ROBORace

[Andres125sx]

I did not see that Audi racing anything, there were no other autonomous cars near it.

No need to be autonomous I guess, but I´d say if technology is good enough to drive between humans with the consequent risk, it must be good enough to drive between driverless cars
http://www.lasexta.com/noticias/ciencia ... 00306.html

Can you show me a video of it doing the lap in the wet at racing speed and a comparison on lap times to a human driven car.

To what purpose?

Nobody said an autonomous car is faster than a human driver, at least not yet. But that´s exactly the reason for RoboRace, developing autonomous driving

[flynfrog]

Are you trying to say that we cant build a computer with more processing power than a fly? Your average cell phone has much more processing power than a fly brain.

We don't need to reach human levels of processing power to drive a car. The inputs are rather small compared to everything a human is processing. Sensor technology is far and away greater than what a human can do. I'm not sure where you think the issue is here. Or do we just need to get our shoes to throw into the machine.

Tell me what is processing power?
What is consciousness?

Processing power: processing power is the ability of a computer to manipulate data

Driving is no more than sensor input followed system output

Consciousness: the fact of awareness by the mind of itself and the world.
not needed for driving a car


Any point to your questions?

[mzivtins]

How can a computer be 'more safe' than a human, when the term 'safe' will simply be a collection of measurements and tolerances with actions applies.

How can it ever be said that a computer is safer when a computer simply cannot determine what is safe or unsafe outside of its boundaries of programming?

They aren't safer, not at all, which is why it is such a good idea to let them race! the only thing at risk, within reason, are the cars themselves! Awesome

[mzivtins]

Are you trying to say that we cant build a computer with more processing power than a fly? Your average cell phone has much more processing power than a fly brain.

But yet it takes teraflops to simulate the actions of a few brain cells... in a dumbed down fashion anyway:

The computer model, which is freely available to explore online, does not simulate every aspect of the cortex: for example, it does not include glia — the brain’s non-neuronal cells — or blood vessels, and leaves out more-complex aspects of neural circuitry, such as plasticity (how synaptic connections change in response to experience).

http://www.nature.com/news/fragment-of- ... er-1.18536

processing power is just the rate at which data can be processed, not manipulated, the higher the processing power (speed, thats all it is) the quicker these 'brains' can make decisions

[SR71]

Interesting responses.
It seems most of you believe in the unproven and not yet developed capabilities of computer technology that as yet does not exist, rather than accepting that human ability has had sole and successful dominance over road car safety and racing car activity for 200 years.

Some might call this opinion arrogant in the extreme.
So computers are better than the human brain, I hope I do not have to rely on any of you for my safety?

Does not exist huh?

This is an old video btw...

http://www.youtube.com/watch?v=Ol3g7i64RAI&sns=em

Yes old indeed, I posted that video and accompanying article over 1 yr. and a half ago
http://www.f1technical.net/forum/viewto ... di#p539625
Audi was not actually show casing an autonomous car in that display,
just the technology that they are using to help develop driving assists such as adaptive cruise control and active lane assist.

.

Audi was definitely showcasing autonomous technology in that video.

They have had several TT's and RS7's built with the sole purpose of being driverless and autonomous.

Lane assist is nothing compared to the technology shown on that RS7.

I believe that car also drove itself from San Francisco to Las Vegas all by itself for a CES promotion.

[SR71]

Interesting responses.
It seems most of you believe in the unproven and not yet developed capabilities of computer technology that as yet does not exist, rather than accepting that human ability has had sole and successful dominance over road car safety and racing car activity for 200 years.

Some might call this opinion arrogant in the extreme.
So computers are better than the human brain, I hope I do not have to rely on any of you for my safety?

Nobody is saying a computer is better at being a human. You need to realize that.
What we are saying is that computers can and in most cases are better at humans in specific and highly focused task.

Specifically driving, which can be boiled down to math, there is no art at the extreme end its pure physics.

What you are really saying is that you have no clue about present technologies that exist and what's on the horizon. You clearly have no understanding about machine learning which is a strong component of this challenge. This wouldn't be a bad thing, except you're telling us these things are years in the future when in reality they are years behind us.

What's wrong with a competition designed to provoke and inspire computer programmers and scientist to evolve autonomous driving and racing?

If you're such a fan of the Human mind why don't you have a little respect for the fact that some of the greatest human minds find this challenge quite fascinating and are dedicating thier time and money to the challenge?

You should read your comment and change driving to controlling a car. driving is much more than pure physics.
How do you conclude that I am unaware of the current state of play?
I have worked on pilotless aircraft so I have some reasonable knowledge.
It is indeed a worthwhile challenge for those in computer development, it might teach some of them about the real world.
You say the technology is years behind us, strange then that all we have so far are military drones and robots with very limited functions.
Go carefully with your virtual reality in ROBRace it might just bite you on the first corner.

You're commenting about 20yr old de-classified drones and saying that has some relevancy to this thread.

You have the same ability as everyone else here to use "the Internet" and educate yourself as to how advanced robotics has come in the last 10 years.

ROBOrace (if it ever happens) will be something new, something untested. It won't be perfect, but I don't think many people here care if it's perfect. We just want to see some cool technology used and hopefully an interesting racing series will be born.

[autogyro]

Hmmm I see the attitude still persists.

So go play with ROBORace if it lasts a season interest will decline just like FE.

Of course it is interesting technology.
I suggested trying an autonomous car at the limit in wet weather.
Try it I think you will find out what I mean.

[Andres125sx]

I suggested trying an autonomous car at the limit in wet weather.
Try it I think you will find out what I mean.

Can you show me a video of it doing the lap in the wet at racing speed and a comparison on lap times to a human driven car.

To what purpose?

Nobody said an autonomous car is faster than a human driver, at least not yet. But that´s exactly the reason for RoboRace, developing autonomous driving

Anycase I´m not sure about the reason you´re discussing about simulating a fly brain. RoboRace is not about developing AI, but about autonomous driving. You don´t need AI to drive a car.

Hmmm I see the attitude still persists.

Yes, I could say the same

[SR71]

Hmmm I see the attitude still persists.

So go play with ROBORace if it lasts a season interest will decline just like FE.

Of course it is interesting technology.
I suggested trying an autonomous car at the limit in wet weather.
Try it I think you will find out what I mean.

Yes because it's impossible for an autonomous car to detect its slip angle and wheel spin better than a human in the wet and apply counter steer and feather the throttle.

Things only a human is capable of!

[Andres125sx]

It looks like ESC is a science fiction technology then, I must have been living into a dream!


BTW, human reaction time is 1.5-2 tenths, computers reaction time is 0.01 tenth

[autogyro]

It is not just reaction time correcting a slide in the wet.
It is predicting the event and considering a response.

[autogyro]

It looks like ESC is a science fiction technology then, I must have been living into a dream!

It is all interesting and very useful technology.
However, a driver with only lightning reaction times will not win races against experienced drivers or be a safe driver on the road.

I do think that many people these days believe computer technology to be the answer to everything.
It certainly has a huge effect on human activity.
Living in virtual reality is not the answer though.
We need to review how we are interacting with the real world if we are to survive into the future.
The decline of F1 and the lack of interest in FE show the essential need for us to make this change.
Developing things simply to meet the needs of global business and greed is not a base to develop human pride in oneself or in others, fabricated competition and gimmicks can never achieve that ideal.

[Andres125sx]

It is not just reaction time correcting a slide in the wet.
It is predicting the event and considering a response.

... for we humans who don´t have instant reaction time.

Similar to turbo era. Drivers had to predict car reaction because of turbo lag, but if there´s no lag, there´s no need to predict anything.

That prediction is to be prepared and reduce reaction time. If reaction time is instant, there´s no need to predict anything, car only need to react when the slide has started.

However, a driver with only lightning reaction times will not win races against experienced drivers or be a safe driver on the road.

True, because drivers also need to know how to react, how much, what should expect after that to not be surprised, etc.

Computers don´t have any of those problems, they never get surprised, they always know how to react (if programmed correctly obviously), they never get over-excited...

[SR71]

It is not just reaction time correcting a slide in the wet.
It is predicting the event and considering a response.

That's all? Should be quite simple to pre-scan the road surface then. Could also load extremely local radar data into the car so it could more accurately predict rain in a specific corner.

Anyways, nobody had said these cars will race in the rain. There is nothing wrong with the first generation needing dry conditions.

You can keep arguing about wet conditions with yourself. As far as I'm concerned it's not an issue.

[Shrieker]

Can't wait. Hope this turns out great ! Wanna watch [insert your fav F1 driver here] go head to head wtih ROBoChamp.