Skip to content
Irvs EbajanIrvs
Back to ProjectsIoT + Mobile + Backend System

SikadVoltz

IoT-enabled indoor cycling ecosystem combining embedded firmware, a Flutter mobile app, and a Node.js backend

Overview

An IoT-enabled indoor cycling ecosystem combining embedded firmware, a Flutter mobile app, and a Node.js backend to deliver real-time cycling analytics, AI-driven training plans, and gamified fitness tracking.

Problem

Traditional fitness tracking platforms lack real-time hardware integration. Most cycling apps provide basic GPS and manual tracking but fail to connect with physical speed/cadence sensors or adapt intelligently to user performance.

Solution

A four-component platform:

ComponentStackRole
FirmwareESP32 + Arduino + NimBLEReads reed switch sensor; computes speed/cadence/power; broadcasts JSON telemetry at 10 Hz over BLE GATT
Mobile AppFlutter 3.16 / Dart 3.2+BLE pairing with ESP32; real-time workout tracker; goal wizard with TDEE-based planning; health screening; achievements
BackendNode.js + Express 5 + MongoDB + RedisREST API + WebSocket; auth (JWT + Google OAuth); FCM push; calorie/TDEE engine; smart plan redistribution; gamification
WebsiteReact 18 + Vite + Framer MotionPromotional landing page with app download links, team info, and public stats

System Architecture

┌─────────────┐    BLE     ┌──────────────┐   REST/WS   ┌──────────────────────┐
│  ESP32       │◄─────────►│  Flutter App  │◄───────────►│  Node.js Backend     │
│  (Reed Sw.)  │  10Hz JSON│  (Mobile)     │   HTTPS     │  (Express + WS)      │
└─────────────┘            └──────────────┘             └──────┬───────────────┘
                                                               │
                                                    ┌──────────▼──────────┐
                                                    │  MongoDB Atlas      │
                                                    │  + Redis (Cache)    │
                                                    └─────────────────────┘

Key Features

  • ▸Live cycling tracker — real-time speed, cadence, power, distance via BLE
  • ▸Smart session management — auto pause/resume via freewheeling detection on ESP32
  • ▸AI training plans — goal wizard with TDEE-based calorie targets and dynamic plan redistribution for missed sessions
  • ▸Gamification — XP, levels, ranks, streaks, badges, milestones, quests
  • ▸Dual notification strategy — WebSocket (foreground) + Firebase FCM (background)
  • ▸Health screening — PAR-Q+ risk assessment before goal creation
  • ▸Offline resilience — ESP32 EEPROM buffering + Flutter offline storage and deferred sync
  • ▸Google Calendar sync — plan integration with external calendars
  • ▸Achievement engine — badge and quest system driven by service-layer logic

Deployment & CI/CD

  • ▸Backend: Render (Singapore, Docker multi-stage, HTTP/2 with fallback, health checks)
  • ▸Website: Cloudflare Pages or Render static site
  • ▸CI/CD: GitHub Actions — lint, test matrix (Node 16/18/20), security scan (Trivy, Snyk, CodeQL), Docker buildx multi-arch (amd64 + arm64), staging → production deploy

Technology Stack

Firmware

ESP32ArduinoNimBLE

Mobile App

Flutter 3.16Dart 3.2+

Backend

Node.jsExpress 5MongoDBRedis

Website

React 18ViteFramer Motion

DevOps

RenderGitHub ActionsDockerCloudflare Pages

Notable Design Decisions

  • ▸Unified dashboard endpoint — single Redis-cached call replaces ~5 separate API calls (~75% latency reduction)
  • ▸Dual Redis clients — SessionManager + simpleRedisClient for resilience and debug comparison
  • ▸MongoDB timeseries — telemetry stored in native timeseries collections for efficient high-frequency sensor queries
  • ▸API versioning — all routes under /api/v1/ with legacy backward compatibility
  • ▸Service orchestrator — Flutter UnifiedServiceOrchestrator manages init order and prevents conflicts
  • ▸UI_ONLY_MODE — feature flag for frontend development without backend dependencies
  • ▸Startup optimization — minimal critical path first, deferred background services, error-tolerant fallback UI