Slalom Ski (Handle) Path Physics part 2

[lundberg]

Back in February I shared a Matlab model that I made to better understand what was going on at the end of the rope.

@MarcusBrown

@AdamCord

Out of that conversation I had an idea of a simple way to collect real skiing data to compare to the model.  Well the boat is running and the water was nice so I gave it a shot.

(Right up front, if this looks interesting to you message me.  If you capture data, I'll process it for you and make some plots!)

Setup:

1) Pylon camera mount (snug up where it tracks the rope so that it does not have any slop).

2) Phone with an app to capture the Gyroscope and GPS (https://www.tszheichoi.com/sensorlogger)

3) Go ski and capture data, note boat speed and rope length.

 

 

 

My buddy @Rodwishart1 and I went free skiing at 32off / 34.2mph

After some filtering and a little math here is what I got.

You loose a little accuracy at width when the rope goes slack so it will underestimate that part but it tracks real nice once it is moving, allowing you to capture where your dynamics peak;  min speed (agrees with the model showing speeds in the mid 25mph after the turn), peak swing speed (late for my simulated gate and before mid-line once in rhythm), and max speed (upper ~48mph for me). 

It is clear from this data that I free ski narrow and with two slow of a rhythm.  

Here is a plot (not to scale) showing where the different dynamics are happening).

Here is the same plot zoomed in but to scale showing the rope angle to the boat and a modeled handle path.

This is real data.  It would be interesting to use this to see (measure) what skiers do different or when you make a change to correlate what you feel with what was happening.

 

 

Edited May 6, 2023 by lundberg

[MDB1056]

Said same with original post. Please crystallize all this down to summary statements

[dchristman]

I started working on an app a few years ago using this method but the gyroscope drift seemed way too much to make the data very useful.

 

[lundberg]

@dchristman

That looks interesting,  I wonder if the major amount of "drift" you had was related to the camera mount not being firmly connected to the rope.  This feature (hysteresis) makes for smoother filming but corrupts the data that I'm looking for by letting the rope change angle without the phone moving at the same time. 

I took the raw gyro data (radians / sec sampled at 50Hz) and heavily filtered it before I did any integration to get angle data.  I'll collect more data to see how repeatable it is but the two data sets we got this week looked promising.

@MDB1056

With only a phone and a simple pylon mount I can "measure" what is happening at the handle (location and velocity (speed and direction)). 

I need more data to make dial it in.  Who's going to help me?

[Buxrus]

@lundberg:  “who’s going to help me?”  

NASA?   Space X?  Elon Musk?   

Good luck.  

[S1Pitts]

@lundberg I applaud your efforts to bring technology to the waterski arena for the purpose of improving performance.

My son is involved in professional sports (hockey) and they have an entire department and outside contractors dedicated to using technology for advancing player performance. It is another tool to bring out the best in athletes. 

It does take a village so I hope you find people to help you with this.

[lundberg]

I captured another data set this morning.  Lot's of info can be mined from this data.

zoomed in and to scale

[Rodwishart1]

It's remarkable how clean your summary plots are given the relative simplicity of the setup, Andrew. It's also fantastically impressive how well the actual data correlates to your Matlab theoretical work from the frozen months of winter. Your work is inspiring, and the picture of your ski buddy is inspiring in other ways - what a handsome guy!

It's too bad the data points out what he already knows - he's too slow and too narrow (and old?) to have a chance at six balls at that length and speed.  But maybe in July? Hope springs eternal.