Jump to content

Artificial intelligence at the service of sport and performance ( skier data )


Horton
 Share

Recommended Posts

  • Administrators

I'm curious if anybody knows any more about this French data acquisition system. Having chased the dream of data acquisition in the past, I'm a bit pessimistic but it would be amazing if somebody actually cracked that nut.

 Goode  KD Skis ★ MasterCraft ★ PerfSki ★ Radar ★ Reflex ★ S Lines ★ Stokes

Drop a dime in the can

 

Link to comment
Share on other sites

  • Baller

Around the 5:10 mark, when they claim "til now no one has tried to understand", that's when they lost me. What an ignorant statement, especially from a company that works with sensors - a concept that's been in many, many fields (including skiing) for decades. 

Link to comment
Share on other sites

  • Baller

@HortonI tried to do some digging, but no dice.

@A_B may have some info. In a 2012 post he said "I tried to track down the original LISA unit, as it was called. Talked to Benzel, who gave me the guy's name who designed it. I called him and he said he basically threw it away, as there wasn't any commercial interest in it. This was years ago."

Edited by Broussard

MasterCraft - Still the Leader. And Pulling Farther Ahead.

Link to comment
Share on other sites

  • Administrators

I would think there would be SOME useful data from rope angle data combined with GPS data showing boat position plus rope length . How useful? IDK

I would think you could get skier speed & path from hookup to edge change. After edge change rope tension becomes an issue. These data sets are no where near what I would ask for.

 Goode  KD Skis ★ MasterCraft ★ PerfSki ★ Radar ★ Reflex ★ S Lines ★ Stokes

Drop a dime in the can

 

Link to comment
Share on other sites

  • Baller

I was very interested in this and was also surprised that there wasn't more response on this forum. From my point of view, it was interesting that they seemed to have managed to do RTK GPS on a ski although it's not clear how successful or reliable it was. It's also an expensive receiver to lose if it comes off!

They also seem to be using it in so called "moving base" mode and so it only gives a relative position between ski and the base (which they have in the boat) and therefore no info relative to where the slalom course is.

The data you can get from this has been done before using rope angle and a tension metre on the rope (LISA) and to my mind this combination is probably more useful (from a coaching point of view) than just plotting skier position relative to the boat. Again, I am surprised there isn't a system out there commercially available. If there was no interest at the time, maybe it was because no-one appreciated at the time the possible importance of skier path variance at crazy short line lengths?

I was talking with a camera operator from one of the webcasters who said that initially it was difficult to track Nate with the camera as he was noticeably quicker through the wakes than any other pro, and yet to watch him visually, this is not obvious as his progress from buoy to first wake is soooo smooth. An analysis of his speed/path/line tension and a comparison with other pros would be hugely interesting.

@dchristman A system for detecting whether a ski has rounded the buoy has been suggested to me quite a few times but unfortunately there are huge problems in terms of cost and reliability.

  • Like 3
Link to comment
Share on other sites

  • Baller

@dchristman I didn't but have done some looking. The best accuracy I have seen quoted is 10-30cm which would not be good enough for buoy scoring but in terms of training analysis, it's certainly worth keeping an eye on.

Another tech to watch is LiDAR. The detectors are coming down in price and up in performance.

Link to comment
Share on other sites

  • Baller

@Broussardyes there was a couple Ballers trying to create another LISA at some point but we didn't have the technical skills necessary so gave up.

The benefit of LISA was it showed the position of skier in the course and the load factor on the handle.  This would be valuable when making changes to where and how you resist the boat.  

There was a Spray or WSM article on it, and they compared Andy to some solid tournament skiers, and of course Andy's load was almost equal and peaked at about the same place on BOTH side of the course.  Amazing consistency versus the tournament skiers that showed a decent onside resistance but offside fell way short and longer, like most skiers.   

Someone who still has all their magazines might be able to look for the article and post it.  I think I have given a date range on here before, but no takers on digging through old Mags.   I think Benzel may have written the article, not sure though.

Link to comment
Share on other sites

  • Baller

I spent some time looking at wearable sensors. Both suits and small sensors. There are some waterproof sensors on the market for about $150 a sensor. I think you could get very meaningful data from 5+ sensors. Suits were nearly $10k, but I'd be interested to see the pro data!

I ended up being too cheap and played with pose estimation from videos. While it was more possible than I expected, it'd be WAY quicker to get some data from sensors.

Being able to draw patterns from ankle, hip, shoulder, wrist position/angles along with GPS speed and position data would be interesting to churn through.

Link to comment
Share on other sites

  • Baller

Interesting. I was thinking from a coaching perspective after watching some of the online coaches drawing lines on videos. There is so much to learn from all of it in sure.

If anyone is interested to work in some tech collectively, I'd be interested. I'm a software guy, and would be interested in putting some money in a pool. 🤷‍♂️

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share

×
×
  • Create New...