Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

My First Ride in a Self-Driving Car Was Harrowing as Hell

This site may earn commission on affiliate links.

Tam

Well-Known Member
Nov 25, 2012
13,301
12,340
California
fgfzjq3vkpe3kygm9tmw.gif



It was a CES demonstration of how Self-Driving Car could be as envisioned and combined by 6 companies Microsoft, NXP, IAV, Swiss Re, Esri, and Cubic Telecom all rolled in specially equipped Volkswagen Golf Estate.

Apparently, stopping for a stationary vehicle is still a challenge.

The driver still had to manually drive for at least 40 percent of this "Highly Automated Driving Smart, Safe And Secure" demonstration.

After reading this report, as well as Uber's attempt to test drive in CA without an Autonomous Vehicle Test Permit, these kinds of demonstrations are a disgrace!
 
  • Like
  • Informative
Reactions: Jaff and Andyw2100
Wow, I'm going to estimate the cross traffic at 160 fps? A pickup is 20 feet long.

Just looking at first I figured 100 mph, and then I tried to do the math.

I've never seen the traffic flow at 109 mph in Vegas.
 
It was a CES demonstration of how Self-Driving Car could be as envisioned and combined by 6 companies Microsoft, NXP, IAV, Swiss Re, Esri, and Cubic Telecom all rolled in specially equipped Volkswagen Golf Estate.

Apparently, stopping for a stationary vehicle is still a challenge.

The driver still had to manually drive for at least 40 percent of this "Highly Automated Driving Smart, Safe And Secure" demonstration.

After reading this report, as well as Uber's attempt to test drive in CA without an Autonomous Vehicle Test Permit, these kinds of demonstrations are a disgrace!


I agree that this demo was a disgrace, however, that doesn't mean that other approaches would not work better (or they might work worse) For example, from the write-up, this us trying to use a V2X communications system with the lights, which apparently wasn't working, as opposed to the Tesla (and Google's) looking at the lights themselves with multiple cameras (he mentioned "a camera"). I'm not saying that the multiple camera approach is better, just that it is different, so we cant use a demo of one to evaluate the other.

It's like saying that since early EV's looked bad and had low range, they all will look bad and have low range. Yup, they might, but on the other hand, different approaches will have different results.

The other demonstrations that I have seen videos of, from a number of manufacturers, seem to show a much more mature system that works better. That could be marketing and cleaned up videos, but even with that, this feels less mature.

As for Uber's attempt to test drive in CA, They're not applying for a permit had more to do with politics than with technical issues from what I read. They didn't think they needed one, then when told to get one, they decided to try to bull their way through. (They did have some technical issues, but nothing that indicates to me, from what I have read, that they are that far off.)

Also, over the years, I have given hundreds of demonstrations to customers and the public in various technical roles. A number of them went wrong, some horribly, that didn't mean that whatever I was demonstrating sucked (though sometimes it did ;) ), simply that some part of the demo wasn't working right.

So, what I am saying is, don't take this (or the Uber thing) as evidence that autonomous driving is doomed to be a long-long way off... include all of the available information, then keep an open mind. I guarantee that Uber doesn't think it is 10+ years off, they would not be spending time and money trying to make it work. And Google would not have started a company to try to sell their approach if they thought it was that far off either... and both of these companies have put lots more thought and research into it than I have. These guys could be wrong, but looking around, Tesla, Volvo, Ford, Nissan, Toyota, Hundai, Chrysler, BMW, Mercedes and others have all been talking about this now, and often showing examples or technology demos. (in fact, I can't think of anyone who isn't, though I am sure there is someone). Again, any of these guys could be wrong, even all of them, but I suspect that if it was impossible, we would be seeing a LOT less of it out there.
 
...that doesn't mean that other approaches would not work better...

True.

To me, it's fine to have a visions and aspirations.

As long as demonstration warns audience that it's not yet "Highly Automated Driving Smart, Safe And Secure" then it wouldn't be a disappointment.

It is still good to get a demonstration to see how well the system work from various companies and compare their progress.