High-performance AI swim app

Leading international swimwear brand Speedo wanted to launch their next-generation swim tracking app called ‘Speedo On’, in time for it to be pre-installed on Samsung’s new Gear Fit2 Pro activity tracker.

However, swim tracking is extremely challenging because of the wide variety of swim techniques and strokes. Lap counting is also complex for a wrist-based device as it needs to differentiate between normal swimming and the hand movements that occur during a turn.

Our Approach

Speedo collected a large data set from swimmers using a worldwide team that worked around the clock. Then our data scientists deployed a mixture of analysis techniques, using both classical and machine learning algorithms to identify the relevant swimming characteristics.

We used these insights to develop sophisticated algorithms to translate data from the tracker’s accelerometer and gyroscope into reliable measurements of lap count, lap time, stroke type and stroke count that are displayed on the Samsung Gear Fit2. Users can then seamlessly transfer their data on to the Speedo On web platform, which acts as a comprehensive training partner, offering additional analysis of swims and tracking progress against fitness goals.

The value we created

The first stage of the development was completed in time for launch and the ‘Speedo On’ swim tracking app was successfully pre-installed on the Gear Fit2 Pro. It included training programs, tips and tailored advice developed by some of the world’s best coaches and athletes, all enabled by our algorithms.