New User? Sign Up  |  Sign In  |  Help
“Logo
 

                 LogoLogo

Question
Answer
Discover
     
Search for questions :
My Profile

philosopher

Open Questions Bookmark and Share

Does this sound like behavior modtication (ABA) for AI cars?

I am not sure how this would work.
TRIAL AND ERROR. Artificial intelligence startup Wayve believes it’s found a better way to train autonomous vehicles (AVs) than the existing methods of sophisticated hardware and detailed 3D maps. On Monday, the company released a video in which a modified Renault Twizy (a two-seated electric vehicle) learns to autonomously navigate a road. It does so through reinforcement learning, a type of machine learning in which a system earns a “reward” for desirable behavior and a “penalty” for undesirable behavior. They detail their experiment in a paper published on arXiv.

THE PROBLEM WITH MAPS. Most in-development AV systems rely on detailed 3D maps to navigate. Companies across the globe are currently racing to create these maps, using sophisticated sensors and cameras to detail city streets and highways alike. AVs, in turn, require their own complex systems of cameras and sensors to then navigate these maps.

Unfortunately, 3D maps are labor-intensive and require frequent updates to account for things like construction. Companies that create 3D maps also tend to focus on highly trafficked roads first, leaving rural areas behind.

TEACHING A TWIZY. Wayve’s technique could render those 3D maps obsolete. The Twizy in the video had only one camera mounted at its front (most other AVs have many more — Tesla’s autopilot, for example, uses eight). That one camera fed information in real-time to a graphics processing unit (GPU) onboard the car. The GPU ran Wayve’s reinforcement learning algorithm, which controlled the vehicle’s acceleration, braking, and steering. A human driver sat behind the wheel during the learning process, stopping the car every time it veered off the road, which “penalized” the system. The longer the system drove without this human intervention, the greater the “reward” it received.

Within about 20 minutes, the car figured out how to follow the gently curving road indefinitely.

If Wayve’s approach to autonomous driving catches on, AVs would no longer need to rely on 3D maps. And that could mean that anyone, anywhere could enjoy the benefits of autonomous vehicles.

READ MORE: AI Algorithm Teaches a Car to Drive From Scratch in 20 Minutes [New Atlas]


Last updated: July 6, 2018 at 3:19 pm
#Autonomous Cars#reinforcement learning#th
https://futurism.com/wayve-ai-teach-autonomous-cars/

9 day(s) ago

    Comment(s) (0)
    Report
   Like  
   Email to Friends  
   Bookmark  
   Subscribe to Answer Alerts  
No comments yet !     Be the first to comment !
Answers (1)

Kravenhead
Ha! Yes, it does.
I find it a little odd that a machine would recognize rewards and penalties.
Those distinctions are usually based on feelings.
It's like your vacuum doesn't care if it's on or off, or whether it's cleaning a rug or the bathroom floor.
I do see a problem with the 3D maps, though. I wouldn't want to be riding in an AV that's using outdated information. Things change so quickly, I don't know how well they'd keep up with updates.
I don't think AVs should go mainstream until they're able to navigate a totally alien environment on their own.... No maps.

Posted 9 day(s) ago

( 0 )
( 0 )
    Comment(s) (0)
   Report
No comments yet ! Be the first to comment on this answer !

Edit your answer. Click save, when done.
Question Title Does this sound like behavior modtication (ABA) for AI cars?
Your Answer
  |         |                            
bold  italic  underline  strike       big  small       superscript  subscript 
  Allows to add a link.
Caption :
Link URL :
( Must starts with "http://" )
Add  |   Cancel
  Allow to insert an image. Must be among the following file types - *.jpg, *.gif, *.png & *.bmp.
Image Url :   Upload New
( Image url must always starts with " http:// " )
Width : pixels
( Must not be greater than 450px. Enter 0px for no resize )
Add  |   Cancel
  Allow to insert YouTube video. Insert the video embed code.
Embed Code :
Add  |   Cancel
Character Count ( Max. - 5000 ) : 2488
  Your comment on this question
  |         |                            
bold  italic  underline  strike       big  small       superscript  subscript 
  Allows to add a link.
Caption :
Link URL :
( Must starts with "http://" )
Add  |   Cancel
  Allow to insert an image. Must be among the following file types - *.jpg, *.gif, *.png & *.bmp.
Image Url :   Upload New
( Image url must always starts with " http:// " )
Width : pixels
( Must not be greater than 450px. Enter 0px for no resize )
Add  |   Cancel
  Allow to insert YouTube video. Insert the video embed code.
Embed Code :
Add  |   Cancel
Max Allowed : 5000 Characters Current Count : 0
  Your comment on this answer
  |         |                            
bold  italic  underline  strike       big  small       superscript  subscript 
  Allows to add a link.
Caption :
Link URL :
( Must starts with "http://" )
Add  |   Cancel
  Allow to insert an image. Must be among the following file types - *.jpg, *.gif, *.png & *.bmp.
Image Url :   Upload New
( Image url must always starts with " http:// " )
Width : pixels
( Must not be greater than 450px. Enter 0px for no resize )
Add  |   Cancel
  Allow to insert YouTube video. Insert the video embed code.
Embed Code :
Add  |   Cancel
Max Allowed : 5000 Characters Current Count : 0
Email this question link to friends
You must enter an email-address, if a name is entered and vice-versa for each friend.
Friend #1 -
Friend #2 -
Friend #3 -
Friend #4 -
Friend #5 -
Enter Security Code : *

 

Copyright ©  All rights reserved.

Contact Us   Terms of Use