Post new topic Reply to topic  [ 79 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6, 7, 8  Next
Author Message
 Post subject: Re: Timing System, Really
PostPosted: Thu Oct 15, 2009 12:42 am 
Offline
User avatar

Joined: Wed Oct 15, 2008 4:24 pm
Posts: 255
shafeeq wrote:
I'm beginning to think the cheapest decent solution here is to use cmutv's finish line camera with a running clock in view. Start the clock from the gun, and after the race, watch the tape in slow mo to see when the buggy crosses the line and what the time is. Precision of at least .01s, Accurate to at best .03s. Cost probably 1-2k depending on how big a clock we need.


I'm just a chemist, but is it impossible to rig a clock to make a loud BEEEEP! sound when it starts and have it in view of a camera with a high enough frame rate for the accuracy required?

Oh, and please guys, we're buggy people. We turned an innocent quasi-parade float contest into this ridiculousness. We can hack, jam or disable an onboard system. I'm convinced we could cure cancer if it meant that PiKA would never win again.


Top
 Profile  
 
 Post subject: Re: Timing System, Really
PostPosted: Thu Oct 15, 2009 10:00 am 
Offline
User avatar

Joined: Mon Oct 13, 2008 3:37 pm
Posts: 135
Location: SoCal
Carl Nott wrote:
shafeeq wrote:
I'm beginning to think the cheapest decent solution here is to use cmutv's finish line camera with a running clock in view. Start the clock from the gun, and after the race, watch the tape in slow mo to see when the buggy crosses the line and what the time is. Precision of at least .01s, Accurate to at best .03s. Cost probably 1-2k depending on how big a clock we need.


I'm just a chemist, but is it impossible to rig a clock to make a loud BEEEEP! sound when it starts and have it in view of a camera with a high enough frame rate for the accuracy required?


That may be the cheapest/simplest solution to replace what we have now with a more accurate measurement of finishing times.

As Shafeeq mentioned there are lots of systems setup to do 3/4 of what we need/want (one thing that complicates some of the off the shelf solutions is that we run point to point rather than a loop with a start/finish line).

If we did something with high-speed video there isn't a simple way to expand it to timing other splits (hills, transitions etc) in the future. So if money is hard to come by for a timing system a high-speed video + clock might be a good fall back solution for a simple system that can't easily be tampered with.


Top
 Profile  
 
 Post subject: Re: Timing System, Really
PostPosted: Thu Oct 15, 2009 10:16 am 
Offline
User avatar

Joined: Wed Oct 15, 2008 4:24 pm
Posts: 255
revo wrote:
If we did something with high-speed video there isn't a simple way to expand it to timing other splits (hills, transitions etc) in the future. So if money is hard to come by for a timing system a high-speed video + clock might be a good fall back solution for a simple system that can't easily be tampered with.


Again, I'm just a simple chemist, but couldn't you just have displays of the same clock in view of cameras at the different split points? Wouldn't even need that terribly high of a frame rate, I'd think, if you just wanted split times to 0.1's. I dunno how you'd sync all the clocks/monitors (wi-fi? intarwebs? microwaves? no clue about this stuff; I'd just use hard wired vga cables running to monitors and going to the same computer) but if need be you could buy a bunch of those self-calibrating clocks that use some sort of atomic clock and radio/space waves and have an extra one in view of the start clock so you could calibrate it.

Edit: Granted, the splits wouldn't be automatic recorded, but the Buggy Alumni Association could sell split time data taken off of the video to those who are interested. Ooh, and the BAA could set up a camera at the top of Hill 2 and at the bottom of Hill 3 with the same system just continually rolling during important/interesting freerolls and sell those times as well.


Top
 Profile  
 
 Post subject: Re: Timing System, Really
PostPosted: Thu Oct 15, 2009 11:41 am 
Offline
User avatar

Joined: Fri Sep 26, 2008 1:56 pm
Posts: 184
Have you never heard of GPS time? How do you think all those aircraft deconflict with each other while flying around. Each timing station gets a GPS signal. Instant psynch. My new HD video camera has a built in GPS and that only cost me 700 bucks. This can be done.


Top
 Profile  
 
 Post subject: Re: Timing System, Really
PostPosted: Thu Oct 15, 2009 11:42 am 
Offline
User avatar

Joined: Fri Sep 26, 2008 1:56 pm
Posts: 184
lemuroid wrote:
hvincent wrote:
I guess assuming that people are going to be honest isn't good enough, huh.


Nope, People are evil.

When I joined the zoo, the list of ideas for how to beat PKA without actually being faster was long and nasty. This RFID hack would not have even made the top 20.

My favorite was:
Spike their milk, OJ, and kool-aid with LSD before the race

Never implemented but always considered.


Thanks for nothing. The LSD would have made CMU much more bearable. :D


Top
 Profile  
 
 Post subject: Re: Timing System, Really
PostPosted: Thu Oct 15, 2009 2:16 pm 
Offline

Joined: Tue Oct 28, 2008 6:40 pm
Posts: 136
Yes, GPS time is a great way for two unconnected devices to both agree on what time it is. The commercial timing systems have a GPS input exactly for that reason.
Without it, the solution is to bring the timing devices together at the start of the event, synchronize them, and then take them to wherever they will be on the course. Depending on the sport, the clock drift may or may not matter (won't for buggy).

If we wanted multiple chip decoders around the course, we'd have to use one of those two methods, too.

The tricky part for us is that the event we're trying to time (the buggy crossed a line) is apparently hard to reliably detect automatically, but easy to do by a human watching a replay (same as deciding a touchdown, anyone?)
Then again, we care more about that at the finish line than anywhere else.

If Bordick's camera can overlay a time with 1/100s digits, we may have a winner.
We might have to worry a little bit about where exactly in the recording process does the time get printed - the camera has around .25-.5s processing delay between when it sees something and when it gets recorded, but as long as it assigns the time consistently, the start and finish will both have the same lag, so it doesn't matter. The same applies for cmuTV's clock, and there's a lot more processing steps there.

There are start strobes (used for deaf runners) that the camera could pick up to signal a start event. Track has lot of point-point events (100, 200), but everyone finishes within a second of each other, which means they need the expensive high-speed video systems anyway.


Top
 Profile  
 
 Post subject: Re: Timing System, Really
PostPosted: Thu Oct 15, 2009 3:55 pm 
Offline
User avatar

Joined: Sun Oct 12, 2008 8:41 pm
Posts: 169
Location: Calabasas CA or Sunnyvale CA
abordick wrote:

Thanks for nothing. The LSD would have made CMU much more bearable. :D


I think your broadcasting buddy took a few hits last year. This year. I think we could both sit at your location in master control and still manage to cover the chute action due to the excellent video feed. (and thus have one less anti-clued on-air speaker).


Top
 Profile  
 
 Post subject: Re: Timing System, Really
PostPosted: Thu Oct 15, 2009 5:45 pm 
Offline

Joined: Thu Aug 27, 2009 11:30 am
Posts: 20
lemuroid wrote:
RFID cloning is real and fairly simple. RFID encryption appears to be non-existent. The equipment involved is not bulky.


Yeah, but not in our case. Sweepstakes hands chips out before raceday. You could clone your own chip/tag but you are limited to making your buggy finish the race faster ('cause if you made your own buggy slower, you'd be an idiot). You have no information about other team's chips. RFID tag numbers are typically not sequential to avoid erroneous correlations on readback.

lemuroid wrote:
Given that the timing systems look for when the repflectd signal is strongest as the basis for the timing, all I need is to clone the RFID from a given transponder and then play it back a bit more strongly than the translonder ( i.e. with a bigger antenna or with a powered antenna) in a manner that causes the peak signal to occur with different timing at the finish line vs. the buggy.


Again, you are not understanding that the transmitter (i.e. the antenna on the finish line) sends out a pseudo-random sequence on each pulse. In order for you to send out a signal "more strongly" (as you put it) you'd need to read the pseudo-random pulse sequence every time and then wait for the buggy to cross the line. You'd then have to send the signal delayed by a certain amount of time (to make the buggy appear to be slower) and by then the first time will have already registered with the race software and your second signal would achieve nothing.

lemuroid wrote:
If it is my buggy, I have plenty of time to make the clone and get set at the finish line to fake an early finish. If my intent is more evil and I want to make a rival’s buggy slower, I would have to clone their RFID signature prior to the start. This could be as late as when they are on the way to the start line. Even then, I would have about 2 minutes and 3 seconds to make my way to the finish line and generate a late crossing.


See the first comment. Yes, you could interrogate the RFID at the beginning. You could even stand next to every truck and identify their RFIDs for every buggy that day. Making RFID clones for your heat competition might be a little trickier since you then need to move your equipment to the finish line and then somehow obstruct the finish line area. You realize these antennas point downwards, right? The antennas also have a narrow field of view. So unless you plan on inconspicuously standing right at the finish line, cheating is not going to happen. All you need to do is wait for some guy to yell out "he's cheating!" and get your ass kicked by a whole bunch of drunk frat guys.

lemuroid wrote:
Even if cloning is not possible, here is a simple hack to consider. Driver slides a foil wrapped box (faraday cage) over the transponder a few feet before crossing the finish line. Driver stashes the box on the way back to the truck. Max signal occurs earlier than if this had not been done, time for the driver is reduced. No electronics required.


Real tinfoil would probably work easier than a faraday shield.

Two things with this:
1) If you are fast that day, I believe you have to open your buggy in front of a sweepstakes person to ensure that all safety requirements have been met (whereby your shenanigans would get caught). You may be able magically take this foil, scrunch it into a ball and somehow eject it from the buggy (loss of mass) and be DQ'd. Or the driver could try hiding it in the buggy by somehow dislocating her shoulder and cramming it into her pockets... NOT LIKELY. Or you could wait and have the slip of foil obscure the drivers vision and have her crash because you wanted to cheat.
2) If your signal is not read because you are obscuring your RFID, technically you won't register with the timing circuit and you'd make yourself slower.


Top
 Profile  
 
 Post subject: Re: Timing System, Really
PostPosted: Thu Oct 15, 2009 5:58 pm 
Offline

Joined: Thu Aug 27, 2009 11:30 am
Posts: 20
hvincent wrote:
Tomas, is it actually relevant that you were the teaching assistant for a vaguely related technology class almost a decade ago? Does this make you somehow more qualified to shoot down suggestions from people who've had current experience with timing systems or have actually talked to people who offer timing services? If you cared that much, you could call up people and ask them questions yourself instead of suggesting that other people do it to save you the trouble.


Yes, because I get paid to figure these things out in real life. I'm shooting them down because things get said like...
Quote:
...you could buy a bunch of those self-calibrating clocks that use some sort of atomic clock and radio/space waves


Fear of hackers messing with the timing system is like worrying about Y2K (remember that?). For all the effort spent trying to hack the timing system, you might as well build a faster buggy. It's that simple! Even if you make it to finals, you still have to beat the person in the other lane. If people are worried about records and such, just figure that the historical times listed 10, 20, 30, or 40 years ago used different timing systems. Really smart people with big PhD's in physics will tell you that time is all relative. Whatever timing system is considered will only be as accurate as at the time it takes the measurement.

What's the race difference between 2:04.999999 and 2:04.999999999999999999999?
It's a rhetorical question, but some smart ass is bound to say "WINNING!".


Top
 Profile  
 
 Post subject: Re: Timing System, Really
PostPosted: Thu Oct 15, 2009 6:13 pm 
Offline

Joined: Wed Nov 05, 2008 1:16 am
Posts: 141
Location: Maryland
Are any of these potential systems any less secure than the system we've been renting?


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 79 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6, 7, 8  Next


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
cron
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group