Workout AI Monitor 💪

🎯 Project Vision

We're transforming manufacturing equipment monitoring by adapting advanced movement detection technology, initially proven through exercise tracking, to create an innovative solution for real-time machinery health monitoring.

💡 Innovation Highlight

Our unique approach demonstrates how human movement analysis technology can be adapted to monitor mechanical movements, providing:

• Real-time anomaly detection
• Pattern-based prediction
• Visual movement analysis

🎥 Demo Setup

⚙️ Current Implementation

Demo Features

Our prototype currently demonstrates capability through:

• Real-time movement tracking using ML Kit
• Pattern recognition and analysis
• Instant anomaly detection

Key endpoints for integration:

Base URL: https://backend-workout-ai.vercel.app/api

Authentication:
POST /userInfo    - Register monitoring system
POST /login       - System authentication
DELETE /logout    - End monitoring session

Monitoring:
POST /pushup      - Submit movement data
GET /getWoInfo    - Retrieve monitoring history

🚀 Technology Stack

Core Technologies

• Frontend: Flutter
• Vision Analysis: Google ML Kit
• Pattern Recognition: Custom algorithms
• Real-time Processing: Native integration

Key Components

• Movement pattern detection
• Real-time analysis engine
• Performance analytics
• Alert system

📊 Technical Demonstration

Pattern Recognition

// Example of our pattern detection system
final bloc = BlocProvider.of<PushUpCounter>(context);
      for (final pose in widget.posePainter!.poses) {
        PoseLandmark getPoseLandmark(PoseLandmarkType type1) {
          final PoseLandmark joint1 = pose.landmarks[type1]!;
          return joint1;
        }

        p1 = getPoseLandmark(PoseLandmarkType.rightShoulder);
        p2 = getPoseLandmark(PoseLandmarkType.rightElbow);
        p3 = getPoseLandmark(PoseLandmarkType.rightWrist);
      }
      if (p1 != null && p2 != null && p3 != null) {
        final rtaAngle = utils.angle(p1!, p2!, p3!);
        final rta = utils.isPushUp(rtaAngle, bloc.state);
        print("Angle: ${rtaAngle.toStringAsFixed(2)}");
        if (rta != null) {
          if (rta == PushUpState.init) {
            bloc.setPushUpState(rta);
          } else if (rta == PushUpState.complete) {
            bloc.incrementCounter();
            bloc.setPushUpState(PushUpState.neutral);
          }
        }
      }

Monitoring History

// Example of our monitoring history
GET https://backend-workout-ai.vercel.app/api/getWoInfo

1. Install dependencies:

flutter pub get

2. Run the application:

flutter run

3. Access the monitoring dashboard:
   • Launch application
   • Login with provided credentials
   • Start monitoring session

🔄 Manufacturing Adaptation

Our exercise monitoring system demonstrates key capabilities needed for equipment monitoring:

Exercise Feature	Manufacturing Application
Movement Detection	Equipment Operation Tracking
Form Analysis	Movement Pattern Analysis
Rep Counting	Cycle Monitoring

📈 Business Impact

Cost Reduction

• Minimize unplanned downtime
• Reduce maintenance costs
• Prevent major equipment failures

Efficiency Improvement

• Real-time monitoring capability
• Predictive maintenance
• Performance optimization

🛠️ Project Structure

lib/
├── models/         # Data models
├── services/       # API integration
├── utils/          # Analysis tools
├── views/          # Monitoring interfaces
├── widgets/        # UI components
└── painters/       # Visual overlays

🎯 Future Development

Planned Enhancements

• Equipment-specific adaptations
• Advanced pattern recognition
• IoT sensor integration
• Expanded analytics dashboard

🤝 Contact

For hackathon-related inquiries:

• Team Name: [Workout.ai]
• Contact: [[email protected]]

📝 License

This project is licensed under the MIT License

Link Github

Github