RFID race timing solutions

Hi,

I'm not sure it would be easier with tags. While there are tags that can be read from a distance, they don't provide accurate positioning, so some sort of break-beam type of barrier might also then be needed (I don't know if that's needed for a 3 km race or not).  Failure to read a tag might be quite likely, especially if the runner is going fast - the long-range tags are designed to be on products, not humans.

Personally I think the easiest solution would be video based, perhaps an app on the mobile phones of race assistants. As a person starts or stops the race, they could be identified on the video at the start or stop time, and then the times could be collected up, since the phone will be aware of the time that the video was captured. The race assistant would merely need to go to the position in the video capture where a contestant has started or finished, and tap to select the contestant name, and let the app handle the rest. It sounds like a nice opportunity for someone to write an app for a mobile phone basically, and perhaps a cloud app for uploading the data once there is mobile coverage. 

I seem to recall start / finish line mat antennas being used for this sort of thing. Something like this perhaps:

https://www.youtube.com/watch?v=MnkCDdUjP5w

Looks like these folks have a how to guide on DIY race timing systems which may be of interest:

https://www.atlasrfidstore.com/rfid-insider/build-your-own-rfid-race-timing-system

Hi Dave, good videos. They're definitely the right sort of tags, I've worked with a solution with them on products, for reading at a distance (can even be read from 10 feet away or more, depending on the antenna, and on the tag shape/size), but the system would take several seconds to read them - although that could partially have been be the back-end doing the processing, rather than the RFID reading portion.  

Interesting how they suggest wearing two tags, just in case one fails to be read. I'm still not sure how accurate it is; perhaps for long distance races it is sufficient for determining the approximate time, and then the precision comes from video footage maybe.

Thanks Dave. I've watched the Atlas videos, but most is out of stock at present so was looking to Element but their products seem to require more setup than maybe I can do, which I was hoping someone had done previous. I think I have given up on the idea as its probably a +$1000 investment for a gamble. I'm stuck between being frugal (there is another word for this) but wanting something useable

Having to be thrifty tends to add to the challenge and often results in creative thinking. However you are often just paying for it in other ways like your own time spent developing it.

Cost perhaps depends on what equipment you already have access to. Having access to laptops / smartphones / radios can sometimes appear to lower the cost.

The mat antennas will likely take a bit of a pounding due to repeated physical contact so would need to be robust which tends to bump up the cost if buying off-the-shelf. They are tried and tested by others which saves you time and expensive of doing it yourself. Non-contact solutions may not require to have to be so robust but may take longer to set up per race. 

You also perhaps need to weigh up using the rider carry the timing information to the finish line as opposed to a separate comms system; or redesigning the race such that the finish line ends up close enough to the start line to avoid the need for the long range comms link. Also the cost and complexity of the tag (x100) versus the cost of the receiver (x2) will likely come into play.   

If the majority of the riders are already carrying smartphones then you may find that you already have access to expensive active tags that can be used. During the pandemic some contact tracing systems leveraged this for detecting proximity between people. You may be able to do similar to detect proximity to the start/finish lines. 

Camera-based solutions can work out quite low cost these days if you already have access to most of the hardware. OpenCV software on a laptop with an appropriate low cost USB webcam could be quick to set-up and also be used to capture an image of the rider crossing the start/finish line with even a GPS synchronised clock in the background as well to provide backup evidence. Not so good at the the finish line though if you will have multiple riders crossing the finish at the same time.

There was an attempt at using OCR for identifying racing kayaks here which may be of interest:

/members-area/personalblogs/b/blog/posts/pynq-ocr-frustration

Could replace the number with a QR tag but it needs to be visible to the camera for it to work.

I too think vision-based systems are the way to go, since everyone already carries a ton of computing and camera capability in their pocket.

It's a common problem for a lot of sports, where the ready-made systems cost a fortune, that the amateur clubs cannot afford. I think a vision based system, whether it's fully automated (e.g. using QR codes etc) or partially automated (e.g. an attendant has to manually tag the names to the video with basic sync/clock information from the phone's clock) would eliminate a ton of cost, and could be reliable, since the video recording is there in case things go wrong.

A friend had the same issue with a dog-racing sport, where the amateur clubs were spending about £6k for (I believe) for photocell based systems, i.e. a barrier through which the dogs had to run between. Timing there needed to be very accurate of course, to detect tens of msec-level racing faults etc.

The dog racing application is an interesting example. I was wondered if BLE tags or beacons could be used, by checking ble signal strength (rssi) with possibly direction analytics (if using BLE 5.2 tags) in combination with a photocell based system for final confirmation. This should give you a 1 second accuracy or possibly millisecond accuracy after calibration/testing. Would be fairly low cost to develop.

BLE would be a neat way to augment photocell. I guess antenna tech is still a way behind to provide accuracy on it's own. I've not seen the BLE location solution in action yet, but from other AoA based systems (in similar frequency range), they tended to home in on the position, i.e. from the first set of readings, they knew the approximate position, then from the next reading it got a bit more accurate, and eventually it would locate the beacon-type device to within a patch of a couple of feet diameter, which was still really impressive. I guess it was averaging the location from multiple readings, because of the need to get a good result from many of the antennas in the array. BLE may be much better hopefully, since it is a more recent implementation.

Also for such sports type stuff, since the area that the measurement is required in is very defined, that could make it easier, since the array doesn't need to locate over a very large space! The timing resolution that was needed was 1/100th of a sec at a minimum (that was what the existing photocell system supported) and they would have liked to go to 1msec. These short times may be difficult, I don't know how fast the BLE solutions could provide measurements.

A lower cost interim solution might be to consider using a pneumatic 'road tube' type sensors, similar to the type used for counting traffic on highways or for ringing bells on forecourts, coupled with a GPS receiver module.

That would automatically log the times when a bicycle wheel passed over the start and stop lines, helping to automate the process. It would offer quite flexible installation, be quick to set up, be weatherproof and be compact to transport.

Never even crossed my mind even though I have used these in the past. I'll start googling now. I know the traffic counters are expensive, but the maybe smaller solutions. In fact we could even get away with infra red beam at present for what we are doing. Thanks for this.

Perhaps try something like this ?

DIY Traffic Counter - Road Tube
https://www.youtube.com/watch?v=x1HM3IvExJE

Add a GPS module to give you your real time clock reference.

Perhaps try something like this ?

DIY Traffic Counter - Road Tube
https://www.youtube.com/watch?v=x1HM3IvExJE

Add a GPS module to give you your real time clock reference.