sense nearby objects and own car measurements
-Computer for processing all information
Correct me if I am wrong.
Last edited:
There are two basic schools of thought.
1. Humans have two eyes and a few other sensors, so a bunch of cameras should work for AI.
2. AI and available sensors are more limited than a human so make up for it with more sensors and non-AI processing.
1. Humans have two eyes and a few other sensors, so a bunch of cameras should work for AI.
2. AI and available sensors are more limited than a human so make up for it with more sensors and non-AI processing.
mspohr
Well-Known Member
I think the problem is not the sensirs but the fact that the brain is a much better processor. Once AI catches up to the brain for image processing,it will be as good.
voyager
Active Member
Did you just hear about the nullification of most (53 out of 56 claims) of Waymo Lidar patent #936 after being exposed by a lone random engineer Eric Swildens?
If Uber had hired him, they wouldn't have to pay Waymo $245 million for patent infringement.
Swildens discovered the implausibility of Waymo Lidar patent claims when he saw a ground wire connecting to a circuit would have current flowing both ways. (I guess it is just like a lightbulb only has a ground wire without a hot wire and still lights up because the design shows both + and - current flow both ways through one single ground wire.)
Looks like Velodyne patent # 558 was also adopted by Waymo as Waymo stopped buying from Velodyne and bragged that it could do its very own design/patent in-house to cut the cost down to 90% from Velodyne's price of $75,000.
Would anyone care to comment on the implication of the story?
It sounds like it's still very hard to work on an autonomous system like invent their own LIDAR system without patent infringement or fake circuit.
If Velodyne sues Waymo then the 90% price reduction might not be achieved any time soon
It's just a small bump along the way.
If Uber had hired him, they wouldn't have to pay Waymo $245 million for patent infringement.
Swildens discovered the implausibility of Waymo Lidar patent claims when he saw a ground wire connecting to a circuit would have current flowing both ways. (I guess it is just like a lightbulb only has a ground wire without a hot wire and still lights up because the design shows both + and - current flow both ways through one single ground wire.)
Looks like Velodyne patent # 558 was also adopted by Waymo as Waymo stopped buying from Velodyne and bragged that it could do its very own design/patent in-house to cut the cost down to 90% from Velodyne's price of $75,000.
Would anyone care to comment on the implication of the story?
It sounds like it's still very hard to work on an autonomous system like invent their own LIDAR system without patent infringement or fake circuit.
If Velodyne sues Waymo then the 90% price reduction might not be achieved any time soon
It's just a small bump along the way.
In June, Waymo safety driver fell asleep and accidentally pressed on the accelerator which disabled the autonomous system and crashed into the highway median in Mountain View, CA. The crash woke the driver up and the driver drove it back to the office successfully.
Maybe Waymo should implement commercial Autopilot protocol by having at least 2 human drivers instead of a lone driver so they can wake each other up every few minutes.
Maybe Waymo should implement commercial Autopilot protocol by having at least 2 human drivers instead of a lone driver so they can wake each other up every few minutes.
We approach the end of 2019 and autonomous vehicles are already here. The technology is now available.
How Reality Has Exceeded Our Expectations About Self-Driving Vehicles
How Reality Has Exceeded Our Expectations About Self-Driving Vehicles
Similar threads
