Detecting timed racing "crack"

General timing system discussions.
User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6856
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Detecting timed racing "crack"

Post by Stan Pope »

This is an exercise to show feasibility of, and study ways to detect, a "crack" for competition based on elapsed time.

Here is the "crack":
1. Competition is based on total (or average) heat times. Competitors run against a variety of other cars. Care is taken that each car runs in each lane.

2. If a wireless signal is received, an interfering device delays the gate open signal by 0.1 seconds, which results in all cars in the heat recording times that are 0.1 seconds faster than actual. (One configuration could be akin to the remote key used for autos.)

Why it works:
Suppose each car runs 4 heats on a 4 lane track, against three other cars in each heat. The "beneficiary" races a total of 12 opponents, each of whom benefits from a 0.1 second reduction in its total time (0.025 second reduction in its average time.) The "beneficiary" gets a total of 0.4 second reduction in its total time (0.1 second reductionin its average time.) This is sufficient to move a car substantially up the rankings!

Why it is difficult to detect:
1. The reduced times have the correct relationship to all of the cars in the heat. The audience can see that the winner of the heat had the fastest time, and so on through the rest of the cars on the heat.

2. The audience can not observe the 0.1 second accross the board "error" in a heat's timings.

3. The interfering device can be located anywhere in the circuitry between the starting gate switch up to (and including) the interior of the timer control box. If a computer participates in the timing, there is that much more exposure, since program hacks are often difficult to detect.

4. The interfering device can be quite small and easy to hide.

5. Analysis of heat times after the race could suggest, but not prove, interference.

Questions:
1. How small could such a separate device be, using off the shelf components?

2. How can the race organizer and the audience assure themselves that there is no such interfering device present?

3. What measures should a race organizer take to eliminate such possibilities.

4. Could the operator of a manually operated gate accomplish the same effects by altering the rate at which the gate is opened?

5. If statistical analysis of heat results suggests interference, what do you do about it?

I'm sure that there are some other questions, so chime in, please.
Stan
"If it's not for the boys, it's for the birds!"
User avatar
terryep
Merchant
Merchant
Posts: 423
Joined: Sat Aug 09, 2003 8:13 pm
Location: Fredericton, Canada
Contact:

Re: Detecting timed racing "crack"

Post by terryep »

I'd suggest a redundant time measuring system. A laser start/finish gate measuring the shortest elapsed time (beams across the track, not individual lanes). This time to be compared to the fastest time measured by the main system. Variances, if any, would need to be accounted for. If the laser were more reliable its time could be used to rank the results.

Terry
ExtremePWD
Master Pine Head
Master Pine Head
Posts: 331
Joined: Sun Nov 16, 2003 5:04 pm
Location: Illinois

Re: Detecting timed racing "crack"

Post by ExtremePWD »

I had the same thoughts as to using a redundant system. But since the question has a way out assumption (though very possible) that the system is wireless, if the redundant system is also wireless then it could also be tampered with. A wired system eliminates the wireless tampering. The switch and detection devices should be sealed with tamper detection devices from the factory. The pre-race procedure should offer inspection of the tamper detection devices by anyone present. (A single official in charge of tamper detection could be part of the scam) The wires should be permanently connected to eliminate a connector where an interrupter device could be easily input. If possible tampering is detected then the race format should default to a points system. However, it would still need to be confirmed that the tampering did not result in software modifications that would favorably reverse the finish order of a predefined car in races that are too close to call by the human eye.
User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6856
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Detecting timed racing "crack"

Post by Stan Pope »

ExtremePWD wrote:I had the same thoughts as to using a redundant system. But since the question has a way out assumption (though very possible) that the system is wireless, if the redundant system is also wireless then it could also be tampered with.
For clarification, the timer system is assumed bo be closed (wired). It has been breached by inserting a device that might be controlled wireless. After posting, I realized that the interference could be controlled by wire, as well, but that wireless would be more easily concealed.
ExtremePWD wrote: If possible tampering is detected then the race format should default to a points system. However, it would still need to be confirmed that the tampering did not result in software modifications that would favorably reverse the finish order of a predefined car in races that are too close to call by the human eye.
Good points! Die hard "clock racers" wouldn't like the fallback, but ... :)

The software hack that recognizes and manipulates "too close to call" results is another risk and affects both timed and finish order electronics! Good catch! Can we treat this in a separate thread?
Stan
"If it's not for the boys, it's for the birds!"
User avatar
terryep
Merchant
Merchant
Posts: 423
Joined: Sat Aug 09, 2003 8:13 pm
Location: Fredericton, Canada
Contact:

Re: Detecting timed racing "crack"

Post by terryep »

Stan's correct, any system can be tampered with, I'd ask the racing official to inspect the start switch and wiring back to the finish line device for added "black boxes". The finish line itself could be inspected a few days before the race and sealed or replaced with an "official" unit supplied by district before the race.

As for the PC and it's program. That should be supplied by district as well.

Our districts judge by placement not time so I have no first hand knowledge of these elapsed time systems only what I've read about them. However, electronics design is my profession. There are plans on the web for crude timing systems that connect to a PC's parallel port. Their accuracy will depend on the PC's interrupt latency (if interrupt driven) or the loop time of the program if the port if polled., both depend on the PC hardware and operating system performance. Commercial systems do all the timing local to the track and report the raw results back to the PC through a serial connection. A printer could be tapped into this line to record raw results for later verification. (To spot tampering with the results at the PC)

I prefer placement judging over times any day. Could we get a poll?

Terry
ExtremePWD
Master Pine Head
Master Pine Head
Posts: 331
Joined: Sun Nov 16, 2003 5:04 pm
Location: Illinois

Re: Detecting timed racing "crack"

Post by ExtremePWD »

What if the "official" is the corrupt person in the loop? An audit committee could force more eyes on the audit making it more difficult to execute the corruption. Your suggestion that timers be provided by the district raises an interesting idea. If a district standardized on a timer system and the packs only had to install finish line sensors and a starting switch, then the timer control box and software could be rented or loaned out to reduce the total system cost to an individual pack. As you suggest, keeping control of the primary timing unit outside of the local group can help protect against corruption. On a side note, I have used a timer diagram from the web combined with a very efficient program (few lines to loop through) running in DOS with a 1+GHz processor. It polls the sensors each loop which results in a very fine resolution. The sensors (infrared led's and phototransistors) are probably the true accuracy limit. This could be another topic.
User avatar
gpraceman
Site Admin
Site Admin
Posts: 4926
Joined: Fri Jun 20, 2003 12:46 am
Location: Highlands Ranch, CO
Contact:

Re: Detecting timed racing "crack"

Post by gpraceman »

I think the scenario of someone slipping some sort of electronic device onto the timing system's circuit to control the start signal and thus influence the race results to be highly unlikely.

Knowing the design of many timing systems, home built and commercial, I'd say the only likely places such a device could be located is at the start gate or at the timer itself. Those are the only locations where you may have start switch connection points to hook up into. At the timer, it would likely be very visible to everyone. At the start switch, it would be easier to hide, but could easily be seen, if someone simply checked the timer connections before racing started.

It would be hard to slip a device onto the track after it is setup and then remove it before it is taken down, all without anyone seeing. The person would have to be very familiar with the timing system connections, and understand the basic operation of the timer in order to achieve the desired result. They would also need to know how to build such a device.

If anyone would be guily of such shenanigans, it would more likely be a race official. If you can't trust your race officials, then they need to be replaced, or maybe you shouldn't be having a race.
Randy Lisano
Romans 5:8

Awana Grand Prix and Pinewood Derby racing - Where a child, an adult and a small block of wood combine for a lot of fun and memories.
User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6856
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Detecting timed racing "crack"

Post by Stan Pope »

I concur with Terry on preferring finish order scoring rather than time. To me, timers are best used when everyone using it wants it to be exactly correct, such as when a car is being tuned before the races.

It is unlikely that a district or council committee (around here, at least) would take on an electronics management job. It is too far away from the council's primary mission, and, anyway, each unit chooses its own racing format. This means that each unit that chooses to "race the clock" has to deal with the issue.

Some of the original questions haven't been answered.

How big? I suspect that a dedicated amatuer could get this down roughly to the size of a black three-connector plug, around 0.25 X 0.75 X 0.75". It is an active device, so size is limited by need for a battery. (Good assumption?)
Stan
"If it's not for the boys, it's for the birds!"
ExtremePWD
Master Pine Head
Master Pine Head
Posts: 331
Joined: Sun Nov 16, 2003 5:04 pm
Location: Illinois

Re: Detecting timed racing "crack"

Post by ExtremePWD »

A person constructing such a device would have knowledge of the timing system that it would have to interface with. With this knowledge he could tap into the power source for the timer and possibly avoid the need to package a battery. The only size reference I have is the wireless PCMIA adapter in my laptop. The portion sticking out of the computer, which I assume is the antenna for send and receive, is 1x2x0.25. The part of the device in the computer is 2x3x0.25 and likely holds additional electronics. Car remotes can be fairly small but their function is to send a constant signal. I am guessing the receiver would be more complicated since it would not only acknowledge receiving a signal but would also need to execute additional activity.

As for a district standardizing and sharing electronics, this comment applies to a general geographic area which may contain multiple districts. In our area there is a person (assumed acting as a private party) that provides a track and race coordination services to many packs that do not have their own track. Taking this to the next level of packs with tracks but no timers, if it could be determined that many packs within a geographic area wanted to contract with a timer/software service then this would be a means to access a timer that does not have local access for tampering. I agree that this is not a suggested business for our scout office.
User avatar
terryep
Merchant
Merchant
Posts: 423
Joined: Sat Aug 09, 2003 8:13 pm
Location: Fredericton, Canada
Contact:

Re: Detecting timed racing "crack"

Post by terryep »

ExtremePWD is correct that the sensors have limitations, if there is a change in the ambient IR the phototransistor's switching time could be effected by 10s of microseconds. A mechanical start switch coupled with manual operation variations could easily vary the race times by several milliseconds. This is not likely to change the placement however.

Terry
User avatar
Stan Pope
Pine Head Legend
Pine Head Legend
Posts: 6856
Joined: Sat Jul 05, 2003 7:01 pm
Location: Morton, Illinois
Contact:

Re: Detecting timed racing "crack"

Post by Stan Pope »

terryep wrote:ExtremePWD is correct that the sensors have limitations, if there is a change in the ambient IR the phototransistor's switching time could be effected by 10s of microseconds.
I wasn't able to find context for this, to see how it relates to the original problem statement.
terryep wrote:A mechanical start switch coupled with manual operation variations could easily vary the race times by several milliseconds. This is not likely to change the placement however.
I'd like to see this tested ... my suspicion is that gate operation variations can cause time variations into the hundreds of ms, which is well beyond the range given in the original problem statement.
gpraceman wrote:If anyone would be guily of such shenanigans, it would more likely be a race official. If you can't trust your race officials, then they need to be replaced, or maybe you shouldn't be having a race.
Valid observation.

Over the years, I've received numerous emails with "horror stories" relating to "shenanigans" by those who organize unit events. The number of occurrences, as a percentage of the total number of races held, is very low, but it does show me that there are at least a few who abuse their position and, in the process, leave a bad taste in the mouth of those who trusted them. I never cease to be amazed at what some folks will do when they think no one is watching. :(

There are lots of ways to do such an ugly deed. Since this one is pretty subtle, I thought to explore it with you knowledgable fellows and to get insight into how it might be detected and prevented. Perhaps by exposing it to light of day (or light of the internet) there will be fewer incidents.

So, thank you for your help, gentlemen.
Stan
"If it's not for the boys, it's for the birds!"
ExtremePWD
Master Pine Head
Master Pine Head
Posts: 331
Joined: Sun Nov 16, 2003 5:04 pm
Location: Illinois

Re: Detecting timed racing "crack"

Post by ExtremePWD »

From what I have witnessed, timed races and manual gates should not be mixed. The worst situation is with a mechanical switch tied to the gate. Even with a laser starting gate a starting official could partially move the gate forward and pause (not releasing the cars) resulting in a reduction in potential energy for that race. Car front end geometry also interacts with gate operation and could lend itself to intentional manipulation of relative car performance.
User avatar
gpraceman
Site Admin
Site Admin
Posts: 4926
Joined: Fri Jun 20, 2003 12:46 am
Location: Highlands Ranch, CO
Contact:

Re: Detecting timed racing "crack"

Post by gpraceman »

ExtremePWD wrote:From what I have witnessed, timed races and manual gates should not be mixed. The worst situation is with a mechanical switch tied to the gate. Even with a laser starting gate a starting official could partially move the gate forward and pause (not releasing the cars) resulting in a reduction in potential energy for that race. Car front end geometry also interacts with gate operation and could lend itself to intentional manipulation of relative car performance.
That may be true for a gate that requires the operator to open the gate fully (spring loaded closed type), but a mechanical start switch paired with a manually triggered spring loaded open gate, should provide a reliably consistent start heat after heat.
Randy Lisano
Romans 5:8

Awana Grand Prix and Pinewood Derby racing - Where a child, an adult and a small block of wood combine for a lot of fun and memories.
ExtremePWD
Master Pine Head
Master Pine Head
Posts: 331
Joined: Sun Nov 16, 2003 5:04 pm
Location: Illinois

Re: Detecting timed racing "crack"

Post by ExtremePWD »

I totally agree with you. My reference to manual gates is that the starting official provides the effort to move the gate to the open position. A spring loaded open gate is the preferred arrangement and shoud be adequately repeatable.
User avatar
terryep
Merchant
Merchant
Posts: 423
Joined: Sat Aug 09, 2003 8:13 pm
Location: Fredericton, Canada
Contact:

Re: Detecting timed racing "crack"

Post by terryep »

I concur there shouldn't be a problem unless there is a lose switch or warn out switch.

Question, I've never experienced a timed derby any comments on the spirit of fairness (perceived) as compared to a placement derby?

Terry
Post Reply