How anshumanfauzdar/predictive-maintenance works
Project Type and Purpose
Project Type:
This project is an embedded mechatronics system designed for predictive maintenance. It integrates various hardware components like microcontrollers, sensors, and actuators with software firmware to monitor and analyze industrial equipment in real-time.
Purpose:
The system aims to provide real-time monitoring of machinery by collecting data from cameras, microphones, and temperature sensors. It leverages WiFi connectivity to send data to the cloud for further analysis and implements machine learning algorithms on the edge for immediate detection of anomalies such as noise, vibration, or temperature irregularities. Additionally, it includes a modular 3D-printed structure for housing the hardware components and a Human-Machine Interface (HMI) for data visualization.
File Structure
Top-Level Directories and Files:
• .github/workflows: Contains CI/CD pipelines for Arduino and Nuttx builds, indicating automated testing and deployment processes.
• 3D: Houses 3D-printed STL files for various components, suggesting a focus on custom hardware enclosures and mounts.
• firmware: Central directory for all firmware-related code, structured into testing, Python scripts, and WiFi configurations.
• LICENSE & README.md: Standard project documentation and licensing.
Key Observations:
• Modular Structure: The presence of separate directories for firmware testing and WiFi configurations points towards a modular architecture, facilitating scalability and maintenance.
• Use of Microcontrollers: Directories like nodemcu and spresense indicate the use of specific microcontrollers for different functionalities.
• Automation: GitHub workflows for Arduino and Nuttx suggest automated building and testing processes, enhancing reliability.
• 3D Integration: The 3D directory with multiple STL files highlights the integration of hardware design with the software system.
README for Additional Insights
• Promised Features: The README outlines features like camera monitoring, noise detection, WiFi connectivity, temperature sensing, speaker notifications, HMI, modular 3D-printed structure, and cloud server integration.
• Development Lifecycle: Emphasis on using Git for version control and structured branching strategies indicates a collaborative and organized development approach.
• Road Map: Details the step-by-step development tasks, emphasizing critical components like bootloader burning, camera setup, WiFi connectivity, sensor integration, and deployment to Edge Impulse for machine learning.
a. Identify the Main Components:
• Hardware Components:
• Microcontrollers: ESP8266/32, NodeMCU, Spresense
• Sensors: Camera, Microphone, Temperature Sensor
• Actuators: Speakers
• HMI Display: For data visualization
• 3D-Printed Structure: Modular enclosures and mounts
• Connectivity Modules: WiFi modules
• Software Components:
• Firmware: For sensor data acquisition and processing
• Testing Scripts: Python-based GUI and data handling
• Machine Learning Models: For anomaly detection on the edge
• Cloud Server: For data storage and advanced analytics
• External Services:
• Edge Impulse: For deploying machine learning models
• Cloud Services: For data aggregation and remote monitoring
b. Determine Relationships and Interactions:
• Data Flow:
1. Sensors collect real-time data (images, noise levels, temperature).
2. Microcontrollers process this data locally, performing initial filtering.
3. Processed data is sent via WiFi modules to the cloud server.
4. Machine Learning Models on the edge analyze data for anomalies.
5. Upon detection, actuators (speakers) provide immediate notifications.
6. HMI Display presents real-time and historical data to users.
7. Data and insights are stored and further processed in the cloud for long-term analysis and predictive maintenance.
c. Highlight Architectural Patterns or Design Principles:
• Modular Architecture: Separates hardware components and firmware into distinct modules, facilitating easy updates and scalability.
• Edge Computing: Implements machine learning directly on microcontrollers for real-time anomaly detection, reducing latency.
• Separation of Concerns: Differentiates between data acquisition, processing, and user interface layers.
• Automated CI/CD Pipelines: Ensures consistent and reliable firmware deployment through GitHub workflows.
d. Include Relevant Technologies, Frameworks, or Libraries:
• Firmware Development: Arduino IDE for ESP8266/32 and NodeMCU, possibly using PlatformIO for Spresense.
• Programming Languages: C/C++ for firmware, Python for testing scripts.
• Machine Learning: Edge Impulse for deploying models on microcontrollers.
• Version Control: Git and GitHub for source code management and CI/CD.
• 3D Design Software: For creating and modifying STL files used in the 3D-printed structures.
Elements
• Clear Labels: Name each component accurately (e.g., ESP32 Microcontroller, Camera Module, Cloud Server).
Importance of Detail and Architectural Elements
• Including All Major Components: Don’t omit any critical hardware or software elements.
• Capturing Interactions: Clearly show how components interact, including data dependencies and communication protocols.
• Avoiding Overcomplication: While detail is essential, maintain clarity by organizing components logically and avoiding clutter.
• Documenting Design Choices: Use annotations or legends to explain significant architectural decisions or patterns.
Final Recommendations
• Iterate and Review: Share initial drafts with team members to ensure accuracy and completeness.