The OSS reads light colors differently dark colors

superchuck500

superchuck500

Team Owner
Joined
May 30, 2017
Messages
495
Points
273
Location
Charleston, SC
As a result, NASCAR has been giving dark cars a higher resolution scan (takes longer) than the low res on the light cars. Some teams are likely trying to exploit weaknesses in this process.

SPARTA, Ky. – NASCAR has had to stay ahead of teams and continues to update its software algorithm for the Optical Scanning Station (OSS) after teams claim they found a way to “confuse” it with contrasting dark and light colors along the wheel wells and edges of the right rear of the racecar, multiple sources confirmed to Kickin’ the Tires.

This is the first season for NASCAR’s at-track OSS, where cars are rolled into a dark tented room and then illuminated with thousands of laser lights that take measurements of all angles of the car. According to NASCAR there are more than 130,000 points of measurement from eight laser projectors that are captured by 17 cameras – 16 around the car and one that is placed underneath the car. At the end of the scan, NASCAR is given a virtual 3D rendering of the surface of the car. There are two types of scans – a high-resolution and a low-resolution.

Teams with all black or extremely dark-colored racecars are given a high-res scan. Teams with light colored racecars usually get low resolution scans because they do not take as long as the high-res scans. Now, it appears that over the past couple of weeks, some teams have figured out that changing the paint scheme and incorporating a dark color – most-commonly black – around the rear fender wells, along the bottom edge of the car and/or up the rear bumper cover tricks the OSS into thinking the car is narrower than its actual measurements.
Click to expand...

http://kickinthetires.net/index.php...ter-teams-claim-to-have-found-laser-weakness/
 
I saw something on this a little while ago. These teams will figure out anything to beat whatever NASCAR comes up with.
 
True, but from a standpoint of competition, shouldn’t the cars all get the same scan?

Even if the reason is practical, it’s still a problem that invites exploitation if some cars get one scan and other cars get a different scan.
 
ChexOrWrex said:
You mean holes in software are being exploited? And the software is probably riddled with inconsistent bugs?

Color me shocked.
Click to expand...
No, holes in the ability of the laser to measure light and dark are being exploited. That's hardware, not software.
superchuck500 said:
True, but from a standpoint of competition, shouldn’t the cars all get the same scan?
Click to expand...
Yes, but the article says since light colors are can be scanned accurately with fewer data points, which means they can be scanned faster. Sounds like NASCAR needs another OSS so they can scan more cars at the same time.
 
wi_racefan said:
In a time that the RTA is going to NASCAR saying you need to help us save money I wonder how much they spent figuring this out?
Click to expand...
It probably didn't cost the teams anything to figure this out.

Crew chief of dark car: "Hey, why do you always get through inspection faster that we do?"
CC of light car: "We ran a dark scheme last week and I notice it took us longer too. I asked and the inspector told us the OSS doesn't scan as many data points on our normal paint job."
Dark CC: "Hm. I wonder if there's a way to exploit that? Maybe with a mix of light and dark? Say, a predominantly light car with dark paint in the areas we'd like to fudge? ... Yeah, that's the ticket!"
 
wi_racefan said:
That would just create more issues..."They got to go thru the good OSS and we got stuck with the bad one!"
Click to expand...
:XXROFL:
Yeah, but scanning the cars at different rates screws the dark cars. If I fail with a dark car, I've lost more time than a failing light car, since his inspection took less time than mine. I lose even more time with each subsequent inspection, pass or fail. Indeed, every car behind me loses more time too.

It would be interesting to see how many dark cars that have failed first inspection wound up running out of time before completing a second one. I suspect NASCAR will soon change their procedure and require the same number of data points be gathered for all cars regardless of color, creed, or national origin. :biggrin:
 
Charlie Spencer said:
It probably didn't cost the teams anything to figure this out.

Crew chief of dark car: "Hey, why do you always get through inspection faster that we do?"
CC of light car: "We ran a dark scheme last week and I notice it took us longer too. I asked and the inspector told us the OSS doesn't scan as many data points on our normal paint job."
Dark CC: "Hm. I wonder if there's a way to exploit that? Maybe with a mix of light and dark? Say, a predominantly light car with dark paint in the areas we'd like to fudge? ... Yeah, that's the ticket!"
Click to expand...

Yes they know that a lighter car scans faster than a darker one, but how much R&D time and money did they spend figuring out that "If I make this corner darker I can trick the OSS into thinking is actually narrower than it really is"? You had to have a light one, a dark one, probably other shades, then numerous sizes to figure out how much you could run thru the OSS and still get away with it. I highly doubt some guy in 5 minutes while eating his lunch one day came up with this idea and said it will give us an extra 0.035" of tolerance and they did it to the car and brought it to the track.
 
wi_racefan said:
Yes they know that a lighter car scans faster than a darker one, but how much R&D time and money did they spend figuring out that "If I make this corner darker I can trick the OSS into thinking is actually narrower than it really is"? You had to have a light one, a dark one, probably other shades, then numerous sizes to figure out how much you could run thru the OSS and still get away with it. I highly doubt some guy in 5 minutes while eating his lunch one day came up with this idea and said it will give us an extra 0.035" of tolerance and they did it to the car and brought it to the track.
Click to expand...
I see, you're wondering about the details. I was just looking at the big picture.
 
From the article ---
NASCAR’s official comment to Kickin’ the Tires was, “NASCAR is fully confident in the ability of the Optical Scanning Station to help provide a level playing field for the garage. We continually update the software to stay ahead of any possible challenges.”
 
I love this stuff...people say ingenuity is "dead" in NASCAR and all the cars are the same and no one can find a grey area to make their car better...but that's exactly what this is, imo. Some figured it out, some didn't...I think it's cool, at least.
 
compares datums (way points),(points) = tolerances on the test car to the Cad Model of the car
 
4tires17gals said:
That's why Furniture row went from brown/white to black !
Click to expand...
:rolleyes:

ef5805b7fbd4d033801954b1c3d168de.jpg
 
Very simple to rectify. Each car goes through a spray booth and all cars come out white.
Those that pass go through a car wash to clean it. Those that fail get loaded for the trip home.
After a couple weeks and sponsors demanding a refund, all cars will pass or stupid people fired.
 
When harvick s car first failed, nascar should have went to every team and laid down the law. Then when other teams started failing pre and post inspections, start sending teams home. A second violation should have been a see you next year. If the owners had a problem with it, question whether they want to save money or keep wasting money on micro manipulating the cars aero.
 
They must've caught on to that pretty early on. Wasn't that like race #4? No pun intended

**After further review it looks like their Atlanta car had it too. Minus the caved in window.
 
This was from Daytona. So yeah, they started doing this as soon as they hit the 1.5ers. Somebody knew about this from the jump. I guess if they had an OSS bay on campus they figured it out on their own.

upload_2018-7-18_7-55-11.png
 
For comparison, this was the 9 at Atlanta, 2nd race of the season. Behind the curb, also one of the last organizations to have their own OSS bay.

upload_2018-7-18_8-1-35.png
 
wi_racefan said:
Yes they know that a lighter car scans faster than a darker one, but how much R&D time and money did they spend figuring out that "If I make this corner darker I can trick the OSS into thinking is actually narrower than it really is"? You had to have a light one, a dark one, probably other shades, then numerous sizes to figure out how much you could run thru the OSS and still get away with it. I highly doubt some guy in 5 minutes while eating his lunch one day came up with this idea and said it will give us an extra 0.035" of tolerance and they did it to the car and brought it to the track.
Click to expand...

My guess is it was much simpler than that. Since paint jobs change pretty often, a car that had previously passed inspection with a dark paint scheme went to a different track with a light paint scheme and their car that previously passed no longer did.

We can take a look at paintjobs over the season and make reasonable assumptions on who figured this out.
 
Are we seeing a trend here? This OSS **** doesn't work any better than anything else....yet cost the teams plenty.
 
StandOnIt said:
Nobody broke their arms and made them buy one, Testing is free for all teams.
Click to expand...

Those little tricks don't come at the R&D center. Now you have your SHR explanation. Can you imagine the ****storm if this had been any of MY Toyota teams? Still could be coming I suppose. Would be disappointed if they are behind on this. Might mean that if the Toyotas are legal, they are dominant again, and you will resurrect your wind tunnel stuff.
 
You must log in or register to reply here.
Back
Top Bottom