TraceUp GPS Unboxing

[Horton]

Get yours! http://www.traceup.com/

https://youtu.be/4MNdNkKqecw

 

 

Sample Exports The data in this XLS sheet shows speed only- Not postion

Use this code at checkout for 20% off: BOFSPRAY-1

Get yours! http://www.traceup.com/

 

https://youtu.be/fo9j1HEVlsY

[Waternut]

I have one of these that I used for snow skiing and the Trace Snow app. Part of the reason I bought it was to track waterskiing but in reality, I don't think I have a good enough grasp on the data to understand what I'm looking at.

 

I think I get that column 1 is speed in mph. Is column X total distance skied in inches and column Y total distance about the courses centerline? Did you have to do any data manipulation to get these numbers?

[bbirlew]

@Waternut,

X is distance traveled down course (in centimeters) with -5500 being the 55's and 0 being the enterance gates.

Y is cross course distance (again in cm).

Yes there was a bunch of data processing done to convert this data from the gps coordinates you get from trace into xy data. I've written an android app that will import the trace data from their data server and (among other things) export the data to excel for @horton.

I'll do a write up on the app in the next few days, but am happy to answer any questions you have.

Cheers,

Brad

[cvaught]

Any chance you might consider open sourcing the Android app and placing it on Github?

[Horton]

Soooo to be clear @bbirlew is the man behind that beta data accusation exports. Without his work I would still be looking at video speedometers.

 

[bbirlew]

@cvaught,

I'm certainly open to collaboration on it and will answer any and all questions about the app.

 

I have given a fair bit of thought to open sourcing it and that is somthing I will consider.

B

[Waternut]

@bbirlew Cool. Centimeters makes sense. Is that app ready by chance or is it still in development?

[ktm300]

How does it sync with and overlay the video? How accurate is the speed data? Unless I'm missing something, Cole is going 50mph at 1 ball (hey, nice turn). Horton 25mph at 2 ball. I am not just being a critic. I really hope this works and if it does, think it will be more useful than a thousand conversations about "carrying speed through the turns", fast this and slow that.

[Horton]

@ktm300 the time sync for video often has to be corrected. It is cool but i do not think it is nearly as valuable as chart data.

 

@Waternut there is no production app. We shoehorned my data from @bbirlew's other project. We need someone to learn from @bbirlew's efforts and run with it. I suspect the programming needs to basically start from scratch.

[Horton]

@ktm300 we think the speed data is good. If nothing else it is consistent. I had recorded from inside the boat and it matches ZO

[cvaught]

@bbirlew If you can share what you have I would be happy to look at it and see if there is anything I can do to move the project forward. Are you only able to extract gps coordinates versus time? Looks like the device also has an accelerometer, gyro, etc. Is trace supporting the data extraction via an API or did you just figure it out? Or is it a manual process? It appears they have dedicated apps for other sports to generate summary and detail information about your activity. The API then gives you access to the results but I don't see anything that discusses how to obtain the raw data from the device programmatically.

[303Skier]

Be cool to see a pro like @CParrish43 using one of these. 32-41 off.

[303Skier]

One thing I've noticed from watching the videos is the difference in cross course speeds generated 34 vs 36. Looks like @Horton tops out at about 48 mph and the other skier at 36 tops out at 58 ish. Almost a 10 mph difference!!!

[Waternut]

@cvaught has a good point. It would be nice to see accelerometer data. Might be useful in determining where your cross course angle is poor and where you're over cooking your turns and losing angle. For example a steady increase in acceleration would lead me to believe it was a clean turn and a good pull through to the wakes. Some jittery numbers show hitting the wake and a steady deceleration shows a good preturn. However, a hard jump in acceleration followed by a drop and with more acceleration shows the turn was too hard and you were forced to back off.

[303Skier]

Looks like an interesting read @Mark_Matis! Looks like the author is from Colorado, cool!

[Horton]

@303Skier Data from the videos is not granular enough. The raw data shows I go 53 - 54 mph into one ball. IDK the speed for a 36mph skiers

[bbirlew]

@cvaught & @Waternut,

Accelerometer data is not available.

 

A bit of background:

I started writing my app general enough to receive, parse and plot data from all kinds of sensors. I envisioned the obvious position, speed and acceleration, but also wanted to support plotting strain gauge data, ski angle and rotation, etc.

 

The skier can input data about the set (speed, rope length, zero off setting, notes, scores, etc) and the app will save that data with the set. You can then view and compare plots between passes in a single set, or between passes from different sets.

 

You can also import and export the data to share (email your latest set to a coach to analyze, or import data from a friend or pro to compare and overlay with your data).

Because it logs your ski and fin settings with each set you can also use the app as a training log book of sorts.

In the future I had envisioned incorporating video as well but have not had time to play with that.

 

@horton and I had some discussions a while back about what we could do with the data from the trace. I offered to support the sensor in my app (through connecting with the trace server through their API) and import the trace high resolution position and speed plots into my app. It quickly became clear that the data plots looked great on my full size tablet, but not so great on hortons phone!!! He was more interested in a desktop app, but I didn't have the time to write one for him. For the time being, we settled on adding the ability to export excel files from my app that john could play with and investigate.

 

As john mentioned, using my app is a bit cumbersome currently. For example, In order to plot the passes on an x/y grid from the gps longitude and latitude coordinates I have to know where the slalom course is. I had imagined mapping a course similar to zero off. However the gps data from the tablet wasn't accurate enough to make that function well. So for now you need to find your course on google maps and grab the lat/long coordinates from the right hand entrance and exit gate balls and input them into the app.

 

Give me a few weeks and I'll see what I can do about getting the code up on github. In the meantime I'm happy to answer more questions you have about it.

Thx,

B

[sailworks]

Hello all

Just stumbled on this thread having seen ad for Trace. I think this data - if mapped to course and granular enough would be fantastic help for those of us interested in dynamics and physics of what's happening on the course. Any follow ups to efforts to tease out and display the information of interest??? Any alternative hardware that is competing??

 

[sailworks]

Looked at and down loaded video of Horton demo trace up and video - looks pretty sync to me. Can you point me to a web site or mention is few words as to how you produced video with superimposed / overlay of path on course and speed!??

Thanks

[Horton]

The video sync is automatic but sadly the sync is just not accurate enough to be helpful. 1/2 second off might as well be a minute.

[DefectiveDave]

Even with the bad sync I still kind of want to make this work. Is it possible to enforce an offset for the video timing before making the overlay videos?

 

Perhaps we could use a secondary device to determine the approximate offset by say, jolting the accelerometer of the trace while simultaneously flashing a LED visible to the camera? Such a device would be quick and easy to build.