Creation of the digital twin model – Athlete Passport
To generate the digital twin of an individual’s response to exercise, we use data collected from historic files, or from a ‘training period’ of several months. We can build this using historical smartwatch data alone, however precision increases with the use of additional sources, including subjective ratings from Readiness Advisor, data from sleep recording devices (e.g. Oura ring), or Continuous Glucose Monitors.
Using leading scientific models and our advanced algorithms, we estimate an individual’s maximal heart rate and lactate threshold, fundamentals for correct fitness assessment. Our algorithms also adjust for external factors, such as terrain, temperature, altitude, and fatigue, to ensure accurate and comprehensive fitness assessments.
With these pieces in place, our algorithms calculate which metrics and correlations are most important for the individual’s performance, recovery and avoidance of injury or illness
Simulation of potential paths – Mimir
With the Athlete Passport in place, our powerful analytics engine uses this profile of known response to simulate what happens when we make changes to the training or wellness regime. A cyclist might wonder how their performance might change if they eat an extra gel during a climb. A runner using svexa’s Ellida might want to know how their training could progress if they add an extra training session per week. A home gym user might be interested in the impact on their aging process if they focus on more strength training for the upcoming 5 years.
Accurate & Adaptive
Our digital twins are not limited to static models based on initial data. Instead, they are dynamic and continuously updated, utilizing information from the wearable devices on progression and compliance during ongoing training. This allows us to incorporate recent data, such as a poor night’s sleep, an elevated resting heart rate, or a report of low mood, to make informed decisions about the training session for upcoming days.