Concept2 Forum

I'm a C2 loyalist / a software engineer.
Yet no like the PM5 display. Remind me of my grandpa rowing each time staring at the b/w text.

Official C2 iPad App has to balance between feature rich and coolness, so probably not adding some cool stuff anytime soon.

So I've devoted nights of last month completing this demo App, just a beginning, not much feature-rich. To provide it to C2 pals like me for free forever.

Most of all, I'd like to make Rowdio to be a real everyone's app , so here's an open-minded idea for all of ya:
1. Every C2 user can post your feature request here;
2. Don't mind if your request is silly or not. Etc, you could request to print your name on front-page. (As long as people vote for it)
3. I'll launch a vote before making the next update. Top 3 most voted features shall be included in the update. (Unless the feature is waaay too difficulty to code)

Let's make an imaginative, free, and community-driven C2 app .
Rowdio introduction:[youtube]https://youtu.be/kw8827hO3Ak[/youtube]
Rowing enthusiast‘s review:[youtube]https://youtu.be/jPmVS2lFvWQ[/youtube]

Search the name "Rowdio" in App Store [only iPad]
download link:https://apps.apple.com/app/rowdio/id1644130099

You are competing with the phone based apps that can get data from a pm5. Ergzone, etc. Take a look at what they offer. There are at least a dozen you can find via google.

Look at competing rowing machines. e.g Biorower, nordictrack rower, See what they are displaying. See what's valuable, and if you can extract values from either additional sensors (e.g. like the pods for "running power" repurposed) or by cranking currently exposed PM5 data. (Example, Ergdata creates drive length from PM5 data. Drive length is not a PM5 measured value).

Look at DCRAINMAKERS reviews of sport watches. You are hunting analytics. For example, the GARMIN training score, HRV approximation, VO2 max approx. Very tough to compete directly without these. There are open source frameworks for some of the items.

Look at sites like trainer road. See what they offer for training plans. Then look at BPP, Wolverine plan, etc. Add training plan capability to your app and base your training plan on rock solid science. Then clearly document the science so people can trust your plan (or plan options). Tailor your plan to hours the rower has available, and assume multi-sport. Many of the requests on this forum are for training advice/guidance. Your tool (assuming it has a reasonable data storage capability) has all the data to form an excellent plan and then give feedback and adjust that plan based on actual workouts done. Or your tool could pull data from C2 and do the same analysis.

Decide if you want to feed social sites or compete with them. For example Strava. I'd guess you will not be competing so you need a first class integration.

Lots of opportunity to win. I'd guess the largest opportunity is *not* being a better glass panel (showing more data better), it's in the training plans and tracking analytics.

Some personal preferences. I do *NOT* want to touch a touch-screen while rowing. It's impossible to use a touch screen while rowing so you need to stop rowing to change things, and you get sweat all over everything. Even the PM5 buttons work better. I don't want an app that assumes a large screen while rowing. I prefer browser based output after rowing so I can display it from a desktop on a very large screen. I want customizable, tracked exercise zones. Any tool must be trusted by C2 and allow upload of verified rows.

In any case, good luck.

Probably would not bother myself because everyone wants something different.

The PM5 itself has plenty of grunt and I have said before it should have a completely different second firmware available for those more interested in performance related stuff.

There was a time I was interested in the likes of the monitor giving you a fitness score and running a heartrate decay graph and timer at the end of your row but reality is I have kind of moved on at age 55 now.

Have switched to the bike these days after a foot injury that's taking forever to come right and use a Model B Wattbike with the same type of mono graphics LCD but once you add Zwift it changes everything, the monitor just becomes a device that enables the connectivity and you don't even look at during the ride.

Tsnor wrote: ↑

November 7th, 2022, 3:23 pm

(Example, Ergdata creates drive length from PM5 data. Drive length is not a PM5 measured value).

driveLength is measured by the PM5 (ErgData can expose this as well). It is a UInt8 in the StrokeData structure with 0.01 meter accuracy.

Tsnor wrote: ↑

November 7th, 2022, 3:23 pm

You are hunting analytics. For example, the GARMIN training score, HRV approximation, VO2 max approx. Very tough to compete directly without these. There are open source frameworks for some of the items.

Could you mention some, because I did quite some research and none are rowing specific. I build an implementation which uses two approaches for VO2Max (see https://github.com/JaapvanEkris/openrow ... /VO2max.js), but in all honesty the numbers aren't credible yet: my walking VO2Max is around 44 (Garmin/Firstbeat algorithm), while my rowing VO2Max is around 29. Might be that my technique is bad/ineffective but I don't think its that bad. So any pointers would be greatly appreciated.

Tsnor, if this forum has upvote, that’s max upvote for your comment. Thank you for the guidance.

BG fyi. This app is made at night spare time. I work as an iOS software engineer at a megacorp. Being a routine indoor rower myself, I figure might do something new with iPad. I checked C2’s official dev manual and found it is fantastically well written and organized. My intention with the App Rowdio is to make a free app incorporates the community’s wisdom that eventually benefits some people including myself, utilizing limited dev capacity I have in spare time. C2 console is classically designed and well tested for many years. I find it joyful in everyday’s use, so questions arises what I could do beyond that, while not to the extent of a complicated app which requires commercialisation to support.

Tsnor wrote: ↑

November 7th, 2022, 3:23 pm

Look at competing rowing machines. e.g Biorower, nordictrack rower, See what they are displaying. See what's valuable, and if you can extract values from either additional sensors (e.g. like the pods for "running power" repurposed) or by cranking currently exposed PM5 data. (Example, Ergdata creates drive length from PM5 data. Drive length is not a PM5 measured value).

As JaapvanE pointed out, the length is collected in PM5 and could be displayed. I find myself hard to relate drive length to performance maybe plotted the correlation in chart would find out any connection.

Tsnor wrote: ↑

November 7th, 2022, 3:23 pm

Look at DCRAINMAKERS reviews of sport watches. You are hunting analytics. For example, the GARMIN training score, HRV approximation, VO2 max approx. Very tough to compete directly without these. There are open source frameworks for some of the items.

Look at sites like trainer road. See what they offer for training plans. Then look at BPP, Wolverine plan, etc. Add training plan capability to your app and base your training plan on rock solid science. Then clearly document the science so people can trust your plan (or plan options). Tailor your plan to hours the rower has available, and assume multi-sport. Many of the requests on this forum are for training advice/guidance. Your tool (assuming it has a reasonable data storage capability) has all the data to form an excellent plan and then give feedback and adjust that plan based on actual workouts done. Or your tool could pull data from C2 and do the same analysis.

Thanks for the list. Will check out. VO2 Max and HRV could be measure through Apple Watch,however syncing with ipad is rather difficult than iPhone. VO2 is better calibrated and recorded via pro mask for precision.
Training plan noted. For now try best to pull, organize and analyze existing data from C2. The plans in this community and you mentioned are incredibly professional and I’ll see if can incorporate any with permission in the long run. BPP is a good one. Maybe a calendar/checklist/plan stuff so no need recording manually.

Tsnor wrote: ↑

November 7th, 2022, 3:23 pm

Decide if you want to feed social sites or compete with them. For example Strava. I'd guess you will not be competing so you need a first class integration.

Exactly, not any social elements in Rowdio thus far, no account required etc. I personally like “lite” social e.g. sharing my pb as an exported file easily via iMessage or WhatsApp to another friend and showing like a ghost rower. if such idea interesting for you.

Tsnor wrote: ↑

November 7th, 2022, 3:23 pm

Lots of opportunity to win. I'd guess the largest opportunity is *not* being a better glass panel (showing more data better), it's in the training plans and tracking analytics.
Some personal preferences. I do *NOT* want to touch a touch-screen while rowing. It's impossible to use a touch screen while rowing so you need to stop rowing to change things, and you get sweat all over everything. Even the PM5 buttons work better. I don't want an app that assumes a large screen while rowing. I prefer browser based output after rowing so I can display it from a desktop on a very large screen. I want customizable, tracked exercise zones. Any tool must be trusted by C2 and allow upload of verified rows.
In any case, good luck.

Well noted and yes, disrupt rowing to touch screen breaks all the record and rhythm. That’s one of the primary reason I start making this new app from scratch, to find a better to integrated many data in a beautiful way, instead of switching any pages or pressing any button. Well said though, it is quite a tough quest, indoor rowing can get professional and deep and many rock solid science data, which to export and which to show, tough calls. Please help with more critics no less.

Carl Watts wrote: ↑

November 7th, 2022, 4:43 pm

Probably would not bother myself because everyone wants something different.

The PM5 itself has plenty of grunt and I have said before it should have a completely different second firmware available for those more interested in performance related stuff.

There was a time I was interested in the likes of the monitor giving you a fitness score and running a heartrate decay graph and timer at the end of your row but reality is I have kind of moved on at age 55 now.

Have switched to the bike these days after a foot injury that's taking forever to come right and use a Model B Wattbike with the same type of mono graphics LCD but once you add Zwift it changes everything, the monitor just becomes a device that enables the connectivity and you don't even look at during the ride.

I agree with the need for expansion of PM5. iPad's screen size and processor power have the potential to be "the second firmware".

C2's logbook collects critical part of a generous data-set PM5 collects. By rearranging and re-design of data could potentially maximize the effect of data on sports performance improvement, such as Strava.

Zwift definitely makes rides more fun, which I use from time to time as well. It definitely inspires me the importance of mixing fun part of sports with professional data analysis. Thx for sharing.

TheFleXing wrote: ↑

November 8th, 2022, 10:34 am

I find myself hard to relate drive length to performance maybe plotted the correlation in chart would find out any connection.

Drive length is interesting, as it shows how long the drive was. But I typically use it for post-training analysis. However, (a lack of) consistency is interesting as real time feedback, as is the consistency of the drive speed (i.e. drive distance divided by drive time).

TheFleXing wrote: ↑

November 8th, 2022, 10:34 am

For now try best to pull, organize and analyze existing data from C2.

And record it and export the data. RowsAndAll has a decent environment to analyse data and uses a pretty decent input format (csv, including force curves). I developed the output generator for OpenRowingMonitor, and it is pretty straightforward to generate (see https://github.com/JaapvanEkris/openrow ... ecorder.js). Not many apps allow detailed capture of all PM5 data.

JaapvanE wrote: ↑

November 7th, 2022, 5:58 pm

Tsnor wrote: ↑

November 7th, 2022, 3:23 pm

(Example, Ergdata creates drive length from PM5 data. Drive length is not a PM5 measured value).

driveLength is measured by the PM5 (ErgData can expose this as well). It is a UInt8 in the StrokeData structure with 0.01 meter accuracy.

Tsnor wrote: ↑

November 7th, 2022, 3:23 pm

You are hunting analytics. For example, the GARMIN training score, HRV approximation, VO2 max approx. Very tough to compete directly without these. There are open source frameworks for some of the items.

Could you mention some, because I did quite some research and none are rowing specific. I build an implementation which uses two approaches for VO2Max (see https://github.com/JaapvanEkris/openrow ... /VO2max.js), but in all honesty the numbers aren't credible yet: my walking VO2Max is around 44 (Garmin/Firstbeat algorithm), while my rowing VO2Max is around 29. Might be that my technique is bad/ineffective but I don't think its that bad. So any pointers would be greatly appreciated.

You know this area much better than me. For some reason I thought that ergdata was creating the drive length (and Ave. Force, Drive Speed, Peak Force). I am sure you are correct that these are native PM5 fields.

For open source analytics I was thinking the Vo2MAX and HRV code stacks that are out there. I know of no specific projects that you should look at.

FWIW I've been tracking HRV two ways now for over a year and have found no use for it (either daily or clumped into weeks or viewed graphically over time). There are many different reported numbers for HRV (e.g. Time domain results and frequency domain results, short term and long term). Fitbit and Oura use RMSSD for their tracking. My fitbit charge 3 reports about 1/2 the HRV score reported by EliteHRV as its overall score, but is consistent with EliteHRV measured short term RMSSD HRV. The various measures are not comparable and move in different ways. Here's an excellent ref for the metrics and when they are useful: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5624990/

As I understand it you do not track HRV specifically during an exercise - For short term measurements you pick a consistent time (like when you get up) and a consistent posture (e.g. sitting, lying down, standing) and you track enough data to get a base line for you (hopefully over a consistent workout load period of constant health, stress, etc) then you use variations from the baseline to get insight in autonomic recovery, then you use that to figure out how hard to workout. Its normal for wearables to do this tracking while you sleep (Fitbit also uses it to help determine sleep zones). So your 44 vs 29 could be different metrics, or they could be the results of different TOD, posture, heart load, etc. In any case, it hasn't worked for me (or it was working and I ignored it because it was not useful).

Tsnor wrote: ↑

November 8th, 2022, 12:53 pm

You know this area much better than me. For some reason I thought that ergdata was creating the drive length (and Ave. Force, Drive Speed, Peak Force). I am sure you are correct that these are native PM5 fields.

Well, my co-developper implemented the entire PM5 spec, I only fill the fields that we can create as well. In general, the PM5 calculates, and ErgData just presents it to the user. I must say, the PM5's Bluetooth implementation is extremely elaborate: the BLE-FTMS profile (Fitness Machine Service, with a specific profle for rowing), provides way less data.

Tsnor wrote: ↑

November 8th, 2022, 12:53 pm

For open source analytics I was thinking the Vo2MAX and HRV code stacks that are out there. I know of no specific projects that you should look at.

The real VO2Max is a simple but extremely tough test (you need to achieve maximum HR). Most people will not want to do that.

Concept2 has a publicly published algorithm (see https://www.concept2.com/indoor-rowers/ ... calculator) which uses your best 2K. For me, that score (39.3) was pretty far of my VO2Max at the time (47). And I don't row a "best 2K" on a frequent enough basis to being able track my progress.

An alternative approach is to use suboptimal measurements (see https://assets.firstbeat.com/firstbeat/ ... 6.2017.pdf, and https://www.frontiersin.org/articles/10 ... 1/full#B12), where typically the power is extrapolated to the power at maximum HR, and then the VO2Max is determined (see also https://sportcoaching.co.nz/how-does-ga ... e-vo2-max/). I tried this aproach, and the results are relatively consistent, but too low when compared to the Concept2's approach and Garmin's data.

Tsnor wrote: ↑

November 8th, 2022, 12:53 pm

FWIW I've been tracking HRV two ways now for over a year and have found no use for it (either daily or clumped into weeks or viewed graphically over time). There are many different reported numbers for HRV (e.g. Time domain results and frequency domain results, short term and long term). Fitbit and Oura use RMSSD for their tracking. My fitbit charge 3 reports about 1/2 the HRV score reported by EliteHRV as its overall score, but is consistent with EliteHRV measured short term RMSSD HRV. The various measures are not comparable and move in different ways. Here's an excellent ref for the metrics and when they are useful: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5624990/

My Garmin HR monitors unfortunatly don't report HRV (only HR), but this data indeed does seem interesting. But it could be that it isn't defined during sports activity.

Another we track (in OpenRowingMonitor V1Beta), is Heart Rate Recovery: after 1, 2, and 3 minutes after rowing, we record the heartrate (see https://www.livestrong.com/article/2608 ... -exercise/). It can easily be measured using the PM5 s well, although you need to record it by hand (which I typically forget to do as I typically am catching my breath). It is quite a stable pattern, and it is a decent indicator for the heart function.

Gentlemen, please allow me to introduce a new prototype feature called After-glow in Rowdio v1.0.2 Update from Apple App Store.

What it does is overlaying multi force curves on on screen, highlight new one, old ones fade away.
It is set to record the past 2mins curves and each curve slowly fading away in 2mins.

Problem trying to solve here is have an overview of my current curve compared so I know if my fatigue is kicking in and my curve is out of shape.




Let me know if you find this useful, or not.

How would you design it?

With criticism/coaching. Those curves are not ideal and could lead to injury.

jamesg wrote: ↑

November 14th, 2022, 2:21 am

How would you design it?

With criticism/coaching. Those curves are not ideal and could lead to injury.

Hi jamesg, these curves are simply overlaying every one you'd find in PM5 console, that said, a pure raw data display with historical comparisons without any suggestion.
I do have questions of improving curves/gestures in my own training, looking for right curve to correct. Many in this forum used curve data and plot to correlate with performance.

overlaying every one you'd find in PM5 console,

Exactly. The late peaks shown here indicate faulty technique, presumably why C2 bothered to put the force curve on PM in the first place.

Since in rowing we can't do anything except work, indices such as W/rating and m/stroke coud help see if we are.

jamesg wrote: ↑

November 14th, 2022, 7:14 am

Since in rowing we can't do anything except work, indices such as W/rating and m/stroke coud help see if we are.

Agree with James on those two metrics.