The importance of accurate measurements

[zapta]

Just watched the olympic Men's 1500m speed skating on TV.   

Brodka (Poland)  1:45.006 vs. Verweij (Nederland) 1.45.009,  and this is in two different races, not side by side. 

By my calculation this is a 0.003% difference.

[EEVblog]

Foerunately tiem is one of the most easily accurately measurable quantities.

[SeanB]

Time is easy to get good accuracy in relation to human perception, with very high resolution mostly limited by the sensors used.

[Rerouter]

so as rare as it would be what actually happens if they tie to the millisecond?

[zapta]

so as rare as it would be what actually happens if they tie to the millisecond?

It becomes a race condition

Sent from my Nexus 5 using Tapatalk

[peter.mitchell]

I expect something like this would be done without budget - or with a budget high enough that it doesn't matter; as such, not only would the time measurement be easily more than accurate enough, the sensor response times would be characterized, the transmission latencies, the propagation delays between sensors, you name it, it's compensated for.

[T3sl4co1l]

Such times are meaningless: what's the variance over the last 20 or 100 races from each athlete, now them's some statistical goodness!

Tim

[Len]

so as rare as it would be what actually happens if they tie to the millisecond?

It happened in one of the skiing events (but they were only scoring with 0.01 s resolution). They awarded two gold medals and no silver.

[zapta]

It happened in one of the skiing events (but they were only scoring with 0.01 s resolution). They awarded two gold medals and no silver.

With large enough resolution, even I can score a gold.

[ConKbot]

How do they detect when the athlete crosses the line? Seems like if you had the right shaped helmet or a glove with slightly longer fingers you could gain some advantage.

A glove with extended fingers would do a lot more for a swim race then let someone touch the sensor at the end of the pool faster and I'm guessing it would be strongly frowned upon 

[zapta]

How do they detect when the athlete crosses the line? Seems like if you had the right shaped helmet or a glove with slightly longer fingers you could gain some advantage.

In the photo finish, they show the front of the skate touching a line on the ice. Possibly this is the end event they measure.

[Galenbo]

I once had 48 hours to make a working time-measurement for a 400m sprint, oldtimer event with 5000+ visitors. The event was announced a year before, but only 48h before they discovered the guy responsible for the time-measurement "disappeared"

So I made it. Worked. All the results, I think 3x 50 racing cars, were saved on a 3.5 floppy, and after programmatical sorting by time column, I discovered there were multiple results like 12,158203125 and 11,6328125

Hmmm, some after-editing was needed before printing the results :-)

[tszaboo]

That is 30ppm. If I would do it, I would go for the manufacturer of the crystal oscillator to tell me the tempco of the crystal. Just sayin.
Although, how do they guarantee that the size of the track did not change if there is a temperature difference?

[zapta]

WHat I mean is how do they sense the athlete/car passing the finish line. Obviously some kind of IR beam break isn't going to work for multiple objects. I thought maybe they used cameras but the times appear instantly on the score boards so that people can see if a record was set, with no delay.

Does the time appears instantly only for the first skater? If so, it may still be a beam.  However, they still need accurate time for other skaters as it may affect their rank.

[SeanB]

Generally for that you have a line scanner above or next to the finish line that detects the crossing of the line by the first, and another camera that gets an image from the front that can be used to determine the athlete. Common on race tracks where they use a line scanner that is used to make a rolling frame of the horses  so you get a picture of the horses as they pass the finish line with the winner in the front. In the older days they used a roll of film that was run slowly past a slit to only expose the film with light from the finish line.

[Galenbo]

... I thought maybe they used cameras but the times appear instantly on the score boards so that people can see if a record was set, with no delay.

Camera above with vision software is possible.
The images are processed in real time, National instruments and Cognex are known brands for this.

Once much accuracy and precision are needed, prices rise fast.