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

Will total autonomy be good?

This site may earn commission on affiliate links.
I am somewhat amused at how many people think total autonomy is right around the corner. Some even seem to feel cheated because their Tesla isn't totally autonomous already. The nerve of Tesla to actually expect them to hold the steering wheel while driving within a few feet of drivers of totally unknown skills, experience, and mental faculty!

I think total autonomy will be tremendous when virtually all vehicles are on line. Until then, I wonder. It's not always about staying between the lines and reading signs and lights. Here are just a couple simple examples of situations I have encountered on a number of occasions which I think could be troubling to a totally autonomous vehicle. I am sure you can think of more, and I would love to see your posts in that regard.

1. I find that often I am trying to enter a steady stream of slow moving traffic. The way I finally get in is by making eye contact with the next driver, and they wave or nod to me to let me know I can proceed without their conflicting with me. How will a totally autonomous car handle those situations?

2. I have also avoided accidents by observing another driver who is obviously distracted by cell phones, kids, applying makeup, etc. I suspect we have all seen these situations and steered clear when we realized they were not going to yield, or stop appropriately.

3. As simple a thing as auto park. My Model S is pretty smart. It looks for an open space, even in a crowded parking lot, that it can maneuver into, and then can handle the parking itself. I am even smarter. I can see someone getting into their car, and I know that space will open soon, so I can just sit and wait. I can even assume that the guy with one small bag will be leaving faster than will the lady with a cart full of groceries and two little children.

These are just a few situations that I wonder how the wizards will have autonomy handle. I am sure they will amaze me, but I do wonder how. Once all cars, or all cars in restricted area are autonomous, they can communicate with each other, making these situations non-issues, but until then, I think it will be quite interesting. Perhaps some of you more advanced tech thinkers can illuminate the process for us.

What situations do you see as problems, and how anxious will you be to get into a car without a steering wheel, on public roads shared by 16 year olds in Mustangs, and 85 year olds in Cadillacs,,,,and deer,,,,,and things which have fallen from truck beds,,,,,,and local flooding,,,,,and heaven forbid,,,,,,ME!

Drive Safe
 
All good points, but:

1. car-to-car COMMs will solve that - all traffic will merge at warp speed!

2. They will all be in the back seat, in autonomous cars.

3. car-to-car COMMs again ... although ... I read that once autonomous arrives it will be cheaper (electricity cost) to drive, slowly, than to pay to park, so you car will just cruise around the streets, clogging them up, whilst you are at work 9-5, so no-more-carparks :D
 
I am somewhat amused at how many people think total autonomy is right around the corner. Some even seem to feel cheated because their Tesla isn't totally autonomous already. The nerve of Tesla to actually expect them to hold the steering wheel while driving within a few feet of drivers of totally unknown skills, experience, and mental faculty!

I think total autonomy will be tremendous when virtually all vehicles are on line. Until then, I wonder. It's not always about staying between the lines and reading signs and lights. Here are just a couple simple examples of situations I have encountered on a number of occasions which I think could be troubling to a totally autonomous vehicle. I am sure you can think of more, and I would love to see your posts in that regard.

1. I find that often I am trying to enter a steady stream of slow moving traffic. The way I finally get in is by making eye contact with the next driver, and they wave or nod to me to let me know I can proceed without their conflicting with me. How will a totally autonomous car handle those situations?

2. I have also avoided accidents by observing another driver who is obviously distracted by cell phones, kids, applying makeup, etc. I suspect we have all seen these situations and steered clear when we realized they were not going to yield, or stop appropriately.

3. As simple a thing as auto park. My Model S is pretty smart. It looks for an open space, even in a crowded parking lot, that it can maneuver into, and then can handle the parking itself. I am even smarter. I can see someone getting into their car, and I know that space will open soon, so I can just sit and wait. I can even assume that the guy with one small bag will be leaving faster than will the lady with a cart full of groceries and two little children.

These are just a few situations that I wonder how the wizards will have autonomy handle. I am sure they will amaze me, but I do wonder how. Once all cars, or all cars in restricted area are autonomous, they can communicate with each other, making these situations non-issues, but until then, I think it will be quite interesting. Perhaps some of you more advanced tech thinkers can illuminate the process for us.

What situations do you see as problems, and how anxious will you be to get into a car without a steering wheel, on public roads shared by 16 year olds in Mustangs, and 85 year olds in Cadillacs,,,,and deer,,,,,and things which have fallen from truck beds,,,,,,and local flooding,,,,,and heaven forbid,,,,,,ME!

Drive Safe
I think you raise two separate issues.

The first is whether a totally autonomous vehicle can be achieved before ALL vehicles have car-to-car comm (I'm less concerned than the OP about random obstructions). Inter-vehicle comm is accomplished today with observation of other vehicles, glances, waves and some amazingly subtle communication between drivers, drivers and pedestrians, drivers and bicycles, etc. I think it will be impossible, or very dangerous, to have total autonomy without 100% comm between vehicles and perhaps some sort of transponder on non-vehicles. But governments operating the way they do, it will take a pile of dead bodies before they back off the rush to autonomy. The German legislature is considering banning the sale of new ICE vehicles after 2030, expecting that to take the majority of ICE vehicles off the road by 2050. A similar time frame is probably realistic for inter-vehicle communication.

The second point is whether we actually want total autonomy. On that point I say, yes, though it will probably arrive after I'm long gone. I've always been a car guy and at times still enjoy driving. But much of the driving I do these days is not at all enjoyable. Last night it took me 40 minutes to travel 10 miles, 5 in city traffic) to see a concert. I received no joy from driving those miles and would gladly delegate it to some form of automation. Not to mention I would also have been free to go out after the concert and have a drink or two.
 
I agree on the desirability of total autonomy. At some point, the only people who drive cars will do so by desire, much like people who ride horses today. This will probably take place in areas out of the main flow of traffic. Perhaps only on closed tracks reserved for such activity. When that day arrives, travel times, and economies will be amazingly advanced compared to today's model.

But I still wonder how total autonomy will mix with non communicative, less intelligent systems.
 
I still wonder how total autonomy will mix with non communicative, less intelligent systems

Any chance of a (compulsory) retro-fit of something that broadcasts information about the (old) car? I'm struggling to think of how useful that data would be - it won't include the "wave or nod to me to let me know I can proceed" but might help with "another driver who is obviously distracted by cell phones, kids, applying makeup" by broadcasting their road position, which a computer might detect is erratic before some sudden move actually takes place.

Is "current speed and heading" useful without "future speed and heading" as well?
 
  • Like
Reactions: GSP
3. As simple a thing as auto park. My Model S is pretty smart. It looks for an open space, even in a crowded parking lot, that it can maneuver into, and then can handle the parking itself. I am even smarter. I can see someone getting into their car, and I know that space will open soon, so I can just sit and wait. I can even assume that the guy with one small bag will be leaving faster than will the lady with a cart full of groceries and two little children.

I think parking is one of the situation where autonomy will be great. The car will drop you off at the entrance and then go look for a parking space. That's the kind of situation that doesn't matter if it takes 1 minute or 10 minutes to find a place ;-)
 
  • Like
Reactions: callmesam
Will autonomous cars be able to mix with manually driven cars? This is an important question. I don't know the answer, but I did have an encouraging thought that leads me to believe that success is one of the possible outcomes.

In the factory where I work, the Manufacturing Engineers installed a new production line for a new product circa 1990. It included Automated Guided Vehicles (AGVs) to carry parts from one machining station to the next. The area where the AGVs operated was fenced off and only personnel that were trained to avoid the AGVs were allowed in.

Ten years later, it was time to install another production line for another new product. AGVs were used again. This time the AGVs carried product on the regular aisle-ways used by all personnel, visitors, forktrucks, tricycles, electric scooters, etc. The technology had progressed enough that while moving at a fast walking speed the AGVs could detect people and vehicles, and come to a safe stop when needed.

Will the same thing be possible at highway speeds? I don't know, but I am hopeful that it will be.

GSP
 
Good point about the Google cars. They have their limitations, but they have been driving autonomously in normal traffic for years, with a very low accident rate.

GSP
 
Being able to share the road with non-autonomous cars is an absolute requirement. To do otherwise will delay implementation pretty much forever. People won't be willing to pay for an autonomous driving system they may never be able to use because their car will go to the junkyard before every car on the road has a similar system.

The early adopters will be willing to wait, mostly impatiently, for the software to arrive for their cars with full autonomous hardware, but the general public isn't going to buy a car with hardware they can't use immediately.

I just took my first long road trip in my Model S. I used AP on long, mostly straight stretches, but also went off AP in a lot of situations where I didn't trust the system. I was observing all the various road conditions that a full autonomous system needs to contend with. It's not an easy problem, even without dealing with other vehicles.

Add other vehicles into the mix and it does get vastly more complex. There were two instances where trucks pulled into my lane without warning and I had to slow down suddenly to prevent an accident, one time screeching the tires. AP was on in both instances and didn't react.
 
I think Google cars are doing it. I would love to hear from someone who has ridden in one of the "driverless" taxis. It would be interesting to see how they liked the ride, and if the "standby driver" ever had to take over, not so much for safety as expediency.