I'm working on a project where I want to analyse sprint times in kayak athletes. I used to do this by video-analysis, but this is very time consuming so I need an alternative.
First I was thinking of using a rotary encoder connected to a vacuum cleaner cord retractor mechanism (to keep tension on the rope and pull it back), but the coil spring wasn't strong enough and centering the rotary encoder was quite difficult (it broke from the forces applied due to being slightly off center).
Now I'm looking at other options.
- Measurement distance should at least be 30 meters (starting from the shore)
- The system should be accurate and sensitive. Improving your performance by 0,1 second is already a big step so it should be more sensitive than that. This translates to approximately 0,4 meter at top speed.
- If a rope is needed, it should keep some tension (at least when paddling away from the device) so it doesn't drag in the water and the measurement is more accurate. Not too much tension though, as I don't want to slow down the athlete too much.
- The system should be relatively mobile.
- The less the person measuring has to do, the better.
- It would be nice if the system could view the results instantly, but not necessary.
How would you approach such a project and which components would you use?
Thanks for the ideas and inspiration!
could you put a transmitter/receiver on a fixed location and on the shell ?
you send a signal. it gets received.
it is sent back.
time of flight = distance between points.
Light beam break between two floating (or mounted on poles) buoys?
Physical ribbon between two buoys that trips a switch when you hit it? You could have a pin on one end that is inserted into a switch, that gets yanked out when you hit it.
Do you have control over where the sprint takes place? If you could do it in a long narrow canal or stream or swimming lane, placing light beams and detectors on opposite sides of the canal would be easy. Or ribbons across the stream. Or parallel to the shore? Between docks?
I beleive that the OP wants to datalog the speed of the shell along the way.
each stroke will cause acceleration
each lifting of an oar will cause deceleration
each bad oar rotation will cause race loosing deceleration.
the data of a practice session would be up and down with each stroke.
I would think that water speed may be a better indicator of how the crew are doing.
A ship's log (little spinny propeller thing) will do that just nicely with little drag.
It may be costly to buy one from a marine shop but good water- and salt-proofing is not cheap.
Speed can be calculated real time from the Doppler effect of sound. Put a bull horn on the kayak with a very accurate tone generator. Record the tone with a microphone on the shore. Analyze the frequency.
Use ultrasonic if it bugs anyone.
Put a GPS receiver on each boat, transmit speed/position to the base station.
GPSs can update 10 times/second, should be good enough for a person powered kayak.
Yes a GPS can update quickly, but what is the absolute accuracy of the unit at any time? My handheld unit with a good view of the sky has an accuracy radius around 10 to 12 feet on a good day. It is a few years old, the GPS in my phone commonly puts me on the wrong side of the street depending on the view of the sky, and if I move a few feet suddenly I am on the correct side of the street, and in some cases the accuracy is off by 100 feet in the city. So what happens to the accuracy of the speed etc. if one measurement is + 10 feet and the next measurement is -10 feet??
seems to me some averaging and checking on the accuracy of the position will need to be done. It sounds like the OP needs an accurate maybe every second, I have not graphed the location of a stationary unit using the newer chips, but over time I would expect some wandering.
Some checking would be in order before going too far down building more than one test unit.
Even if the GPS is 100ft off the true position, it doesn't wander very fast. So the next position measured is almost the SAME 100ft off. The speed calculated from adjacent positions is very accurate. I consider it accurate down to about 0.1kph and I'm sure the real accuracy is better.
Thank you for all your answers.
To make the situation a bit more clear, here you can find an example (not my video or test situation, but similar): 25m Canoe polo speed test - YouTube
Unfortunately the situation is that the exact position of the start cannot be controlled exactly and a stationary start is required. The first meter takes a lot of time, so starting 20cm or 1m behind a beam that triggers the start makes too much of a difference. I used to do this sprint test at one location, where I had two poles 35 meters apart with a cord just below the water surface between them at both the start and the finish. I could see where the cord was on the video and approximately determine where the kayaker was in relation to the cord. If the body was slightly in front of the start cord, I would determine the finish time at the moment the body was at the same position in relation to the finish cord.
Now I want to be able to test at different locations without too much setup time. Putting poles in the water or placing buoys accurately on the water (and that they stay there) is too difficult/time consuming.
I like the idea of LIDAR. Considering there's quite some splashing involved, would LIDAR (for example this one: LIDAR-Lite v3 - SEN-14032 - SparkFun Electronics) still measure accurately?
edit: I also cannot control if an athlete paddles in an exactly straight line, although they roughly will.
Anyone that can comment on the LIDAR system? Is the splashing water an issue and is it difficult to target someone at a distance of 30-40 meters?