Skip to content

AnshumanFauzdar/predictive-maintenance

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

73 Commits
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Video Demonstration

Samvedan 2021 - TEAM MECHATRONICS

Documentation hosted on sony.anshumanfauzdar.me

Engineering guidelines to be followed:

Guidelines to be followed for this project:

  • KISS - Keep it stupid simple.
  • No extra features until and unless it is a bug.
  • Revisit official documentation in case of no results found on google or forums.
  • Create safe fallback features and try to re-create until and unless it is a bug.
  • In case of creative block discuss with team ASAP.

Promised Features:

  • Camera Monitoring - Images only
  • Microphone - noise detection
  • WiFi connectivity - ESP8266/32

For predictive maintenance:

  • Temperature sensor - real time data
  • Speakers - for notifications
  • HMI - for displaying data
  • Modular structure - 3D printed + aluminum extrusion structure
  • Server - cloud implementation

Development lifecycle

All the software, CAD, electronics development will be done with git and uploaded here to GitHub

  1. Create a branch off of master:

Only in case of working on new feature not aligned with master branch. Name the branch with your first name pre-pended [do not create unnecessary branches]: spresense/cool-feature

  1. Writing Code:

GitHub upload instructions:

  • Standard PR should be made to merge features with log including changes and bug fixes.
  • To make standard changes follow engineering guidelines.
  • For making revision of CAD models mention version ex. Mount_v2.0.1
  1. Submit for review:
  • Assign the task to the appropriate reviewer.
  • Once tested and approved from all team members PR will be merged.

Software Architecture

  graph TD;
    A[Start Software] --> B{{Check Internet Connection}};
    B --> C(connected to internet);
    B --> E(Not connected to internet);
    C --> D(update to latest dataset and firmware);
    E --> F(use local dataset and firmware edgeML);
    D --> G;
    F --> G{{Data acquisition from camera and microphone/vibration/temperature sensors}};
    G --> H(Image/video data in realtime)
    G --> I(Microphone/vibration/temperature sensor real time data)
    H --> J(Monitoring with ML and AI on Edge)
    I --> K(Monitoring with noise filtering and AI/ML on the edge)
    J --> L;
    K --> L{{Leakage, excess, noise/vibration, wrong part handling, heat etc. detected}}
    L -->|YES| M{{Notify nearest production line worker and visual/audio notifications}}
    M -->|Yes| N[\Action Taken\]
    M -->|No| O[\No action taken after certain time\]
    O --> P{{MACHINE AUTO TURN OFF and notify all}}
    P --> Q[\Acquire data and improve detection dataset\]
    Q --> R[Repeat Process]
    N --> R

Road Map

  • [⚠️ VERY IMPORTANT ⚠️] Burning Bootloader
  • [⚠️ VERY IMPORTANT ⚠️] Setting up camera
  • Wi-Fi Connection - ESP8266
  • Try Examples - Wi-Fi functionality only
  • Try multicore functionality
  • Attach microphone and temperature sensors
  • Deploy to edge impulse
  • Initial image detection
  • Noise detection
  • Temperature variation detection
  • Mix all the detection in single application
  • Make webserver to show all data