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

Research on autonomous cars

This site may earn commission on affiliate links.
Hi everyone! I am currently reseraching which characteristics drivers would like to have in autonomous vehicles. It would really help if you could fill this quick survey!! At the end of the project I will share the results with you!
Thank you very much!!

Autonomous Driving - User Control Options
There's no need for the survey as the results don't matter.

The first self driving cars will all be about safety, anything to compromise that safety will be frowned upon. Therefore, they aren't even really options. These next few years are going to be all about proving the technology to the public.
 
There's no need for the survey as the results don't matter.

The first self driving cars will all be about safety, anything to compromise that safety will be frowned upon. Therefore, they aren't even really options. These next few years are going to be all about proving the technology to the public.
While I agree that the first self driving cars will focus mainly on safety to get the users to trust them, I do not think they will be willing to let go their freedom. This is what the survey is about.
 
While I agree that the first self driving cars will focus mainly on safety to get the users to trust them, I do not think they will be willing to let go their freedom. This is what the survey is about.
Elon is a big proponent of allowing you to drive your own car. Although you'll have the ability to turn on the autonomy features on a safe, self-driving car, you also have the ability to manually drive it anyway you see fit.

These two things must NOT fade into one another for the public good. A user should not be allowed to legally make a self driving car self-drive in a less safe manner. It defeats the entire purpose.
 
Elon is a big proponent of allowing you to drive your own car. Although you'll have the ability to turn on the autonomy features on a safe, self-driving car, you also have the ability to manually drive it anyway you see fit.

These two things must NOT fade into one another for the public good. A user should not be allowed to legally make a self driving car self-drive in a less safe manner. It defeats the entire purpose.

The premise of this research is not to question the need for safety in an autonomous vehicle. The technology a person uses defines their behaviour and reflects their personality to some extent. We would like to consider that as autonomous vehicles get popularised in the future, people will be able to maintain their characteristics through technology, instead of learning to live with the given systems.
We are thinking a step ahead, to level automation 5 (total automation), not to level 3.
 
The premise of this research is not to question the need for safety in an autonomous vehicle. The technology a person uses defines their behaviour and reflects their personality to some extent. We would like to consider that as autonomous vehicles get popularised in the future, people will be able to maintain their characteristics through technology, instead of learning to live with the given systems.
We are thinking a step ahead, to level automation 5 (total automation), not to level 3.

The general consensus in industry is to skip SAE level 3 and go straight into levels 4-5.
At least all the major auto manufacturers and Tesla feel this way, including non-auto manufacturers like Google, Apple, Otto/Uber, etc.

If we assume the survey covers SAE level 2 like the current Tesla Autopilot or even the future Enhanced Autopilot then it becomes quite relevant. However, you'd need to remove the word "autonomous" because at that point it's not level 5 self-driving or when people refer to "fully autonomous".

Survey: Advanced Driver Assistance Systems (ADAS) - User Control Options
 
Last edited:
The general consensus in industry is to skip SAE level 3 and go straight into levels 4-5.
At least all the major auto manufacturers and Tesla feel this way, including non-auto manufacturers like Google, Apple, Otto/Uber, etc.

If we assume the survey covers SAE level 2 like the current Tesla Autopilot or even the future Enhanced Autopilot then it becomes quite relevant. However, you'd need to remove the word "autonomous" because at that point it's not level 5 self-driving or when people refer to "fully autonomous".

Survey: Advanced Driver Assistance Systems (ADAS) - User Control Options

Jeff I think you are not understanding. The survey covers SAE level 5. We are trying to understand to what extent people are willing to embrace a technology that limits their freedom. This together with interviews to experts in the field of automotive and control systems will allow us to come up with a research paper on this topic.
 
Jeff I think you are not understanding. The survey covers SAE level 5. We are trying to understand to what extent people are willing to embrace a technology that limits their freedom. This together with interviews to experts in the field of automotive and control systems will allow us to come up with a research paper on this topic.

"to what extent people are willing to embrace a technology that limits their freedom" is very different than the questions posed in the survey which explicitly ask how would use use the technology if it allowed such freedoms... We already know it will not and should not have such freedoms.
Example:
Which of the following driving styles options would you like the self-driving car to have as a feature?
  • Aggressive Driving Style (overspeeding, tailgating)
Pointless. It doesn't matter what you'd like because it's going to be limited by what the manufacturer deems fit and regulators deem fit. It may allow modest yet still safe options but nothing like a human really drives and for good reason. If a 4000 lb vehicle is tailgating, computer controlled or not, I promise it's stopping distance will be greater than a light weight vehicle in an emergency situation and will therefore end up slamming into the front vehicle.

Here's what you should be asking:

Would you be willing to give up ALL control to an autonomous vehicle and ride as a passenger to your destination?
  • Yes
  • No
If they answer no then you've pretty much eliminated SAE 4 and 5.

That said, the manufacturer can choose to add some different driving styles, but your "freedoms" are still limited by available choices. Perhaps you can call it "perceived freedom"

Before coming up with a research paper on the topic you'd also need to do some research into the background of your topic:

http://ais.informatik.uni-freiburg.de/publications/papers/kuderer15icra.pdf

Autonomous Cars Could Determine Your Driving Style by Gently Probing You

https://pdfs.semanticscholar.org/bb4c/3821b197b11fc4f818a7538e8bff5484612a.pdf

https://mediatum.ub.tum.de/doc/1294967/1294967.pdf

http://www.auto-ui.org/15/p/workshops/4/Comfort Determination in Autonomous Driving Style.pdf

This is a widely researched topic. Look specifically at the last link, it's a proposal... was this experiment ever performed?
 
"to what extent people are willing to embrace a technology that limits their freedom" is very different than the questions posed in the survey which explicitly ask how would use use the technology if it allowed such freedoms... We already know it will not and should not have such freedoms.
Example:

Pointless. It doesn't matter what you'd like because it's going to be limited by what the manufacturer deems fit and regulators deem fit. It may allow modest yet still safe options but nothing like a human really drives and for good reason. If a 4000 lb vehicle is tailgating, computer controlled or not, I promise it's stopping distance will be greater than a light weight vehicle in an emergency situation and will therefore end up slamming into the front vehicle.

Here's what you should be asking:

Would you be willing to give up ALL control to an autonomous vehicle and ride as a passenger to your destination?
  • Yes
  • No
If they answer no then you've pretty much eliminated SAE 4 and 5.

That said, the manufacturer can choose to add some different driving styles, but your "freedoms" are still limited by available choices. Perhaps you can call it "perceived freedom"

Before coming up with a research paper on the topic you'd also need to do some research into the background of your topic:

http://ais.informatik.uni-freiburg.de/publications/papers/kuderer15icra.pdf

Autonomous Cars Could Determine Your Driving Style by Gently Probing You

https://pdfs.semanticscholar.org/bb4c/3821b197b11fc4f818a7538e8bff5484612a.pdf

https://mediatum.ub.tum.de/doc/1294967/1294967.pdf

http://www.auto-ui.org/15/p/workshops/4/Comfort Determination in Autonomous Driving Style.pdf

This is a widely researched topic. Look specifically at the last link, it's a proposal... was this experiment ever performed?


JeffK, I honestly do not understand all your negativity and critiques. I asked for your help to fill the survey, not to improve or change it.
The survey is a small part of our research and I don't expect you to understand the meaning its results are going to have.
If you have constructive advices they are welcome, but I do not accept your pointless and exaggerated criticism.
 
All messages above are very constructive.

Your tone and reactions to above advice are not social norms (either in an academic setting or internet.)

Honestly they did not seem constructive at all, rather disruptive and with a sense of superiority.
And the assumption that no further research on the topic has been done before coming up with the survey proves my point.
 
Honestly they did not seem constructive at all, rather disruptive and with a sense of superiority.
And the assumption that no further research on the topic has been done before coming up with the survey proves my point.
I'm trying to help here. I've seen these surveys, news articles, and research papers go down rabbit holes attempting to prove common sense items wasting both time and money without actually advancing academia.

This particular survey isn't asking anything new and the results of the survey can be derived from previous works:
http://bdd.berkeley.edu/system/files/project/document/Report_CarMotion.pdf
For autonomous cars, driving dumb is key - SlashGear

The general theme I'm seeing is that people want and probably need a level of customization to feel "comfortable". If, as an aggressive driver, you've ever been in a car with someone who brakes for everything, accelerates super slowly, and only goes the speed limit, or leads (limiting) an entire caravan of cars while there is no one front of them, it's really aggravating. You want to strangle them and take the wheel. Likewise, if you get in the car with someone more aggressive than yourself you might feel unsafe.

Many groups are suggesting that the basics be set by the manufacturer and to adapt within some bounds to a driving style learned from the user themselves. This would allow a driving style the user is comfortable with while at the same time remaining safe.

This might also mean that a car might have more than one user profile if it has more than one regular driver....

We know that Tesla gave certain options with AP1 such as speed to a degree and how close to follow another vehicle. We don't yet know if self driving AP2 vehicles will have the ability to be "personalized" (or other manufacturers for that matter).
 
the assumption that no further research on the topic has been done before coming up with the survey proves my point.

For example, if I want to do a research to prove that white car cabin temperature is much cooler than a black car is because of the different degrees in reflection.

Then someone says there's no need because there's no difference with modern Air Conditioning system.

I may feel offended for such a superiority claim but for the sake of science, I would ask for evidences rather than accusing "disruptive."
 
There is something fundamentally paradoxical about self-driving... Will we be left the pleasure of 'actual driving' only in a virtual world?

maxresdefault.jpg