Au contraire! It nicely summarizes a common theme in those posts. And it is factual, not an ad hominem post in the least.
charlie hebdo Au contraire! It nicely summarizes a common theme in those posts. And it is factual, not an ad hominem post in the least.
I guess that settles it, then.
If GM "killed the electric car", what am I doing standing next to an EV-1, a half a block from the WSOR tracks?
Guess you people need to look up what "ad hominem" actually means.
aegrotatio Guess you people need to look up what "ad hominem" actually means.
Isn't "you people" one of those bad phrases?
I'm not thinking self-driving is coming soon. I've seen what a Tesla model Y can do. Impressive, but not anywhere near as savvy as even a 20%ile driver - and then only under ideal conditions. I can't get Alexa to do what I ask about half the time unless I phrase things exactly the right way.
These things as a backup or aid? Sure. On their own while we sleep? Don't hold your breath.
-Don (Random stuff, mostly about trains - what else? http://blerfblog.blogspot.com/)
It means "of an argument or reaction directed against a person rather than the position they are maintaining."
There were no personal attacks on him, just disagreements with what he stated.
This is what Greg Posted:
"ttrraaffiicc has a gloomy chrystal ball. listing of your posts
Others may have made comments that were less specific. What did you see as personal?
oltmannd I'm not thinking self-driving is coming soon. I've seen what a Tesla model Y can do. Impressive, but not anywhere near as savvy as even a 20%ile driver - and then only under ideal conditions. I can't get Alexa to do what I ask about half the time unless I phrase things exactly the right way. These things as a backup or aid? Sure. On their own while we sleep? Don't hold your breath.
Me: "Alexa, can you drive a car?"
Alexa: "That's not one of the things I can do."
Me: "Damn straight!"
oltmanndI'm not thinking self-driving is coming soon. I've seen what a Tesla model Y can do. Impressive, but not anywhere near as savvy as even a 20%ile driver - and then only under ideal conditions. I can't get Alexa to do what I ask about half the time unless I phrase things exactly the right way. These things as a backup or aid? Sure. On their own while we sleep? Don't hold your breath.
I agree. As impressive as the advancements have been, there is still a ways to go.
I watched a program recently about a company that makes the software for cars to recognize pedestrians standing along the road, and if the pedestrians will start to cross the road in front of the car. Basically, the software has to be taught to recognize every single "look" of what a pedestrian might look like. There are still issues.
I have no doubt one day in the future, people will be looking back at us primitive people who actually had to drive a car ourselves.
But I don't think that day is anywhere near. Since I'm old, I imagine it will be long after I quit thinking (or breathing).
York1 John
York1 oltmannd I'm not thinking self-driving is coming soon. I've seen what a Tesla model Y can do. Impressive, but not anywhere near as savvy as even a 20%ile driver - and then only under ideal conditions. I can't get Alexa to do what I ask about half the time unless I phrase things exactly the right way. These things as a backup or aid? Sure. On their own while we sleep? Don't hold your breath. I agree. As impressive as the advancements have been, there is still a ways to go. I watched a program recently about a company that makes the software for cars to recognize pedestrians standing along the road, and if the pedestrians will start to cross the road in front of the car. Basically, the software has to be taught to recognize every single "look" of what a pedestrian might look like. There are still issues. I have no doubt one day in the future, people will be looking back at us primitive people who actually had to drive a car ourselves. But I don't think that day is anywhere near. Since I'm old, I imagine it will be long after I quit thinking (or breathing).
A few more...
Can it tell a pot hole and avoid it depending on deep it is? Will it remember where the pot hole is and recognize it getting worse or repaired?
Can it keep out of ruts in the lane to reduce risk of hydroplaning in the rain? Will it adjust position in the lane to avoid worn/rough/patched area? Will it adjust position in the lane to add extra cushion against wildlife on the side of the road?
Teslas can read road signs and adjust speed accordingly, but they don't recognize when school zone limits are not in effect.
Can it follow temporary lane markers when the originals are poorly marked out in work zones? Can it decipher a flagman?
Can it anticipate a large truck making a wide swing on a curve at an urban intersection?
Anybody got any others?
Don:
I'm still working on pot-hole recognition.
Along with plastic garbage bin detection when backing from my driveway.
Yeah, the "look" from a pedestrian.
I was in Japan visiting Advanced Telecommunications Research (ATR), which is in a rural part of Japan where people drive -- real fast.
My instinctive crossing drill "look left, look right, look left, step out" almost got me killed more than once.
Lefthand side-of-the-road driving.
Can it steer through a skid?
GrampCan it steer through a skid?
Keep in mind that a much more complex skid-control scenario, for OTR trucks, has been under development a very long time. Multiple-trailer setups that do controlled backup do it in part with controlled differential brakes on both bogie and trailer wheels -- this can also address 'wind dance' for multiples that may build up harmonic oscillations if uncorrected... or that have the wrong interaction between tires and road imperfections. All these things can be continually monitored and reacted to, with very short latency, very short error correction, but high 'physics prediction' based on other factors, ensuring the best possible mix of quick and correct action. Skids of various kinds require more complex response, but no faster and not involving dramatic increases in technological cost to implement.
What a human driver can do, for example, with trailer brakes in a tractor skid is greatly limited by comparison.
Of course there will still be plenty of times accidents would be unavoidable to 'recover' from -- by either human or robot driving. This opens up another concern: anticipation of potential issues, and slowing/redirecting/stopping etc. to keep within the stability envelope. These were important topics in guided-vehicle design over 70 years ago, and still are part of sensible system design.
Overmod Gramp Can it steer through a skid? For many years, better than a human driver. Keep in mind that a much more complex skid-control scenario, for OTR trucks, has been under development a very long time. Multiple-trailer setups that do controlled backup do it in part with controlled differential brakes on both bogie and trailer wheels -- this can also address 'wind dance' for multiples that may build up harmonic oscillations if uncorrected... or that have the wrong interaction between tires and road imperfections. All these things can be continually monitored and reacted to, with very short latency, very short error correction, but high 'physics prediction' based on other factors, ensuring the best possible mix of quick and correct action. Skids of various kinds require more complex response, but no faster and not involving dramatic increases in technological cost to implement. What a human driver can do, for example, with trailer brakes in a tractor skid is greatly limited by comparison. Of course there will still be plenty of times accidents would be unavoidable to 'recover' from -- by either human or robot driving. This opens up another concern: anticipation of potential issues, and slowing/redirecting/stopping etc. to keep within the stability envelope. These were important topics in guided-vehicle design over 70 years ago, and still are part of sensible system design.
Gramp Can it steer through a skid?
For many years, better than a human driver.
https://www.youtube.com/watch?v=kL_IzyZ9d2g
Never too old to have a happy childhood!
Daycab. I doubt he had a change of shorts with him...
BackshopI doubt he had a change of shorts with him...
And he obviously had the option to pull to the side at any point, although I sympathize with the intent to get him off there -- yes, I appreciate the reasons not to leave the truck in such a location. Why they did not call for trucks to drop salt or calcium chloride in front of him is another mystery.
A proper autonomous truck would have been able to use slightly more braking and antilock modulation on the rear trailer axle than on the front, which would in turn be more than the resultant of any engine and service braking on the tractor drive axle. That would eliminate any of the effect you see of lateral grade or superelevation sliding the rear trailer out. Here the driver evidently doesn't have a proper independent trailer-brake controller, let alone a way to modulate both trailers, the tractor brake, and the tractor steering all at once, let alone try to do so continuously with both prediction and situational learning. (In fact this would be a good application for artificially-intelligent antilock-braking control, but I doubt that would be marketed separate from 'autonomous' control at this point in time...)
Professional drivers: aren't chains mandatory for that location in that weather? And would they have helped or hindered on icing of that kind?
This has some implications for Templeton the rat's little sleeper cars. At what point do they 'hole up' for conditions like ice or storms? What happens if the available charge or shore-power facilities become overused or themselves start to fail due to conditions? Will there be the 'moral equivalent' of package boilers or mobile generators -- quite possibly involving properly-equipped locomotive consists -- that can be directed ahead of time to potentially affected areas?
Overmod Backshop I doubt he had a change of shorts with him... This is almost a non-event, except insofar as he shouldn't have been allowed on that grade with pups in known ice in the first place. At no time does he appear badly out of control (although of course that's no guarantee he might have progressed into worse in the next few seconds at any particular time we see) And he obviously had the option to pull to the side at any point, although I sympathize with the intent to get him off there -- yes, I appreciate the reasons not to leave the truck in such a location. Why they did not call for trucks to drop salt or calcium chloride in front of him is another mystery. A proper autonomous truck would have been able to use slightly more braking and antilock modulation on the rear trailer axle than on the front, which would in turn be more than the resultant of any engine and service braking on the tractor drive axle. That would eliminate any of the effect you see of lateral grade or superelevation sliding the rear trailer out. Here the driver evidently doesn't have a proper independent trailer-brake controller, let alone a way to modulate both trailers, the tractor brake, and the tractor steering all at once, let alone try to do so continuously with both prediction and situational learning. (In fact this would be a good application for artificially-intelligent antilock-braking control, but I doubt that would be marketed separate from 'autonomous' control at this point in time...) Professional drivers: aren't chains mandatory for that location in that weather? And would they have helped or hindered on icing of that kind? This has some implications for Templeton the rat's little sleeper cars. At what point do they 'hole up' for conditions like ice or storms? What happens if the available charge or shore-power facilities become overused or themselves start to fail due to conditions? Will there be the 'moral equivalent' of package boilers or mobile generators -- quite possibly involving properly-equipped locomotive consists -- that can be directed ahead of time to potentially affected areas?
Backshop I doubt he had a change of shorts with him...
This is almost a non-event, except insofar as he shouldn't have been allowed on that grade with pups in known ice in the first place. At no time does he appear badly out of control (although of course that's no guarantee he might have progressed into worse in the next few seconds at any particular time we see)
As he brought the truck to a stop - it looked like the right rear brakes on the rear trailer were locked and that was what was causing the loss of proper tracking on a low grip driving surface.
1. I doubt if there were salt trucks available to get ahead of the truck.
2. Chains are only for drive axles for traction.
3. It's only a "non event" if you're a Monday morning quarterback.
4. I'm sure he was doing everything in his power to keep the truck as straight as it was.
Backshop1. I doubt if there were salt trucks available to get ahead of the truck.
That is also the basis for calling it a 'nonevent' -- it was continued only because there was some understanding that getting the truck down off the grade under controlled observation in minimum time was considered important. Any actual 'danger' other than circumstantial could have been avoided with a controlled stop -- that being part of the skill being demonstrated by the driver in question precisely in keeping the sliding controlled at relatively low speed.
And there is no question the driver was doing 'everything in his power to keep the truck as straight as it was'. The concern is that the brake arrangement on the truck did not facilitate better control than he had, and there was a patent loss of stability as a result. A loss of stability that better methods of differential brake modulation and control, including those appropriate to autonomous trucks, would likely have experienced to a much less extreme degree. Which was, I believe, the major point of the discussion concerning the video and the event.
Our community is FREE to join. To participate you must either login or register for an account.