The road to a high-performance algorithm for Swimtraxx One.

The information obtained with Swimtraxx One is highly valuable for coaches. The intelligence inside the system provides coaches and athletes with an accuracy down to tenths of a second. In this blog post, we give some insights into how we achieved those accuracy levels.

Swimtraxx One quantifies 9 parameters essential to monitor in swimming and share them in real time with the coach. First, we measure heart rate, allowing you to keep track of the internal load of the body during workouts. Additionally, we detect what stroke you are swimming, how fast you turn, how long you stay underwater, we time your laps, calculate stroke rate, stroke efficiency and finally we count strokes and breaths.

Swimtraxx One is created after many years of scientific research. This to obtain the best possible accuracy using cutting edge algorithms combined with artificial intelligence and state-of-the-art machine learning techniques. At the foundation of the Swimtraxx algorithms, there are large data sets that are obtained by multiple sensors we have built into the system. Those “sets” you can take literally as we have collected data while swimmers from different levels were working out. We have been processing the raw sensor data by making signal annotations, performing video footage analyses and using multiple data collection protocols. This has resulted in high performance algorithms that analyze all sensor data in real-time while swimming. Every stroke, breath, and lap Swimtraxx One gets smarter. The data obtained by Swimtraxx One is transmitted to a tablet, giving coaches live insights into their swimmer's performance. This allows them to fully focus on the swimming technique of their athletes instead of losing precious coaching time while gathering all data manually.

The placement of Swimtraxx One also plays an import role in the accuracy level of certain parameters. Research taught us that positioning Swimtraxx One on the temple is the best location to measure all of the above-described parameters. Let us take the identification of breathing patterns as an example. When breathing, a swimmer makes a very specific head movement. Where better than any other place to identify that specific movement than on the head itself. This is only one example of why our symmetric shaped tracker is mounted on a swimmers very own and favorite goggles, and thus his head. Imagine we would try to measure breathing patterns with a device attached to the wrist or ankles.

To end this blog post we have a small disclaimer for you: head banging while swimming might confuse our algorithm and will not be recognized. So, keep that head straight and your technique flawless. 😉