Jump to content


Photo
* * * * - 3 votes

Roborace - Driverless Formula E support series


  • Please log in to reply
946 replies to this topic

#901 phrank

phrank
  • Member

  • 1,082 posts
  • Joined: May 01

Posted 11 September 2018 - 14:30

Racing 20 cars against each other on a closed, controlled circuit is a LOT easier task than navigating all the intricacies of urban and rural traffic on the open roads in direct interaction with thousands of other participants...

And both are not possible at this moment



Advertisement

#902 Jordan44

Jordan44
  • Member

  • 7,273 posts
  • Joined: May 15

Posted 11 September 2018 - 14:37

And both are not possible at this moment

 

Technology wise it is, legislation is preventing Audi from releasing a vehicle

 

https://www.fleetnew...mous-technology


Edited by Jordan44, 11 September 2018 - 14:41.


#903 maximilian

maximilian
  • Member

  • 3,969 posts
  • Joined: February 10

Posted 11 September 2018 - 14:41

Correct, the technology absolutely already exists, it's just a matter of the legal system catching up.



#904 phrank

phrank
  • Member

  • 1,082 posts
  • Joined: May 01

Posted 11 September 2018 - 14:46

Technology wise it is, legislation is preventing Audi from releasing a vehicle

 

https://www.fleetnew...mous-technology

 

 

When active, the AI Traffic Jam Pilot takes charge of driving in slow-moving traffic, controlling all necessary driving functions, up to a speed of 37mph.

Hardly "navigating all the intricacies of urban and rural traffic on the open roads in direct interaction with thousands of other participants"



#905 Jordan44

Jordan44
  • Member

  • 7,273 posts
  • Joined: May 15

Posted 11 September 2018 - 14:53

Hardly "navigating all the intricacies of urban and rural traffic on the open roads in direct interaction with thousands of other participants"

 

Level 3 automation is exactly that.

 

Level 3 -> Get from A to B without any human intervention required, but the user must be ready to take back control if something goes wrong.


Edited by Jordan44, 11 September 2018 - 14:55.


#906 phrank

phrank
  • Member

  • 1,082 posts
  • Joined: May 01

Posted 11 September 2018 - 14:57

Level 3 automation is exactly that 

I thought level 4 would be that? 



#907 Jordan44

Jordan44
  • Member

  • 7,273 posts
  • Joined: May 15

Posted 11 September 2018 - 14:59

I thought level 4 would be that? 

 

Level 4 specifies that the driver is present in the car, but does not need to pay attention.

 

Level 5 is no driver at all.



#908 BRG

BRG
  • Member

  • 16,631 posts
  • Joined: September 99

Posted 11 September 2018 - 16:39

Level 3 automation is exactly that.

 

Level 3 -> Get from A to B without any human intervention required, but the user must be ready to take back control if something goes wrong.

So level 3 is no human intervention required, except when human intervention IS required?  Total technofrolic BS.



#909 maximilian

maximilian
  • Member

  • 3,969 posts
  • Joined: February 10

Posted 11 September 2018 - 20:07

So level 3 is no human intervention required, except when human intervention IS required?  Total technofrolic BS.

 

It's more a legalofrolic BS at this time.  People have a hard time believing the car can actually drive itself better than they could, and so do legislators - despite all evidence to the contrary.  Once these legal/legislation hurdles clear, we are looking at a rapid reduction in human operators being replaced by Level 5 vehicles, whether it's soccer moms no longer having to drive their kids to 3 different sports team practices (and just sending the car instead), or giant logistics fleets running their driverless trucks 24 hours a day 7 days a week.


Edited by maximilian, 11 September 2018 - 20:22.


#910 pdac

pdac
  • Member

  • 7,425 posts
  • Joined: February 10

Posted 11 September 2018 - 20:13

So level 3 is no human intervention required, except when human intervention IS required?  Total technofrolic BS.

 

Exactly why (correctly) these things are still prohibited in a lot of places.



#911 scolbourne

scolbourne
  • Member

  • 506 posts
  • Joined: January 09

Posted 12 September 2018 - 03:13

I am not a great fan of driverless cars on the road. I think we are a long way from these being safe, but I guess they only have to be as safe as the worst human drivers to help improve road safety.

 

On the race track I think they should be able to perform better than most drivers in everything except avoiding other cars. If giving way to other cars the other car will get an advantage, if not you may crash. Rules may help here to apportion blame and give rights of way.

Is Roborace planning to simply set up the series or are they designing the software and vehicles ? I think they should leave the car design to others and just make the rules.

 

If it is for entertainment they might as well make the cars as fast as possible (but keep the weight to a minimum to help safety), and make the tracks as difficult as possible with low grip (dirt) sections and water sprinklers.

 

I think this series could be quite entertaining and may help to advance technology at the same time if they choose the right rules and circuits.



#912 4444

4444
  • Member

  • 138 posts
  • Joined: June 17

Posted 12 September 2018 - 03:27

This is the most road relevance series. F1 can shove their road relevance marketing else where.



#913 thegamer23

thegamer23
  • Member

  • 7,429 posts
  • Joined: November 13

Posted 09 November 2018 - 09:58

Roborace's first full-season racer revealed, Robocar plan dropped

https://www.autospor...-racer-revealed

 

 

9122b77b7424061bb63ca3381065f6db.jpg

 

b6d807e3f0f8036fd4f7d3410300d93f.jpg



#914 Vielleicht

Vielleicht
  • Member

  • 3,723 posts
  • Joined: June 16

Posted 09 November 2018 - 10:04

Curious. Bit of an innovative idea to have a ‘driver swap’ to AI. I can see that working more both from a machine learning and human sporting element front.

#915 phrank

phrank
  • Member

  • 1,082 posts
  • Joined: May 01

Posted 09 November 2018 - 10:49

Of course its a disappointment they cannot use their purpose build autonomous racer, but as far as I know this is only a small independent team developing this, without much outside help. What they really do well is designing attractive cars (unlike Formula E). Still hoping this can be a success



#916 JavierDeVivre

JavierDeVivre
  • Member

  • 604 posts
  • Joined: September 15

Posted 10 November 2018 - 19:48

Roborace's first full-season racer revealed, Robocar plan dropped

https://www.autospor...-racer-revealed

 

 

9122b77b7424061bb63ca3381065f6db.jpg

 

b6d807e3f0f8036fd4f7d3410300d93f.jpg

 

That is one good looking car, except for the over-sized diffuser, that based on this picture alone will do very little.

 

A simple less-is-more design that should have been used on the FE car, instead of going for a bloated sci-fi-esque design that tries too hard to look "futuristic".



#917 Vielleicht

Vielleicht
  • Member

  • 3,723 posts
  • Joined: June 16

Posted 11 November 2018 - 10:17

Personally, as trying to look futuristic goes I think that looks pretty tame.

 

edit: in fact, I see echos of Group C particular in those rear wing supports and smoother one-piece nose.

roborace-wont-use-a-fully-driverless-car


Edited by Vielleicht, 11 November 2018 - 10:21.


#918 ViMaMo

ViMaMo
  • Member

  • 6,376 posts
  • Joined: September 03

Posted 11 November 2018 - 10:20

Would you accept a cockpit less car, strange looking.

#919 ViMaMo

ViMaMo
  • Member

  • 6,376 posts
  • Joined: September 03

Posted 11 November 2018 - 10:21

Would you accept a cockpit less car, strange looking.

Advertisement

#920 phrank

phrank
  • Member

  • 1,082 posts
  • Joined: May 01

Posted 11 November 2018 - 10:39

Personally, as trying to look futuristic goes I think that looks pretty tame.

 

edit: in fact, I see echos of Group C particular in those rear wing supports and smoother one-piece nose.

 

I think that's where Formula E went wrong, they think that being futuristic means wild, crazy, over-the-top, while I think most racecar enthousiasts would agree these Roboracecars look modern and aesthetically wel design



#921 midgrid

midgrid
  • RC Forum Host

  • 7,459 posts
  • Joined: April 09

Posted 11 November 2018 - 12:05

As a growing and non-traditional racing series, Formula E needs to have a radical design to continue to get the attention of non-traditional racing fans.

#922 Vielleicht

Vielleicht
  • Member

  • 3,723 posts
  • Joined: June 16

Posted 11 November 2018 - 12:11

I think it’s way too early to conclude on whether Formula E has got the design style right or wrong. That will be measured in how many people identify positively with the new car, and by extension, the series over it’s lifetime. But for their stated goal, it doesn’t seem like a fundementally bad direction in my opinion.

#923 Vielleicht

Vielleicht
  • Member

  • 3,723 posts
  • Joined: June 16

Posted 10 April 2019 - 10:29

Behold, a real Season Alpha Roborace car

D3vfsuDXoAAoUcl.jpg



#924 phrank

phrank
  • Member

  • 1,082 posts
  • Joined: May 01

Posted 10 April 2019 - 10:53

The design department of Roborace is on point



#925 Kalmake

Kalmake
  • Member

  • 3,051 posts
  • Joined: November 07

Posted 10 April 2019 - 11:25

I think it's missing a point. We can't see the human driver not driving the car. Should have gone open cockpit.



#926 Clatter

Clatter
  • Member

  • 35,070 posts
  • Joined: February 00

Posted 10 April 2019 - 11:27

I think it's missing a point. We can't see the human driver not driving the car. Should have gone open cockpit.

 


Maybe they are trying disguise the fact that half the race will have a driver.

#927 pdac

pdac
  • Member

  • 7,425 posts
  • Joined: February 10

Posted 10 April 2019 - 11:35

Maybe they are trying disguise the fact that half the race will have a driver.

 

Nah. It's just that the car is the driver and the car has a really big helmet on.



#928 maximilian

maximilian
  • Member

  • 3,969 posts
  • Joined: February 10

Posted 10 April 2019 - 11:38

The design department of Roborace is on point

 

And that's about the ONLY thing they've gotten halfway right so far...

 

Although in fairness, at least partial credit probably to Munich Tech Uni...


Edited by maximilian, 10 April 2019 - 11:41.


#929 7MGTEsup

7MGTEsup
  • Member

  • 2,028 posts
  • Joined: March 11

Posted 10 April 2019 - 13:04

Level 3 automation is exactly that.

 

Level 3 -> Get from A to B without any human intervention required, but the user must be ready to take back control if something goes wrong.

 

You might as well be driving if you still have to pay attention and be ready to take control. I'm assuming that is so the manufacturer can blame the driver and not the AI in event of a crash.



#930 Jordan44

Jordan44
  • Member

  • 7,273 posts
  • Joined: May 15

Posted 10 April 2019 - 18:14

You might as well be driving if you still have to pay attention and be ready to take control. I'm assuming that is so the manufacturer can blame the driver and not the AI in event of a crash.

It's more because that's where the technology is at the moment.

Tesla are about to release full self driving, edging towards L4. Will be interesting to see what the performance is like.

Edited by Jordan44, 10 April 2019 - 18:16.


#931 kumo7

kumo7
  • Member

  • 1,749 posts
  • Joined: May 15

Posted 11 April 2019 - 01:43

Roborace's first full-season racer revealed, Robocar plan dropped

https://www.autospor...-racer-revealed

 

 

9122b77b7424061bb63ca3381065f6db.jpg

 

b6d807e3f0f8036fd4f7d3410300d93f.jpg

 

 

SO is this the level 3 race car?

Why should there be a place and possibility to accommodate a race drivers in a LMP style?

 

 

Older design was much more roboracer. 

What is the fun of this new design? 

 

I totally missed the point. 



#932 Ben1445

Ben1445
  • Member

  • 2,706 posts
  • Joined: December 13

Posted 11 April 2019 - 06:47

SO is this the level 3 race car?
Why should there be a place and possibility to accommodate a race drivers in a LMP style?


Older design was much more roboracer.
What is the fun of this new design?

I totally missed the point.


Seems like they’re zeroing in on a race format than involves having a human driver for the opening stint, after which they jump out and hand over to the AI.

Gives it data to learn from based on what the drivers do. Or something like that. Hence calling it ‘Season Alpha’ as an indication that it’s work in progress.

#933 Spillage

Spillage
  • Member

  • 5,427 posts
  • Joined: May 09

Posted 11 April 2019 - 13:15

I think I'd rather they held off altogether until the AI can handle doing the full race on its own. No point rushing it before the technology is ready.



#934 7MGTEsup

7MGTEsup
  • Member

  • 2,028 posts
  • Joined: March 11

Posted 11 April 2019 - 13:25

Seems like they’re zeroing in on a race format than involves having a human driver for the opening stint, after which they jump out and hand over to the AI.

Gives it data to learn from based on what the drivers do. Or something like that. Hence calling it ‘Season Alpha’ as an indication that it’s work in progress.

 

Starting to sound less like AI and more like copying what the human does then repeating it metronomically. 



#935 Ben1445

Ben1445
  • Member

  • 2,706 posts
  • Joined: December 13

Posted 11 April 2019 - 14:01

Starting to sound less like AI and more like copying what the human does then repeating it metronomically. 

That would be an interpretation that does a disservice to machine learning. Whilst the driver is in control it is merely providing reference data for the AI rather than hard set instructions on how to do it. It's giving it a leg up - a point to check itself against. It's like how a young child will imitate what the parent does and thereby learn from doing so. 



#936 BalanceUT

BalanceUT
  • Member

  • 1,741 posts
  • Joined: February 16

Posted 11 April 2019 - 14:18

I've been fairly positive on the advent of AI and have overtly expressed here and on other forum threads that I believe it is coming fast.  But, I've recently become a bit pessimistic on the possibility of self-driving cars in the near future. The issue is recent work in neuropsychology on consciousness suggests AI driving is unlikely to meet the promise many imagine, including myself as recently as a few months ago.

 

We've long thought consciousness was a kind of 'ongoing record of our awareness of the world'. But that description was difficult to test and didn't fit a good bit of evidence. A revised theory has emerged relatively recently: Consciousness appears to emerge from the integration of multi-sensory inputs forming an interpretable whole. That whole is formed from our experience with the world and, in comparison with our experience, we can infer the near future, hidden information, possible reactions from our actions, etc.* Consciousness is the ongoing continual meta-experience of the comparison of our current experience with our past experiences and inferences for the future. That meta-experience is, essentially, our personal history, sense of self... our conscious knowledge of who we are, in this moment, extending back in time and into the future. These new theories are being put to test by creating blocks on some senses and noting how it interrupts conscious memory. 

 

We also have 'theory of mind' in making inferences about other persons' behaviors based on the situation and what we know of their dispositions, with situation being a strong driver of behavior. We do a lot of communication with, and inference from other drivers by non-verbal behaviors. We see the person in the car ahead on their cell phone or animated discussion with a passenger and we know, implicitly, that their attention is not fully on the road and that we should be a bit more aware of their possibility of an unexpected movement. We keep a closer eye on that 'point of risk'. Or, we have a moment of eyes meeting eyes of the person walking across the street. We know they saw us in the car, and they know they are safe to cross because they know we saw them. 

 

Therefore, we are not simply looking at the world as a collection of objects, but as a collection of linked, hidden, and interacting objects which includes ourself. We don't just see a car at the crossing street's stop sign at a 4-way stop having arrived a couple of seconds after we arrived at our stop. We also see the driver's facial expression and interpret if they appear 'aware' of us, intend to move immediately, intend to wait, seems to want to 'break the 4-way rule', etc. We 'know'. We are conscious of all this in a way that it is unlikely we can code into software and hardware, at least for a long time yet to come.

 

That somewhat rules out the essential transition phase of automated vehicles. If ALL vehicles were AI and communicating with each other, as we cross, with all of us wearing smart glasses to tell us a car has 'seen us' and is attending to our crossing the street, etc., then an AI world of vehicles becomes possible. Because they are doing the communication that we do non-verbally. Until then, we are stuck behind meat brains that do certain things much better than computer brains, but also make mistakes that others must account for. 

 

Yes, I know they are training computers to recognize human facial expressions...While there are 6 major universal facial expression, we express tremendously more intent and emotion. That tech is lagging human abilities by leaps and bounds, and its integration into automobile driving tech is probably barely even considered as of yet. 

 

And, even then, the thoughts are that linear designed programming (even if using parallel processors, the essence is still linear) seems unlikely to capture the simultaneous processing of integrated information to create a useful whole that filters the irrelevant  (the tree behind the fence) and pulls in the tiny but essential: the eyes of the driver in the other car directed downward to their phone.

 

*For instance, in child development we know that infants go through a stage in which an object that moves out of their field of vision is treated as literally vanished. You take a toy, move it behind your back and it is GONE to that child. They don't yet infer that it is still existing, and only out of view. That shift to being able to see the toy as out of view, but still there is likely an essential step in the development of consciousness according to these new theories of consciousness. (This is called object permanence and develops between ages 4 and 7 months).

 

Every noticed that you don't experientially remember anything from before about age 3? Anything you 'remember' from before is really just stuff you were told by others, you have no 'experience' of it to call on. That's also connected to this notion of consciousness. Before about age 3, we simply have not yet stored enough information about the world sufficient recognition of self being different from other (that happens about age 18 months), etc. to begin to form a continual 'history of self', the essence of consciousness.

 

Lecture over. 


Edited by BalanceUT, 16 April 2019 - 14:59.


#937 Clatter

Clatter
  • Member

  • 35,070 posts
  • Joined: February 00

Posted 11 April 2019 - 14:18

I think I'd rather they held off altogether until the AI can handle doing the full race on its own. No point rushing it before the technology is ready.

I think doing it this way will help them advance at a better pace. It's one of the problems when doing something new on a shoestring budget.

#938 ElectricBoogie

ElectricBoogie
  • Member

  • 140 posts
  • Joined: March 19

Posted 12 April 2019 - 13:51

While Roborace is working towards another goal, might there be a "risk" of their car being quicker over the same lap than Formula E?
F for an "open" wheeler may be relatively slippery, but the downforce seems to be largely absent, tyres basically road car take-offs. 
It's my understanding that a more performance focused all-electric car could be heaps faster than Formula E. Roughly around F2 levels, especially if stints can be kept short. I'd love to see battery swaps during pit stops. Even choice of battery size/weight and C rate, bringing lots of strategy. Pits stops cost time, weight slows a car down. We need a pit stop for new battery to go fast to the finish, but which combination of range and power pack do we take?

Roborace seems to have struggled with just that ons small team for much too long. Bringing in universities or teams to do better would have brought them a lot forward. On demonstrations, a novice lady driver was quicker in the car than the car by itself without her weighing it down.

What I can't wrap my head around is that race simulators can easily keep cars driving competitively, even many together into a turn with reasonably realistic interactions, but Roborace can't show a single proper hot lap to save their lives. How much difference is there between measuring where you are on a track and KNOWING where you are on the track?



#939 absinthedude

absinthedude
  • Member

  • 576 posts
  • Joined: June 18

Posted 12 April 2019 - 14:28

At least it looks like a serious car....unlike the monstrosity that currently roams the Formula E tracks...and which looks like an 8 year-old's idea of a "futuristic racing car". It's a fair baseline from which to develop the concept. 



Advertisement

#940 Ben1445

Ben1445
  • Member

  • 2,706 posts
  • Joined: December 13

Posted 12 April 2019 - 14:31

.... the monstrosity beautiful creature that currently roams the Formula E tracks...

Fixed it for you  :p



#941 7MGTEsup

7MGTEsup
  • Member

  • 2,028 posts
  • Joined: March 11

Posted 15 April 2019 - 13:19

That would be an interpretation that does a disservice to machine learning. Whilst the driver is in control it is merely providing reference data for the AI rather than hard set instructions on how to do it. It's giving it a leg up - a point to check itself against. It's like how a young child will imitate what the parent does and thereby learn from doing so. 

 

Isn't that just another way of saying what I said? Monkey see monkey do.



#942 Ben1445

Ben1445
  • Member

  • 2,706 posts
  • Joined: December 13

Posted 15 April 2019 - 13:46

Isn't that just another way of saying what I said? Monkey see monkey do.

Not exactly.  It's not the same as copying and repeating - that's not 'intelligent', that's just following instructions. 

 

Think of it the way. Put a human who's never seen a kart in a kart they will eventually work out how the thing works and start turning laps. With enough practice they will eventually get faster of their own accord. But if they start following an experienced driver around, or are driven around in two seater by one, they will get faster much quicker than if they are left to work it out themselves. That's not changing the intelligence of the person in the kart, it's just providing them with more information to learn faster. It's the same for AI. 



#943 BRG

BRG
  • Member

  • 16,631 posts
  • Joined: September 99

Posted 15 April 2019 - 16:47

Not exactly.  It's not the same as copying and repeating - that's not 'intelligent', that's just following instructions. 

 

Think of it the way. Put a human who's never seen a kart in a kart they will eventually work out how the thing works and start turning laps. With enough practice they will eventually get faster of their own accord. But if they start following an experienced driver around, or are driven around in two seater by one, they will get faster much quicker than if they are left to work it out themselves. That's not changing the intelligence of the person in the kart, it's just providing them with more information to learn faster. It's the same for AI. 

I am sure you are right here.  Which begs the question, why didn't Roborace know this?  And why have they spent several years trying to do it the hard way?



#944 Ben1445

Ben1445
  • Member

  • 2,706 posts
  • Joined: December 13

Posted 15 April 2019 - 16:55

I am sure you are right here.  Which begs the question, why didn't Roborace know this?  And why have they spent several years trying to do it the hard way?

That is one I do not have an answer for! Your guess is a good as mine. 



#945 Clatter

Clatter
  • Member

  • 35,070 posts
  • Joined: February 00

Posted 15 April 2019 - 22:23

I am sure you are right here.  Which begs the question, why didn't Roborace know this?  And why have they spent several years trying to do it the hard way?

 


Probably because that was the original remit. They now realise it is harder than envisaged, are behind in the schedule, so need to take a different approach.

#946 johnnyw

johnnyw
  • Member

  • 530 posts
  • Joined: November 09

Posted 27 April 2019 - 10:27

Pace is looking pretty bad again. I for sure hope they get big upgrade package for spain.

#947 BRG

BRG
  • Member

  • 16,631 posts
  • Joined: September 99

Posted 29 April 2019 - 15:39

Pace is looking pretty bad again. I for sure hope they get big upgrade package for spain.

:confused: