Endurance by Endulytics – fitness app with wearables & training plans

Accuracy, device diversity, and athlete expectations

Endurance athletes expect reliable real-time metrics and trustworthy history: heart rate, power (watts), cadence, and distance must stay stable outdoors, indoors, and when connectivity fluctuates. Sensor vendors implement different BLE GATT profiles and firmware behaviours; pairing, background reconnects, and multi-sensor setups do not behave identically across devices.

The app must also work when the network drops: recorded sessions cannot be lost if the user is briefly offline. When WiFi or mobile data returns, data should merge with the backend without duplicates or manual fixes.

Battery use, background work, and privacy

Continuous BLE scanning and connections, combined with screen-on use and occasional GPS scenarios, stress CPU and battery. iOS and Android apply different background limits—a flow that works on one test device may drop connections on older phones or under load. Health and performance data must be processed in line with GDPR, from on-device storage to synced history in the cloud.

Native clients, BLE layer, and robust sensor integration

We implemented the Endurance app natively for iOS (Swift) and Android (Kotlin), wiring the BLE stack per platform: scan intervals, reconnect strategies, and cached device metadata are tuned so typical combinations of chest straps, power meters, and bike computers remain stable. Where vendors use proprietary extensions, we isolate logic in modular adapters so SDK or firmware updates stay manageable.

Training planning combines structured intervals, target zones, and clear day/week views. Users see at a glance whether they trained in the planned range—with practical guidance without turning the app into a medical device.

WiFi sync, API design, and offline-first behaviour

Backend sync uses encrypted REST APIs; after a session, batches upload when connectivity is stable. A local queue with idempotency keys prevents duplicate entries if uploads fail mid-way, keeping history consistent whether the athlete syncs at home on WiFi or on mobile data.

Field stability and maintainable architecture

Throughout the project we prioritised reproducible testing with real sensor sets and an architecture ready for follow-on features: more sports, deeper analytics, or third-party integrations. Collaboration with Endulytics was iterative—feedback from athletes and coaches fed into sprint planning and UX refinements.

User experience and performance

The UI targets fast pre-workout capture and clear post-workout review: minimal taps, readable metrics, and sensible defaults for zones and units. Profiling on mid-range devices keeps charts and animations smooth so the app stays responsive after long sessions.