Go Down

Topic: Dog racing timing (Read 246 times) previous topic - next topic

davavastone

Hi guys,
First of all, hello all, I'm new to the forum and very new to Arduino. I have a reasonable understanding in electronics but need your help in my latest project...

I am part of a running dog group (lurchers) which travel all over the UK doing different tests with the dogs by chasing a lure (plastic bag attached to a nylon cord power by a motor); simulated coursing, straight line running and long jump etc (plus a few others), much like greyhound racing if you will but much more varied and not on a track. The dogs are separated into two groups, over 23" and under 23" (to the shoulder), this works OK, but seeing as there is such a vast range of Lurcher crosses and ages, we have been in talks of breaking the classes up solely by their times against a bogey time for the group/distance (I shan't go too much into it)

Anyway, I have recently completed the prototyping stage of my single dog timer (lasers and LDRs) with threshold adjustment using two Arduinos with Xbee wireless modules so they can be placed at a distance apart in the field with MPH as well as time upon finish (MPH is a bit of a novelty for the dog owners), obviously the major down side to this is that only one dog can be run at a time which has its caveats; taking up a lot of time to do the runs and some dogs don't run so well on their own.

Before now all dogs got through to the finals purely by a knockout scheme running in batches of 4, we'll continue the idea of knockout but time's will be needed by the 4 dogs at a time to keep on top of their classing.

Now, the technology for multiple dog timing

Just some things to bear in mind

  • Needs to be suitable for all weathers
  • the start/finish gates need to be approximately 10-12 meters wide
  • I cannot have anything on the ground (lure and dogs can get caught up
  • Easily portable.


I had in my head maybe omni directional transmitter on the dogs' collars and a directional receiver at the start/finish gates?

Happy to brain storm as much as needed discussing the for/against on all ideas.

Hopefully I've been clear enough on the explanation :)

Thanks all

Dave

DrDiettrich

For proper timing of multiple objects, in arbitrary positions across the gates, tags on the objects seem inevitable to me. The tags should be actively sending signals (beacons), of a unique (distinguishable) encoding.

Ultrasonic modules are discouraged, dogs may hear those frequencies.
Laser beams are too directional, will hardly hit any sensor in a fixed place.
RF is too undirectional, will hardly trigger a receiver when exactly crossing a line.

Remains light, with distinct colors or pulse codes. The dog tags could shine upwards, where a number of directional sensors (in tubes) over the gates can catch the light. The sensors must be fast (no LDR) and sensitive, but insensitive to ambient light.

davavastone

Brilliant thanks, I had considered RFID but with it being so undirectional exactly as you say, I had sort of ruled it out.

Maybe modulated IR for the light? Or will it scatter too much?

GoForSmoke

First, please not lasers where dogs or humans could be blinded in stupid accidents. Please no.

With any kind of light beam break sensor it is the view angle of the detector that makes the difference. You don't need a 'beam', just a restricted view with good contrast between blocked and not.

Have you ever seen those collars that detect an 'electronic fence' buried cable and give them a shock depending on how close they are to 'the line'? That detection system could be used to do something else like bluetooth a time to a laptop.

If the dogs stay in lanes then the sensing could be easier telling them apart, other simpler, cheaper methods might work.

For a physics teacher, such a thing might make a good class project.
Nick Gammon on multitasking Arduinos:
1) http://gammon.com.au/blink
2) http://gammon.com.au/serial
3) http://gammon.com.au/interrupts

TomGeorge

Hi,
The technology you are looking for is being used in motor racing around the world, NASCAR, V8 Supercar, Touring Car, LeMans and F1, unfortunatly it very expensive.

Some examples.

http://blog.atlasrfidstore.com/build-your-own-rfid-race-timing-system

http://rfidtiming.com/

https://www.mylaps.com/en/our-systems

They may need investigating, as your speed is relatively slow, you may be able to copy some principles.

Tom... :)
Everything runs on smoke, let the smoke out, it stops running....

DrDiettrich

@GoForSmoke: Good ideas, except the buried cable is not very portable, and not very precise. But it should be possible to mount the cable above the lines, just like any other sensors.

I found laser pointers (cheap laser diode modules) harmless. I wanted to use one to shoo pidgeons from my balcony, but they only shook a bit when I hit them directly into their eyes. Much less power is required for a light barrier. Nonetheless they can startle the dogs, so that these possibly reject to cross the target line any more ;-)


Also consider that the results should not contradict a finish photo. The times of crossing the lines must be sufficiently precise.

Modulated signals are used in IR remote controls, to distinguish the signals from ambient light. But consider that the transmitted codes must be short enough, so that they can be decoded just when the line is crossed. Decoding signals from multiple tags, just in time, may exceed the capabilities of an Arduino.

Furthermore a mix of signals, reaching one sensor, will be very hard to separate. This would require that the tags send in a synchronized manner, so that only one dog signal is sent at a time. Then the delays between the signals add to the overall time resolution.

It may be easier to let the senders transmit at different carrier frequencies, which can be separated e.g. by discrete band pass filters. Carrier frequencies can exceed 100kHz, so that a couple of frequency bands can be separated with little efforts.

GoForSmoke

I've got a buddy who was swinging his cheap pointer around the room and caught a curved wide-view mirror. After about 90 minutes he got sight back in that eye. You could try it yourself.....

A cable on the ground can work. A flat vertical loop (rectangular) can work even better as the field change will be strongest in the plane of the loop, detect it will a linear Hall sensor and watch for rise to peak and fall means line crossed.
For timers, a loop at the start pulsed once can synch the lot and the fall from peak for each is when to record the finish. Would it matter if the data from each isn't delivered instantly? No, because it is already determined by the collar and won't change until reset. But I would go about getting the collars to detect the field first.

Will the collar always see an IR light? If it's on top in daytime, the sky might flood it out and you'd do better having a shadow across the finish line.
Nick Gammon on multitasking Arduinos:
1) http://gammon.com.au/blink
2) http://gammon.com.au/serial
3) http://gammon.com.au/interrupts

Go Up