I’ve been thinking a lot lately about what features we’ll see in endurance sports apps in the future and which of them will be part of my app development journey. I’ve come across the following topics:
Training Plans & Progression
Training Plan Generation:
Automatically created based on training goals, level of experience, availability, training frequency and fitness level in any kind of sport. You give the AI guidelines on how a real trainer would create the training plan, with a few elements to rely on. It could look something like this:
TRAININGPLAN METHODOLOGY:
- Follow 90/10 RULE: 90% low-moderate intensity, 10% high intensity
- Use 3:1 PERIODIZATION: 3 weeks progressive build, 1 week recovery
- PROGRESSIVE LOAD: 5-10% weekly increase in mileage
- WEEKLY STRUCTURE: 1 Anchor Quality session per week (+ optional second for advanced), 1 long run, rest easy
- HARD/EASY RULE: Always separate hard days with at least 2 easy day between
...Based on this the AI generates a readable data format to import it into the app. The difficulty here is probably ensuring the quality of the plans.
Progress Monitoring & Suggestions for adapting:
Through sports gadgets like watches we get a lot of data to analyse the progress we make with training plans. The art lies in reducing and processing the data, for example using vector databases. Progress is only visible when you compress, contextualize, and compare data over time. This data can then be used to adjust training plans.
Comparison between Feedback & Context Data
Athlete Feedback Processing:
Based on user inputs like “How do you feel today” inputs, RPE scores in workouts and personal notes suggestions for the training today and in the future were given. It will be interesting to compare the recorded workout data such as HR, power, or pace with the wellness data measured by wearables such as HRV, sleep, stress, recovery scores, and the athlete’s perception (RPE, feeling, stress). These comparisons can then be used for training in the short or long term.
Injury Risk Prediction:
The more data is feed in into the system it can predict probability of injuries and help to avoid them. The use of AI is not necessarily required for this in the first step. This can also be done on a rule-based basis by calculating a score from various data points.
Workout-Level Analytics
Session Analytics:
Session Analytics will automatically analyze a workout to provide actionable insights. It detects intervals without requiring manual input, identifies deviations from target zones by highlighting where pace, power, or heart rate exceeded or fell short of the planned ranges, and compares the intensity distribution to show how time was spent across different training zones relative to the planned session. This transforms raw workout data into meaningful feedback that helps athletes and coaches understand performance and make informed adjustments. Basically Strava AI intelligence but done right ;).
Live Adaptive Coaching:
Live Adaptive Coaching will provide real-time voice feedback during workouts through a smartwatch or headphones. The system monitors metrics such as heart rate, pace, or power and gives immediate instructions – for example, “Slow down, HR is too high” or “Decrease pace for the next interval.” This allows athletes to adjust their effort on the fly, stay within target zones, and optimize training effectiveness without constantly checking their devices.
AI Coach Interaction
Virtual Coach Chat:
Users will be able to selects their “coach type” with personalized coaching attributes like tone, philosophy and personality. Chat Interactions will be possible and help the user to make decisions about their training. Also motivation can be a big part of a virtual coach.
Hybrid Coach Mode:
The combination of AI-based data and the influence of a human coach will become the gold standard. It combines AI-driven data analysis with the guidance of a human coach. The AI continuously processes performance metrics, recovery scores, and athlete feedback to provide actionable insights and adaptive training suggestions. The human coach then focuses on higher-level decisions, motivation, and nuanced adjustments, using AI insights to make coaching more precise and personalized. This synergy allows for scalable, data-informed coaching while preserving the empathy and experience of a human trainer.
