Almost three months since we released our teaser video, we’ve come up for just enough air to look back at what we’ve learned and share the path we’ll forge through this data revolution in rowing. We are incredibly excited to be here.

Our mission remains to empower rowers and coaches at all levels with the tools to achieve peak performance on race day, so how do we get there?

What we’ve learned

Going into this, my partner Peter and I already knew how saturated some parts of the wearable device market are today. So many companies now play in the quantified self sandbox, and their devices — heart rate straps and watches attempting to capture more and more information about the world around us — already find themselves wrapped around rowers’ torsos and wrists. While the incumbents like Garmin and Polar are busy competing with Android Wear and the Apple Watch — all tech which focuses on the individual — the technology that empowers teams as a whole is sorely lacking. So guess where the opportunity lies for us?

Along the way we connected with a small group of coaches* here in the Pacific Northwest and learned which data impacts their decision-making and which data, though interesting on the surface, is utterly useless. The variety of answers is fascinating and helped us steer through the vast landscape of technology that we could invest in. To give you an idea of some options, each combination of answers to these questions yield some different technological direction:

While we would love to build everything for everyone, having only two people on our team forces us to focus on our early adopters. Here are our answers right now.

Who’s it for? Live coaches of competitive junior and collegiate teams, who are already using data to drive decisions or desire to use data to drive decisions.

What are we optimizing? Boat speed. Each of the metrics in the table above focus on the individual, and while building sensors for individual rowers or their seats/oars/foot stretchers/oarlocks is on the horizon, even that data needs data from the entire boat (hence, boat speed) to put it in context.

When can you use it? In real-time. Obvious to most rowers is that immediate feedback after a change reinforces good habits and staves off bad ones.

Every coach prides themselves on their technical eye and their training plan, and we eventually envision RowHero as the toolset to evaluate how effective their plan is for every single athlete on their team. First we start by helping coaches give better real-time feedback; then we develop clear metrics on how all of their feedback and their training plan impacts the growth of every athlete, giving them opportunities to course correct quickly when the team or an individual is not making expected progress.

What we’ve done

To this end, our first goal is to identify a crew’s racing weaknesses and minimize or even eliminate them entirely. Most coaches we’ve talked to target goal splits for race pace in calm conditions (some even track standard times for all intensities), and keeping crews accountable to those splits cements the connection between speed and boat feel. So we first wanted to build a way for coaches to see their boat speeds, live, in the launch.Peter and I split up and tackled two different ends of the problem. We had already decided that building our own device was ideal for many reasons (one of the biggest ones was that people still hesitate to bring their phones out in a shell). So Peter started building prototype devices to assess their feasibility as well as understand the constraints we were up against (battery life, storage space, getting data off the device, etc.).

Evolution of devices and proofs of concept. Top Row: Data logging proofs of concept experimenting with different transmission channels (bluetooth vs wifi). Middle Row: Evolution of an “AutoLog” device from left to right, which captures boat data automatically. Eventually will go into a waterproof housing with a rechargeable battery. Bottom Row: Field Logger. Used for the past three months gathering motion data from shells and things that move in them (seats, oars…). Accelerometer and initial orientation sensors on the underside.

In the meantime, I needed to evaluate an experience with coaches as soon as possible, even before we built our own device, so we could use real feedback to guide our own device implementation. To start we chose a straightforward scenario: as a coxswain, I can record a piece and send the results of that piece to my coach immediately we finish, so that they can deliver more precise feedback to the crew.

We prototyped on Garmin’s ConnectIQ platform, available for its latest set of watches. In this scenario:

  1. The coxswain uses a special RowHero app (sorry, not widely available) on the watch to record the time of the piece.
  2. As soon as they hit “stop,” the details of piece are transmitted through their phone to a service called PushBullet. (PushBullet is an easy way for us to send push notifications without building anything ourselves.) The coxswain doesn’t need to touch their phone at all for this to happen.
  3. PushBullet then pushes a notification with the distance, time, and average split of the piece directly to the coach’s phone in the launch.

It’s not rocket science, but it shouldn’t have to be. One of our principles is to build insanely simple experiences for coaches and coxswains on the water. Simple to learn, simple to use, and simple to teach. Both coaches and coxswains still have to steer boats after all.

Perhaps our greatest takeaway is affirming the power of Murphy’s Law: if someone can drop your device in the lake, they will!


Coming into the last part of my training for Head of the Charles this year, I was curious if better data visualization of our race pieces could impact how we train, so I took a week to spin up a site which ingests data from the NK SpeedCoach GPS and plots it on a map, colorizing our path for speed and also providing some summary data on each piece. From looking at recent race pieces, it was very apparent that we had some problems getting back to 100% pressure after turning. (If you’ve ever coxed a blind boat in a head race, you can sympathize.)

See more at http://app.rowhero.com

“But David, didn’t you say you were focusing on real-time feedback?”

Yes, absolutely. We firmly believe analytics, metrics, and visualizations must be more than pretty pictures — they must provide insights that spur the team to action, either through a coach or through an athlete. The question is can we train our technology to detect these patterns and then give us signals in real-time when we repeat our mistakes? For example, if we raced again in practice and our coach saw a signal that our relative speed dropped significantly after coming out of a turn, could they have said something to help us create the right habit of going back to 100% pressure? And these patterns are everywhere: fading at the 1100m mark in a 2k race, being slow off the line on a start, even understanding how different motivation tactics work for different athletes.

What’s next?

Complete our first prototype device. That’s priority #1.

With the early prototypes we’re looking to prove out our concept of taking measurements, converting this to information and getting that to decision makers, and de-risk device related technologies. Starting with measuring shell motion and pushing session data to a to web page for post practice review gives us an end-to-end scenario. Naturally, having a working demonstration is important, but why this particular scenario? It requires us to master a number of technical issues and link them together in concert. The demonstration will provide a great vehicle to carry conversations about needs and requirements forward, while having learned about the technologies required will improve our agility for our next iteration.

The transmission of race piece information is the proverbial toe in the water when it comes to minimizing race weaknesses; there are so many other things we look forward to, like real-time transmission of split and stroke rate and detection of technical problems, no matter how slight, which contribute to decreased speed (e.g. boat set, slumping, timing…). We can’t wait to test common coaching wisdom about good technique against what the data actually says.

This is just the beginning!


If you like what you just read, consider signing up for our mailing list to keep updated on our technology and content to make you and your rowing team faster.

Also check out RowHero for more info on how we help rowing coaches and athletes like you make the most of your valuable training time. We are also on Instagram and Facebook.

Have a great day!

Share on facebook
Facebook
Share on twitter
Twitter
Share on linkedin
LinkedIn

More to explore

Leave a Comment

Your email address will not be published.

ready to be a rowhero?

Fill out the form below, and we'll be in touch within 24 hours on getting your team setup!